American Journal of Physical Medicine and Rehabilitation · American Journal of Physical Medicine...

American Journal of Physical Medicine and

Rehabilitation

December 2007, Volume 86, Issue 12,PP 957-1038

Research Article

957 Musculoskeletal Disorders in Referrals for Suspected Lumbosacral Radiculopathy. Daniel E. Cannon, BS; Timothy R. Dillingham, MD; Haiyan Miao, MS; Michael T. Andary, MD; Liliana E. Pezzin, PhD

Research Article

962 Understanding Pain After Traumatic Brain Injury: Impact on Community Participation. Jeanne M. Hoffman, PhD; Kathleen F. Pagulayan, PhD; Nadya Zawaideh, BA; Sureyya Dikmen, PhD; Nancy Temkin, PhD; Kathleen R. Bell, MD

Research Article

970 Myopathic Dropped Head Syndrome: An Expanding Clinicopathological Spectrum. Jenny P. Liao, MD; Andrew J. Waclawik, MD; Barend P. Lotz, MD; Sharhriar M. Salamat, MD, PhD; Brad R. Beinlich, MD; Benjamin R. Brooks, MD

Research Article

977 Upper-Limb Prosthetics: Critical Factors in Device Abandonment. Elaine Biddiss, PhD; Tom Chau, PhD

Research Article

988 A Dynamic Seating Intervention for Wheelchair Seating Discomfort. Barbara A. Crane, PhD, PT, ATP; Margo B. Holm, PhD, OTR/L; Douglas Hobson, PhD; Rory A. Cooper, PhD; Matthew P. Reed, PhD

CME Article * 2007 Series * Number 5

994 CME Cyproheptadine for Pediatric Intrathecal Baclofen Withdrawal: A Case Report. Joseph A. Saveika, MD; Jean E. Shelton, MD

Media Review

997 Physical Medicine and Rehabilitation Board Review: Palm Edition. Lam Nguyen, DO

Research Survey

1001 Perceived Preparedness for Physiatric Specialization and Future Career Goals of Graduating Postgraduate Year IV Residents during the 2004-2005 Academic Year. Vishwa S. Raj, MD; Diana H. Rintala, PhD

Review & Analysis

1007 Outcome Assessment in Randomized Controlled Trials of Stroke Rehabilitation. Katherine L. Salter, BA; Robert W. Teasell, MD; Norine C. Foley, MSc; Jeffrey W. Jutai, PhD

Review & Case Reports

1013 Persistent Hiccups During Rehabilitation Hospitalization: Three Case Reports and Review of the Literature. John A. Schuchmann, MD; Barry A. Browne, PharmD

Case Report

1019 Persistent Hiccup Associated with Intrathecal Morphine Infusion Pump Therapy. Xiulu Ruan, MD; John Patrick Couch, MD; Rinoo Shah, MD; Frank Wang, MD; Hai Nan Liu, MS

Commentary

1023 Ashley X. Kristi L. Kirschner, MD; Rebecca Brashler, LCSW; Teresa A. Savage, PhD, RN

Visual Vignette

1030 Exacerbation of Habitual Dislocation of Ulnar Nerve by Concurrent Dislocation of Triceps Muscle: Complementary Role of Dynamic Ultrasonography to Electrodiagnosis. Myung Jae Yoo, MD; Dennis DJ Kim, MD; Mooyeon Oh-Park, MD

Letters to the Editor

1031 Re: Introduction to Nanotechnology: Potential Applications in Physical Medicine and Rehabilitation. Govindasamy Agoramoorthy; Chiranjib Chakraborty

1032 Re: Introduction to Nanotechnology: Potential Applications in Physical Medicine and Rehabilitation: Response Letter. Assaf T. Gordon, MD

1032 Re: Sacroiliac Joint Pain: Anatomy, Biomechanics, Diagnosis, and Treatment. Myron M. LaBan, MD, MMSc, FACP, FAAPMR

1033 Re: Sacroiliac Joint Pain: Anatomy, Biomechanics, Diagnosis, and Treatment: Response Letter. Brian S. Foley, MD; Ralph M. Buschbacher, MD

1033 Re: The Myofascial Trigger Point Region: Correlation Between the Degree of Irritability and the Prevalence of Endplate Noise. Jay P. Shah, MD; Sagar Parikh, MS IV; Jerome Danoff, PT, PhD; Lynn H. Gerber, MD

1034 Re: The Myofascial Trigger Point Region: Correlation Between the Degree of Irritability and the Prevalence of Endplate Noise: Response Letter. Chang-Zern Hong, MD; Ta-Shen Kuan, MD

1034 Re: Characteristics and Treatment of Headache After Traumatic Brain Injury. Myron M. LaBan, MD, MMSc

1035 Re: Characteristics and Treatment of Headache After Traumatic Brain Injury: Response Letter. Henry L. Lew, MD, PhD; David J Clark, MD, PhD; William C. Walker, MD

1035 Re: Shoulder Impingement Syndrome: Relationships Between Clinical, Functional, and Radiologic Findings. Ilknur Aktas, MD; Kenan Akgun, MD

1036 Re: Shoulder Impingement Syndrome: Relationships Between Clinical, Functional, and Radiologic Findings: Response Letter. F Figen Ayhan, MD; Z Rezan Yorgancioglu, MD

Erratum

1038 Erratum.

Authors:Daniel E. Cannon, BSTimothy R. Dillingham, MDHaiyan Miao, MSMichael T. Andary, MDLiliana E. Pezzin, PhD

Affiliations:From the Medical College ofWisconsin, Milwaukee, Wisconsin(DEC); Departments of PhysicalMedicine and Rehabilitation (TRD)and Medicine (HM, LEP), MedicalCollege of Wisconsin, Milwaukee,Wisconsin; and Department ofPhysical Medicine and Rehabilitation,Michigan State University College ofOsteopathic Medicine, East Lansing,Michigan (MTA).

Correspondence:All correspondence and requests forreprints should be addressed toTimothy R. Dillingham, Departmentof Physical Medicine andRehabilitation, 9200 W WisconsinAvenue, Milwaukee, WI 53226-3596.

0894-9115/07/8612-0957/0American Journal of PhysicalMedicine & RehabilitationCopyright © 2007 by LippincottWilliams & Wilkins

DOI: 10.1097/PHM.0b013e31815b614a

Musculoskeletal Disorders inReferrals for SuspectedLumbosacral Radiculopathy

ABSTRACT

Cannon DE, Dillingham TR, Miao H, Andary MT, Pezzin LE: Musculoskeletaldisorders in referrals for suspected lumbosacral radiculopathy. Am J Phys MedRehabil 2007;86:957–961.

Objective: An electrodiagnostic evaluation is often requested for pa-tients with suspected lumbosacral radiculopathy. Although musculoskele-tal disorders can produce lower-limb symptoms, their prevalence in thisreferral population is unknown. The purpose of this study was to determinethe prevalence of common lower-limb musculoskeletal disorders in pa-tients referred for lower-limb electrodiagnosis and determine whetherthese findings predict study outcome.

Design: Subjects undergoing an electrodiagnostic study for lower-limbsymptoms and suspected lumbosacral radiculopathy constituted the sam-ple. A standardized clinical and electrodiagnostic evaluation was per-formed for all patients.

Results: There were 170 subjects included. The mean age was 52(SD � 17), and 45% were males. The total prevalence of musculoskel-etal disorders in the sample was 32%. The prevalence in those with anormal study was 55% as compared with 21% in those with lumbosacralradiculopathy (P � 0.0001).

Conclusions: Musculoskeletal disorders are common in patients sus-pected of having lumbosacral radiculopathy. The high prevalence amongboth patients with normal studies and those with radiculopathy and otherdisorders limits the usefulness of this information in predicting studyoutcome. In particular, it is common for patients to have two or moreproblems and the presence of a musculoskeletal disorder should notpreclude such testing.

Key Words: Radiculopathy, Electrodiagnosis, Musculoskeletal Diseases, MyofascialPain Syndromes

December 2007 Musculoskeletal Disorders in Radiculopathy 957

RESEARCH ARTICLE

Musculoskeletal

A large number of conditions are included inthe differential diagnosis of patients referred forelectrodiagnostic testing. Suspected radiculopathyis a common reason patients are referred to anelectrodiagnostic laboratory.1,2 Lumbosacral radic-ulopathy is a pathologic condition involving thelumbosacral nerve roots. The most common symp-tom is pain that is often increased or precipitatedby standing or sitting. Other symptoms that canoccur are parasthesias and weakness in a part of thelimb. The distribution of the symptoms depends onthe nerve roots involved.3 Likewise, musculoskele-tal conditions are common causes of lower-limbsymptoms and may be mistaken for lumbosacralradiculopathy.4–7

Myofascial pain is characterized by pain inskeletal muscles that originates from triggerpoints, areas within the muscle that are highlysensitive and reproduce the patient’s symptomswhen palpated.8 Trochanteric bursitis is a condi-tion of inflammation or irritation of either thegluteus maximus bursa or the bursa separating thegreater trochanter from the gluteus medius andgluteus minimus. It presents with pain over thelateral aspect of the hip.9 Iliotibial band syndromeoccurs when the iliotibial band, a dense fascia onthe lateral aspect of the knee and hip, impingesover the lateral femoral epicondyle. It most com-monly presents with pain over the lateral knee thatcan radiate up into the lateral thigh or down toGerdy tubercle.10 In the present analysis, trochan-teric bursitis and iliotibial band syndrome aregrouped together. Plantar fasciitis is an inflamma-tory condition of the plantar aponeurosis of thefoot. It typically presents with tenderness on pal-pation of the sole anterior to the calcaneus.11

The purpose of this study was to determine theprevalence of these common musculoskeletal disor-ders of the lower limb in patients referred for elect-rodiagnostic testing. In addition, whereas previousstudies have examined the influence of physical examfindings on predicting electrodiagnostic study out-comes,12,13 in this study the influences of these spe-cific musculoskeletal disorders on electrodiagnosticstudy outcome prediction were examined.

METHODSSubjects

A prospective study was conducted betweenJuly 1996 and April 1998 at the following medicalcenters that treat diverse patient populations andhave high-volume Physical Medicine and Rehabil-itation electrodiagnostic services: Johns HopkinsUniversity, Baltimore, MD; Ingham Regional Med-ical Center, associated with Michigan State Univer-sity, East Lansing, MI; Madigan Army Medical Cen-

ter, Tacoma, WA; Womack Army Medical Center,Fort Bragg, NC; and Walter Reed Army MedicalCenter, Washington, DC. Patients were asked toparticipate if they were referred for lower-limbsymptoms or suspected lumbosacral radiculopathyto the electrodiagnostic laboratory in the depart-ment of physical medicine and rehabilitation at oneof the participating institutions. Subjects partici-pated on a voluntary basis after giving writteninformed consent. The study protocol was ap-proved by the appropriate institutional review com-mittee at each of the participating institutions.

Data CollectionStandardized procedures and data collection

sheets were implemented at all participatingcenters. Patients completed a questionnaire thatcollected detailed information on symptoms andmedical history. For all patients, a standardizedphysical examination was completed that consistedof the following: (1) a neurologic examination,including manual muscle testing, sensation (vibra-tion and pinprick), and reflex assessments; (2) amusculoskeletal examination; and (3) special tests,such as the straight-leg raise, and the Patrick test.After the questionnaire and physical examination,electrodiagnostic evaluations, conducted by at-tending physicians and their supervised residentphysicians, were performed for all patients. All par-ticipating attending physicians were certified bythe American Board of Electrodiagnostic Medicineand were skilled musculoskeletal physicians. Suchexpertise was sought to ensure electrodiagnosticand physical exam findings across participatingcenters were consistently interpreted.

Musculoskeletal disorders were diagnosed ifcertain criteria were met in the course of the phys-ical exam. Myofascial pain was diagnosed if palpi-tation of the back or buttock region while avoidingthe area of the sciatic notch reproduced symp-toms.8 Trochanteric bursitis/iliotibial band syn-drome was diagnosed if palpation in the lateral hipregion reproduced symptoms.14,15 Plantar fasciitiswas diagnosed if palpation of the sole at the originof the plantar fascia at the calcaneus reproducedsymptoms.16

The standardized electrodiagnostic study con-sisted of at least (1) one lower-limb motor nerveconduction study, (2) one lower-limb sensorynerve conduction study, and (3) needle EMG witheither monopolar or concentric needles of a stan-dard set of 11 muscles. The electrodiagnosticconclusions were classified into three mutuallyexclusive categories: normal study, lumbosacral ra-diculopathy, or other condition. A stratified data-collection strategy was used to recruit sufficientpersons in each category because these proportionsin the referral population were different.

958 Cannon et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

Lumbosacral radiculopathy was diagnosed us-ing methods established earlier.17 Specifically,lumbosacral radiculopathy was diagnosed if EMGfindings (spontaneous activity, increased polypha-sics, complex repetitive discharges, or reduced re-cruitment) were found in two or more musclesinnervated by the same nerve root but differentperipheral nerves, or if the paraspinal musclesshowed spontaneous activity. Overall, 96% of thisgroup had spontaneous activity in at least onemuscle. Other diagnoses included sciatic neuropa-thy, femoral neuropathy, tibial neuropathy, pero-neal neuropathy, polyneuropathy, myopathy, andMyasthenia Gravis. The electrodiagnostic out-comes were then analyzed as they related to thepresence of musculoskeletal disorders.

Data AnalysisSAS Windows version 9.1 was used to compare

the prevalence of musculoskeletal disorders be-tween study groups using �2 analysis or forwardtwo-sided exact test. To further examine the influ-ence of musculoskeletal conditions on study out-come, a multivariate analysis was undertaken us-ing STATA 9.0. Age, gender, and duration ofsymptoms were controlled for to prevent any po-tential confounding effects these variables mayhave had on the outcome of this analysis. Studyoutcomes were defined as normal study (referencecategory), lumbosacral radiculopathy, or other di-agnosis.

RESULTSThere were 170 subjects included in the study.

The mean age of the subjects was 52 (SD � 17) and45% were men. The mean duration of symptomsreported by the subjects was 26 mos (SD � 54).The prevalence of musculoskeletal disorders isshown in Table 1. Those with a normal studytended to have higher prevalence of the conditionsstudied. However, persons with electrodiagnosti-

cally confirmed conditions also had these disor-ders, but at lower prevalence than those with nor-mal studies.

The only significant finding from the multi-variate analysis showed that the presence of myo-fascial pain meant about a fivefold lower likelihoodof having lumbosacral radiculopathy comparedwith in a normal study (P � 0.001), given thatother variables in the model were held constant.

DISCUSSIONMusculoskeletal disorders are common in pa-

tients suspected of having lumbosacral radiculop-athy. However, within the study sample there weresignificant differences in the prevalence of muscu-loskeletal disorders when subjects were groupedaccording to the outcome of their electrodiagnosticstudies. More than half (55%) of those demonstrat-ing a normal electrodiagnostic study had a muscu-loskeletal disorder diagnosed on the basis of phys-ical exam findings. In comparison, there was a 21%prevalence of musculoskeletal disorders in thosewith lumbosacral radiculopathies (P � 0.0001) anda 29% prevalence in those with other electrodiag-nostic diagnoses.

Whereas the presence of a musculoskeletal dis-order suggests that a patient will have a normalstudy, the fairly high prevalence of musculoskeletaldisorders in those with a confirmed lumbosacralradiculopathy or other electrodiagnostic conclu-sions limits the usefulness of this information inpredicting the results of an electrodiagnostic study.A study should not be curtailed solely on the basisof finding a musculoskeletal disorder on physicalexam.

Myofascial pain in particular was common insubjects that had a normal study (39%), signifi-cantly more so than in those with a lumbosacralradiculopathy (12%, P � 0.0001) and those withanother electrodiagnostic diagnosis (12%, P �0.0362). This result could be an indication of the

TABLE 1 Prevalence of musculoskeletal disorders in patients with suspected lumbosacralradiculopathy

Musculoskeletal DisorderNormal Study, %

(n � 51)

LumbosacralRadiculopathy, %

(n � 102)

Other EDXDiagnosis,‡ %

(n � 17)Total Sample, %

(n � 170)

Myofascial pain syndrome 39 12* 12* 20Trochanteric bursitis/iliotibial

band syndrome29 12* 18 18

Plantar fasciitis 4 3 18† 5One or more of the above 55 21* 29 32

* Significant at P � 0.05 compared with normal study.† Significant at P � 0.05 compared with lumbosacral radiculopathy.‡ Other diagnoses included sciatic neuropathy, femoral neuropathy, tibial neuropathy, peroneal neuropathy,

polyneuropathy, myopathy, and Myasthenia Gravis.


difficulty in distinguishing between myofascialpain and radiculopathy and the degree to which thetwo disorders mimic each other. The referred painpatterns seen in myofascial pain can be very similarto the dermatomal patterns of the spinal nerveroots and are poorly recognized by most practitio-ners.4 Therefore, it is possible that many patientsbeing referred for electrodiagnostic evaluations ac-tually have myofascial pain, either alone or alongwith radiculopathy. If this is the case, an electro-diagnostic evaluation can help clarify the nature ofthe problem.

Trochanteric bursitis/iliotibial band syndromeshowed a significantly lower prevalence in subjectswith lumbosacral radiculopathies (12%, P � 0.007)than in those with normal studies (29%). It ispossible that this finding reflects difficulty in dis-tinguishing between these conditions and lumbo-sacral radiculopathy based solely on physical examfindings, as was proposed with myofascial painabove. Both cases emphasize the importance ofelectrodiagnostic testing in helping distinguishmusculoskeletal disorders from lumbosacral radic-ulopathy when evidence is present for both andtreatment strategies for the musculoskeletal con-ditions are not effective. However, electrodiagnos-tic studies are not advocated for all patients withmusculoskeletal disorders. Rather, the clinicianshould consider the possibility of radiculopathy orother disorders even in the presence of such con-ditions.

There are some potential limitations to thisstudy. It is possible that there was heightened sus-picion for certain musculoskeletal conditions giventhe nature of the prospective study and the carefuland systematic examination for these conditions.In addition, there are interrater differences inphysical examination even for skilled practitionersthat could influence our findings. For example, theexact finger pressure used to elicit findings duringpalpation was not standardized. The underlyingpain system and peripheral nervous system physi-ology is quite complex and interrelated. These sys-tems may augment the pain responses and percep-tions in persons with radiculopathies such thatphysical exam maneuvers are perceived as morepainful than they otherwise might be in personswith radiculopathy.

Finally, not all radiculopathies that producesymptoms can be detected by an electrodiagnosticevaluation. For example, radiculopathies that in-volve only sensory nerve roots will not produceabnormal EMG findings.18 Also, some radiculopa-thies may have associated myofascial pain as acomponent of their disease process. Electrodiag-nostically proven radiculopathies with substantialaxonal loss may have myotomal pain in the affectedmuscles that may mimic more typical myofascial

pain. Each of these factors may have contributed tosome of the data being misclassified with respect tomusculoskeletal conditions present, electrodiag-nostic conclusions, or both.

The multivariate analysis showed that thepresence of myofascial pain meant that a patientwas about five times less likely to have lumbosacralradiculopathy than a normal study. This provides aplausible indication of what the results of electro-diagnostic testing would be in a patient diagnosedwith myofascial pain.

These findings are also in line with what wasconcluded in other studies regarding the predictiveability of physical exam on electrodiagnostic out-come. Physical exam findings are important andhelpful in guiding the course of an electrodiagnos-tic study, but are not sufficient to predict the out-come of the study.12,13 Similarly, the presence ofmusculoskeletal disorders gives some indicationabout what the outcome of an electrodiagnosticstudy will be but does not provide enough infor-mation to definitively predict the results of thestudy.

CONCLUSIONSMusculoskeletal disorders are common in pa-

tients referred for electrodiagnostic assessment.However, the presence of a musculoskeletal prob-lem does not accurately predict who will have anormal electrodiagnostic study or an electrodiag-nostically confirmed lumbosacral radiculopathywith sufficient discriminative ability to curtail theelectrodiagnostic evaluation. Whereas there is asignificantly higher prevalence in those with a nor-mal study compared with those with lumbosacralradiculopathy, the fairly high prevalence in allgroups makes it difficult to predict the outcome ofelectrodiagnostic testing based on the presence ofmusculoskeletal disorders.

REFERENCES1. Levin KH: Electrodiagnostic approach to the patient with

suspected radiculopathy. Neurol Clin N Am 2002;20:397–421

2. Dumitru D, Zwarts MJ: Radiculopathies, in Dumitru D,Amato AA, Zwarts MJ (eds): Electrodiagnostic Medicine, ed2. Philadelphia, Hanley and Belfus, 2002, pp 713–76

3. Ellenburg M, Honet JC: Lumbar radiculopathy, in FronteraW, Silver J (eds): Essentials of Physical Medicine and Re-habilitation, ed 1. Philadelphia, Hanley and Belfus, 2002, pp243–8

4. Lauder TD: Musculoskeletal disorders that frequentlymimic radiculopathy. Phys Med Rehabil Clin N Am 2002;13:469–85

5. Facco E, Ceccherelli F: Myofascial pain mimicking radicularsyndromes. Acta Neurochir Suppl 2005;92:147–50

6. Flax HJ: Myofascial pain syndromes—the great mimicker.Bol Asoc Med P R 1995;87:167–70

7. Keplinger FS, Gupta N: Trochanteric bursitis, in FronteraW, Silver J (eds): Essentials of Physical Medicine and Re-

960 Cannon et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

habilitation, ed 1. Philadelphia, Hanley and Belfus, 2002, pp297–300

8. Borg-Stein J, Simons DG: Myofascial pain. Arch Phys MedRehabil 2002;83:S40–7

9. Anderson M: Trochanteric bursitis, in Ferri FF (ed): Ferri’sClinical Advisor: Instant Diagnosis and Treatment, ed 8. St.Louis, Mosby, 2006, p 867

10. Akuthota V, Stilp SK, Lento P: Iliotibial band syndrome, inFrontera W, Silver J (eds): Essentials of Physical Medicineand Rehabilitation, ed 1. Philadelphia, Hanley and Belfus,2002, pp 328–33

11. Glazer JL, Hosey RG: Soft-tissue injuries of the lower ex-tremity. Prim Care Clin Office Pract 2004;31:1005–24

12. Lauder TD, Dillingham TR, Andary M, et al: Effect of historyand exam in predicting electrodiagnostic outcome amongpatients with suspected lumbosacral radiculopathy. Am JPhys Med Rehabil 2000;79:60–8

13. Lauder TD: Physical examination signs, clinical symptoms,and their relationship to electrodiagnostic findings and thepresence of radiculopathy. Phys Med Rehabil Clin N Am2002;13:451–67

14. Shbeeb MI, Matteson EL: Trochanteric bursitis (greatertrochanter pain syndrome). Mayo Clin Proc 1996;71:565–9

15. Khaund R, Flynn SH: Iliotibial band syndrome: a commonsource of knee pain. Am Fam Physician 2005;71:1545–50

16. Aldridge T: Diagnosing heel pain in adults. Am Fam Phy-sician 2004;70:332–8

17. Dillingham TR, Lauder TD, Andary M, et al: Identifyinglumbosacral radiculopathies: an optimal electromyographicscreen. Am J Phys Med Rehabil 2000;79:496–503

18. Wilbourn AJ, Aminoff MJ: AAEM minimonograph 32: elec-trodiagnosis of radiculopathies. Muscle Nerve 1998;21:1612–31


Authors:Jeanne M. Hoffman, PhDKathleen F. Pagulayan, PhDNadya Zawaideh, BASureyya Dikmen, PhDNancy Temkin, PhDKathleen R. Bell, MD

Affiliations:From the Departments ofRehabilitation Medicine (JMH, KFP,NZ, SD, NT, KRB) and Department ofNeurological Surgery (SD, NT),University of Washington, Seattle,Washington.

Correspondence:All correspondence and requests forreprints should be addressed toJeanne Hoffman, PhD, University ofWashington School of Medicine,Department of RehabilitationMedicine, 1959 NE Pacific Street, Box356490, Seattle, WA 98195.

Disclosures:Supported by the National Instituteon Disability and RehabilitationResearch, U.S. Department ofEducation (Grant No. H133A020508).

Part of the material presented in thismanuscript was presented at theAmerican Congress of RehabilitationMedicine Conference, Chicago,October, 2005.


DOI: 10.1097/PHM.0b013e31815b5ee5

Understanding Pain After TraumaticBrain InjuryImpact on Community Participation

ABSTRACT

Hoffman JM, Pagulayan KF, Zawaideh N, Dikmen S, Temkin N, Bell KR:Understanding pain after traumatic brain injury: impact on community participa-tion. Am J Phys Med Rehabil 2007;86:962–969.

Objective: To examine the prevalence of pain 1 yr after moderate tosevere traumatic brain injury (TBI) and identify predictors from the time ofinjury. Additionally, factors related to pain at 1 yr after injury were exam-ined along with the impact of pain on community participation.

Design: Prospective cohort study of 146 individuals enrolled duringacute inpatient rehabilitation for TBI and community follow-up at 1 yr afterinjury.

Results: Higher reports of depressive symptoms during inpatient re-habilitation and at 1 yr after injury were significantly related to reports ofpain at 1 yr when controlling for demographic and injury characteristics.Being female and nonwhite were also factors related to increased reportsof pain. Pain and community participation were significantly related untildepression was entered into the model. Depression is a significant factorin the relationship between pain and community participation.

Conclusion: Whereas pain was frequently reported 1 yr after injury,injury-related factors were surprisingly unrelated. Further evaluation of therole that depression plays in the relationship between pain and communityparticipation will be important to determine appropriate management ofpain and depression and to optimize participation in individuals with TBI.

Key Words: Pain, Traumatic Brain Injury, Rehabilitation

962 Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

RESEARCH ARTICLE

Brain Injury

The experience and explication of pain amongindividuals with disability is developing into a focusof rehabilitation research. Studies have been con-ducted to understand the prevalence and impact ofpain on individuals with a variety of disabling con-ditions such as spinal cord injury, amputation, andmultiple sclerosis.1,2 Pain after traumatic brain in-jury (TBI) has only been partly examined, but hasbeen noted frequently as an area in need of furtherresearch.3–8 Much of the existing research has pri-marily focused on posttraumatic headache, andtypically for those with mild TBI.9–11 In addition,most studies on pain after TBI have enrolled indi-viduals presenting to outpatient brain injury clin-ics, rather than prospectively enrolling subjects,which limit the ability to generalize results.12,13

Pain has also been documented as a significantconcern after more severe injury. For instance,Walker and colleagues14 observed 109 consecutiveVeterans Administration patients with moderate tosevere TBI for 1 yr. Of those patients observed, 38%reported headache during their inpatient rehabili-tation, and approximately 50% of those who ini-tially reported headache continued to endorseheadache at 1 yr after injury.

Whereas headache has been more widely exam-ined, other painful conditions that can develop afterbrain injury have not been as frequently evaluated.Such conditions include spasticity, peripheral nerveinjury, reflex sympathetic dystrophy, contracture,and heterotopic ossification.15 A case series con-ducted on patients presenting to a brain injury clinicfound that 58% of individuals with mild TBI and 52%of those with moderate to severe TBI reportedchronic pain.12 Headaches were the most commonsource of pain for both groups (47 and 34%, respec-tively), followed by neck/shoulder, upper-limb, andlower-limb pain. Bryant and colleagues16 observed 96patients with severe TBI for 6 mo after injury andfound that 31% reported chronic pain associated withhead, neck, back, and limb pain. Other reports of painvary in the literature but typically exceed 50% of allpatients assessed.13

The problems associated with the presence ofpain are particularly worrisome for individualswith TBI. In noninjured individuals, pain has beenfound to interfere with cognition and exacerbateproblems with sleep and fatigue and is stronglyassociated with anxiety and depression, as well asother psychological disorders such as posttraumaticstress disorder.16–18 Mooney and colleagues19 foundthat poor recovery after mild TBI was explained inlarge part by depression and pain. Depression wasalso found to be highly correlated with headache painat 1 yr after injury.14

The aim of the current study was to deter-mine the prevalence of pain at 1 yr after injury

using a secondary analysis of a prospectivelycollected sample and to examine possible factorsrelated to overall pain reports in individuals withTBI. Possible predictors of pain at 1 yr wereexamined, as well as factors associated with painat 1 yr. In addition, the relationship of pain,depression, and community participation was ex-plored.

METHODSStudy Population

One hundred forty-six individuals who com-pleted the SF-36 at their 1 yr after injury evalua-tion were included in the current study from a totalof 202 individuals who were eligible and consecu-tively enrolled in the University of Washington’s(UW) Traumatic Brain Injury Model System(TBIMS). Twenty-three of the 202 subjects werelost to follow-up, and 33 did not complete theSF-36 because of limited testing time. No signifi-cant differences between those who completed theSF-36 and those who did not were found on gen-der, age, and lowest Glasgow coma scale (GCS)score within 24 hrs of injury. However, those thatdid not complete the SF-36 were more likely to benonwhite individuals. Individuals were enrolled be-tween 1998 and 2001 during their inpatient reha-bilitation stay. The TBIMS is a longitudinal data-base and research program sponsored by theNational Institute on Disability and RehabilitationResearch. Criteria for enrollment include (1) pri-mary diagnosis of TBI on admission to the inpa-tient rehabilitation unit, (2) age greater than 16yrs, (3) arrival at the emergency department within24 hrs of injury, and (4) receipt of both acute careand inpatient rehabilitation within the Model Sys-tem of care. Individuals were assessed during theirinpatient rehabilitation stay (baseline) and at 1 yrafter injury. During inpatient rehabilitation, indi-viduals were not consented or evaluated until theyhad two consecutive scores of 75 or above on theGalveston Orientation and Amnesia Test (GOAT),20

which places them out of posttraumatic amnesia.In addition, study examiners discontinued testingif there was any question on validity of responses.The majority of 1 yr after injury evaluations wereconducted in person. Alternatively, when an indi-vidual was unable to come in person due to dis-tance from the center they were conducted overthe telephone. The study was approved by the in-stitutional review board at the University of Wash-ington.

MeasuresDemographics

Age at the time of injury, years of education,gender (female � 1, male � 2), and race (white �1, nonwhite � 0) were included in analyses.

December 2007 Pain and Community Participation After TBI 963

Injury CharacteristicsSeverity of injury was measured using the low-

est GCS within 24 hrs of injury. GCS measureslevel of consciousness as indicated by eye opening,verbal response, and motor response to stimuli.Scores can range from 3 to 15, with scores of 3–8indicating severe injury, scores of 9–12 indicatinga moderate injury, and scores of 13–15 indicating amild injury.21 Scores were prorated if an individualhad a score for the best motor component but wasmissing the eye opening and/or verbal responsecomponent of the GCS.22 In the current sample,verbal response was missing in 78% of cases be-cause subjects were intubated at the time of assess-ment. Eye opening was missing in only one case.Cause of injury was categorized into four groups:(1) motor vehicle related, 2) fall, 3) violence, and 4)all other causes, including pedestrian, bicycle, andsports-related injuries. In addition, the number ofassociated injuries that occurred at the time ofinjury was included to examine the impact of non–TBI-related injuries on reports of pain. The totalnumber of injuries, including fracture, amputa-tion, peripheral nerve injury, brachial plexus in-jury, intraabdominal injury, spinal cord injury,and/or traumatic pneumothorax/hemothorax wascalculated.

Functional ImpairmentThe Functional Impairment Measure (FIM) in-

strument is a widely used measure of functionalstatus after TBI. This measure includes 18 itemsthat assess level of independence or dependence insix domains of functioning (self-care, sphincter con-trol, mobility, locomotion, communication, and so-cial cognition). Ratings range from 1 (completelydependent) to 7 (independent). This measure wascompleted for each patient at time of discharge andat 1 yr after injury.

Depression Symptom SeverityThe depression subscale, one of the nine pri-

mary dimensions from the Brief Symptom Inven-tory, was used to determine the level of distress ordepression.23 This measure was administered dur-ing inpatient rehabilitation (baseline) and at 1 yrafter injury. The depression subscale has high in-ternal reliability (Cronbach � � 0.85).23 Rawscores from the subscale were transformed intoT-scores using nonpatient adult norms.16 Higherscores indicate more depression.

Community ParticipationThe Community Integration Questionnaire

(CIQ),24 a widely used measure for assessing homeand social integration as well as community in-volvement after TBI, was used to assess participa-

tion at 1 yr after TBI. The measure includes ques-tions about home and social integration andproductive activity level. Scores range from 0 to 29,with higher scores representing more communityparticipation. Test–retest reliability coefficientsrange from 0.83 to 0.97.24

PainThe two-item Bodily Pain scale of the SF-36

Health Survey (SF-36)25 was used to assess pain.26

This subscale contains two items that assess painintensity (ranging from none to very severe) andpain interference (ranging from not at all to ex-tremely), which are combined into a compositescore of pain severity and transformed into a 0–100(no pain) scale, with lower scores indicating morepain. The Bodily Pain scale has demonstrated goodtest–retest reliability (0.78) and validity.26 In addi-tion, it has been found to have good internal con-sistency among individuals with TBI.27 The com-posite score was calculated, using the scoringcriteria outlined in the SF-36 manual. Norms arealso given in the manual from a United Statespopulation sample (mean, 75.15, SD, 23.69).25

Statistical AnalysisPearson correlation coefficients were calcu-

lated among independent and dependent variablesto assess relationships among variables. Two linearregressions were conducted to evaluate (1) baselinepredictors of pain level reported on the Bodily PainIndex of the SF-36 at 1 yr and (2) potential asso-ciated variables, measured at the same time as thelevel of pain. Baseline predictors included demo-graphic variables (age at time of injury, education,race, and sex), injury characteristics (GCS, numberof associated injuries, cause of injury, FIM score attime of hospital discharge), and baseline BSI de-pression score. Dummy coded indices of each ofthe cause of injury factors (motor vehicle accident,violence, fall, or other cause, which included pe-destrian, bicycle, or sports injuries) were used inthis regression analysis. Potential associated vari-ables included baseline demographics as well as theFIM scores, BSI depression scores, and CIQ scoresthat were completed at 1 yr after injury.

When the above analyses were completed, ex-ploratory analyses were conducted to further elu-cidate the relationship between pain, depression at1 yr after injury, and community participation. Aseries of analyses were conducted to evaluate theimpact of depression on the relationship betweenpain and community participation and involvedcomputation of three regression equations: (1)pain predicting community participation (�), (2)pain (�=) and depression (�) predicting communityparticipation, and (3) depression predicting pain

964 Hoffman et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

(�). All equations controlled for the demographicvariables.28

RESULTSOne hundred forty-six individuals who com-

pleted the Bodily Pain subscale of the SF-36 at 1 yrafter injury were included in the study. Prevalenceof pain as well as demographic and injury charac-teristics of this sample are shown in Table 1. Themajority were white males with an average age of

36. The average amount of education was 12.4 yrs(range 3–20 yr). Only 27% of individuals reportedhaving no pain (a score of 100 on the bodily painsubscale), and 45% reported no interference withactivities from pain. The average pain rating on theBodily Pain Index was 66.7. Approximately 50% ofthe sample sustained their injury in a motor vehi-cle accident. Descriptive statistics for FIM scoresand BSI depression scores at baseline and 1 yr afterinjury, and community participation measuredwith the CIQ, are also displayed in Table 1. FIMscores and BSI depression scores at 1 yr weresignificantly higher than at baseline, indicatingthat functional status improved and depressionworsened (P � 0.001 and 0.05, respectively).

Pearson correlations are presented in Table 2.Results suggest that higher reports of pain at 1 yrafter injury were significantly related to being fe-male, nonwhite, and having lower function as mea-sured by the FIM instrument at 1 yr after injury.Additionally, higher reports of pain were signifi-cantly related to higher rates of depressive symp-toms both at baseline and 1 yr after injury, andlower community participation as measured by theCIQ. Interestingly, neither severity of TBI nornumber of associated injuries was associated withpain at 1 yr after injury. Higher rates of depressivesymptoms at baseline and 1 yr after injury wererelated to being female, having lower function at 1yr and lower community participation. Baselinedepressive symptoms were also related to lowerfunction at baseline. Reports of higher levels ofcommunity participation at 1 yr were related toyounger age, being white, having more education,having more injuries, and highly related to betterlevel of function measured by the FIM instrument.

Evaluation of cause of injury suggests thatindividuals who were violently injured had signifi-cantly higher levels of pain than those in the otherinjury category (mean 50.94 compared with 78.32,

TABLE 1 Sample characteristics

n Percentage of Sample

GenderMale 111 76.0Female 35 24.0

RaceWhite 114 78.1Nonwhite 32 21.9

Cause of injuryMotor vehicle

related71 48.6

Fall 33 22.6Violence 17 11.6Other 22 15.1Missing 3 2.1

Mean (SD) RangeAge 146 36.1 (14.9) 17–88Education 145 12.4 (2.7) 3–20Glasgow coma scale 143 9.2 (3.1) 3–15Number of associated

injuries146 2.0 (1.7) 0–7

FIM scoresBaseline 144 112.5 (14.4) 32–1261 yr after injury 140 122.7 (5.5) 86–126

BSI depressionBaseline 132 52.9 (9.6) 36–781 yr after injury 136 55.8 (12.2) 36–80

CIQ functional score 142 17.7 (6.0) 2.5–29SF-36 bodily pain 146 66.7 (28.4) 0–100

TABLE 2 Correlations among variables

BodilyPain Index

BaselineDepression

1-yrPostinjury Depression CIQ

Age �0.07 0.10 0.15 �0.21*Gender 0.19* �0.21* �0.36† �0.08Race 0.19* 0.04 0.01 0.17*Education 0.11 �0.13 �0.11 0.28†

Number of injuries 0.07 0.01 �0.07 0.17*Glasgow coma scale �0.07 �0.08 0.00 0.14Discharge FIM score 0.10 �0.17* �0.02 0.36†

1-yr FIM score 0.23* �0.30† �0.33† 0.50†

Baseline depression �0.30† —1-yr postinjury depression �0.54† 0.50† —CIQ community participation 0.30† �0.37† �0.33†

* P � 0.05; † P � 0.001.


respectively). No other significant differences werefound.

Predictors and Correlates of Pain at 1 yrAfter Injury

Table 3 displays the results of the regressionmodel with the outcome of pain at 1 yr after injuryand predictors from baseline, including demo-graphic factors, injury characteristics, function atdischarge, and depressive symptoms. Overall, thesevariables accounted for 25% of the variance in theBodily Pain measure (P � 0.001). Higher levels ofdepression during inpatient rehabilitation, and be-ing female and nonwhite were associated withhigher levels of reported pain at 1 yr after injury.Consistent with the bivariate correlational analy-ses, age and injury characteristics were not signif-icant predictors of pain at 1 yr. Similarly, functionat baseline was not found to be a significant pre-dictor of pain reports at 1 yr.

Table 4 displays the results of the regressionmodel with the outcome of pain at 1 yr after injuryincluding demographic variables, level of function,depressive symptoms, and community participa-

tion measured at 1 yr after injury. Overall, thesevariables accounted for 36% of the variance in theBodily Pain measure (P � 0.001). In this analysis,higher reports of depression at 1 yr after injury andbeing nonwhite were significantly associated withhigher levels of pain reported at 1 yr after injury,whereas other demographics, function, and com-munity participation were not significantly relatedto reports of pain when the other factors are takeninto account.

Relationship Among Pain, Depression,and Community Participation

Given the finding that community participa-tion was related to reported pain on a bivariatelevel, but not when included in the regressionmodel, further analyses were conducted to deter-mine how pain, depression, and community partic-ipation interrelate. The three regression equationswere conducted as described in the Statistical Analy-sis section, and results are shown in Figure 1. Aftercontrolling for demographic variables, pain was asso-ciated with community participation. However, whendepression was added into the model, the relationshipbetween pain and community participation becomesnonsignificant. This set of relationships suggest thatdepression is an important variable in the relation-ship between pain and community participation.

TABLE 3 Summary of regression analyses forvariables predicting pain at 1 yrafter injury

B SE b t

Age �0.09 0.20 �0.45Education 1.14 1.07 1.08Race 15.87 5.79 2.74**Sex 13.22 5.85 2.26*Glasgow coma scale �0.73 0.86 �0.84Associated injuries �1.65 1.65 �1.00Fall �1.80 6.54 �0.27Violence �9.90 8.28 �1.20Other Injury 7.01 7.22 0.97Discharge FIM score 0.32 0.23 1.37Depression at baseline �0.68 0.27 �2.54*

R2 � 0.25; adjusted R2 � 0.18. * P � 0.05; ** P � 0.01.

FIGURE 1 Exploration of the relationship betweenpain, depression, and community partic-ipation. Standardized values are reportedin the figure above. � and � indicate val-ues when only pain or depression predictsCIQ including only demographic vari-ables. �= and �= indicate the value whenboth pain and depression are included inthe model along with demographic vari-ables.

TABLE 4 Association with pain at 1 yr afterinjury

B SE b t

Age �0.01 0.15 �0.05Education �0.52 0.87 �0.59Race 16.70 5.38 3.10**Sex 3.68 5.28 0.701-yr FIM score �0.02 0.45 �0.05Depression at 1 yr �1.10 0.20 �5.46**CIQ 0.66 0.46 1.44

R2 � 0.36; adjusted R2 � 0.32. ** P � 0.01.


DISCUSSION

This study shows that pain is a significantproblem at 1 yr after injury for individuals withTBI, with 74% of participants reporting some levelof pain and 55% reporting interference with dailyactivities from pain. Whereas these findings sug-gest higher rates of reported pain than previouslypublished in the literature,12,14,16 the average painrating of 66.7 is similar to that reported by Findlerand colleagues,27 who found average pain ratingson the Bodily Pain Index of 67 for individuals withmoderate to severe TBI. The higher rates of re-ported pain may be attributable to a variety offactors, including the population included, howpain was assessed, and the development of unmea-sured complications (e.g., contractures, hetero-topic ossification, and spasticity), which may de-velop in the year after injury.

Initial evaluation of relationships among vari-ables suggested that pain at 1 yr was correlated withbeing female, nonwhite, lower functional status, andwith higher reports of depressive symptoms at base-line and 1 yr after injury and with more limitedcommunity participation at 1 yr after injury. Simi-larly, in the regression models with baseline charac-teristics predicting pain at 1 yr, gender (being fe-male), race (being nonwhite), and higher reports ofdepressive symptomatology were related to higherpain ratings at 1 yr. Injury characteristics includingGCS and number of associated injuries were notfound to be related in correlational or regressionanalyses. This finding is consistent with Walker andcolleagues,14 who report that headache pain was un-related to injury severity.

At 1 yr after injury, race (being nonwhite) andincreased depression were the only variables inde-pendently related to higher rates of pain. The sig-nificant impact of depression on individuals afterTBI has been suggested in the literature and hasbeen found to be related to pain after injury. Forexample, Mooney and colleagues19 suggest that pooroutcomes after mild brain injury were related to bothdepression and pain. In addition, Chiu and col-leagues29 found that psychological distress, depres-sion, and poor sleep can lead to lower pain thresholds,which may impact reports of pain. The relationshipbetween being nonwhite and higher reports of painis difficult to interpret in the current sample, giventhe small sample size and the heterogeneous pop-ulation included in the nonwhite category. Futureexamination of the impact of race on pain reportsafter TBI is warranted.

Another interesting finding in this study wasthat community participation was associated withpain on a bivariate level but was not significantlyassociated with pain in the regression analysis.Further exploration of the relationship among

pain, depression, and community participationsuggests that depression influences the relation-ship between pain and community participation at1 yr after injury. In other words, the relationshipbetween pain and community participation be-comes nonsignificant when depression is takeninto account. This is a complex relationship thatwill require further exploration to determine po-tential causal relationships and the direction ofthat causality. Although direction of causality can-not be explored with the data available in thisstudy, these results do offer some suggestions forfuture causal hypotheses. One such hypothesis maybe that depression acts as a mediator in the rela-tionship whereby pain leads to increased depres-sion which leads to decreased participation.30 Al-ternatively, depression may lead to increased painreports and decreased participation which wouldsuggest that depression is the influential variablewhich should be the focus of interest. Whereasreports of depression assessed at baseline and 1 yrwere significantly associated with pain at 1 yr afterinjury, both pain and community participationwere only assessed at 1 yr making any determina-tion of causal relationships impossible in the cur-rent study. Also, it is possible that variables notmeasured in this study are important to under-standing the relationship among depression, pain,and community participation. Other potential vari-ables of interest may include psychological diag-noses, such as posttraumatic stress disorder, whichwas found to potentially mediate the relationshipbetween pain and depression and pain and commu-nity participation in a sample of patients with se-vere TBI.16 Further examination of the possiblerelationship of depression with pain and function isneeded to evaluate the impact of other unmeasuredvariables.

Results of the current study do suggest thatdepression and pain are significant problems forindividuals with TBI. Further evaluation of therelationship may assist clinicians in determiningbest treatment approaches. If depression is the keyvariable leading to increased reports of pain anddecreased participation, then treatment for depres-sion should lead to improvements in both. How-ever, if depression is a mediator in the relationshipbetween pain and participation then treatmentwould need to focus on decreasing pain first, thenon treating depression to increase participation.Clinicians would benefit from conducting moredetailed evaluations of pain, depression, and func-tional participation when considering treatmentapproaches for individuals with TBI such as thosedescribed by Martelli and colleagues.31,32 Consid-eration of alternative treatment interventions willbe needed given the potential sedative effects ofpain medications in individuals with TBI and the


reduced ability to participate in cognitive behav-ioral treatments.33 In general, there has been areluctance to treat pain in cognitively impairedindividuals. However, lack of treatment can nega-tively impact not only individuals with TBI, buttheir families and caregivers as well.34 Therefore,further study is warranted to identify where tofocus treatment and on the best treatment meth-ods for this population.

This study has several limitations that shouldbe considered in interpreting the results. First,assessments were conducted only at baseline and 1yr; thus, we cannot determine whether pain reportswere ongoing since injury or whether they involvedcomplications developing after TBI. Second, painevaluation was limited to questions asked on theSF-36, which do not include the site of pain (headvs. body, etc.) or how much pain was associatedwith other injuries (fractures, amputations, etc.)that occurred at the time of injury. However, boththe severity and interference of pain were included.More comprehensive assessment over time mayprovide information that would determine causalrelationships and allow for the evaluation of therole of depression after TBI. In addition, we had noinformation on whether individuals had been re-ceiving ongoing treatment for their pain or depres-sion or how they were coping with their pain moregenerally. Third, given the types of measures thatwere chosen, we have no ability to account for anyresponse bias that may exist and could influencethe results. Although it is possible that individualsmay have exaggerated their levels of pain and de-pression, they were instructed as to the confiden-tial nature of their responses, including a certifi-cate of confidentiality. Finally, the proportion ofnonwhites was lower in the group that completedthe SF-36 compared with those that did not com-plete it (or were lost to follow up). Individuals whoare nonwhite have been found to report higherpain although mixed results in the literature sug-gest that this finding may be due to other factors.35

In the current study, it is not possible to assesswhether individuals who participated in the studymay differ from those who were lost to follow-up orunable to respond because of other factors. Lesserrepresentation of nonwhites in the sample is likelyto underestimate pain. This study was not able toexamine the effects of cultural influences on painand pain perception in TBI patients, but this is animportant topic for future study.

The current study suggests that pain is a fre-quently reported problem at 1 yr after injury inindividuals with TBI. Thorough evaluation of painby clinicians assisting individuals with TBI wouldbe beneficial. In addition, further research isneeded to understand the relationship among pain,depression, and community participation to deter-

mine causal relationships. Such research may in-form clinicians on appropriate assessment andtreatment recommendations for this population.

REFERENCES1. Dudgeon BJ, Ehde DM, Cardenas DD, Engel JM, Hoffman

AJ, Jensen MP: Describing pain with physical disability:narrative interviews and the McGill Pain Questionnaire.Arch Phys Med Rehabil 2005;86:109–15

2. Ehde DM, Jensen MP, Engel JM, Turner JA, Hoffman AJ,Cardenas DD: Chronic pain secondary to disability: a review.Clin J Pain 2003;19:3–17

3. Nicholson K, Martelli MF: The problem of pain. J HeadTrauma Rehabil 2004;19:2–9

4. Martelli MF, Zasler ND, Bender MC, Nicholson K: Psycho-logical, neuropsychological, and medical considerations inassessment and management of pain. J Head Trauma Re-habil 2004;19:10–28

5. Ivanhoe CB, Hartman ET: Clinical caveats on medical as-sessment and treatment of pain after TBI. J Head TraumaRehabil 2004;19:29–39

6. Branca B, Lake AE: Psychological and neuropsychologicalintegration in multidisciplinary pain management afterTBI. J Head Trauma Rehabil 2004;19:40–57

7. Hecht JS: Occipital nerve blocks in postconcussiveheadaches: a retrospective review and report of ten patients.J Head Trauma Rehabil 2004;19:58–71

8. Walker WC: Pain pathoetiology after TBI: neural and non-neural mechanisms. J Head Trauma Rehabil 2004;19:72–81

9. Mihalik JP, Stump JE, Collins MW, Lovell MR, Field M,Maroon JC: Posttraumatic migraine characteristics in ath-letes following sports-related concussion. J Neurosurg2005;102:850–5

10. Martelli MF, Grayson RL, Zasler ND: Posttraumaticheadache: neuropsychological and psychological effects andtreatment implications. J Head Trauma Rehabil 1999;14:49–69

11. Gunstad J, Suhr JA: Cognitive factors in postconcussionsyndrome symptom report. Arch Clin Neuropsychol 2004;19:391–405

12. Lahz S, Bryant RA: Incidence of chronic pain followingtraumatic brain injury. Arch Phys Med Rehabil 1996;77:889–91

13. Beetar JT, Guilmette TJ, Sparadeo FR: Sleep and pain com-plaints in symptomatic traumatic brain injury and neuro-logic populations. Arch Phys Med Rehabil 1996;77:1298–1302

14. Walker WC, Seel RT, Curtiss G, Warden DL: Headache aftermoderate and severe traumatic brain injury: a longitudinalanalysis. Arch Phys Med Rehabil 2005;86:1793–1800

15. Bell KR, Pepping M, Dikmen S: Rehabilitation after Trau-matic Brain Injury. Trauma Rehabilitation. Philadelphia,Lippincott Williams & Wilkins, 2005

16. Bryant RA, Marosszeky JE, Crooks J, Baguley IJ, Gurka JA:Interaction of posttraumatic stress disorder and chronicpain following traumatic brain injury. J Head Trauma Re-habil 1999;14:588–94

17. Nicholson K: Pain, cognition and traumatic brain injury.NeuroRehabilitation 2000;14:95–103

18. Benjamin S, Morris S, McBeth J, Macfarlane GJ, Silman AJ:The association between chronic widespread pain and men-tal disorder: a population-based study. Arthritis Rheum2000;43:561–67

19. Mooney G, Speed J, Sheppard S: Factors related to recoveryafter mild traumatic brain injury. Brain Inj 2005;19:975–87

20. Levin HS, O’Donnell VM, Grossman RG: The GalvestonOrientation and Amnesia Test. A practical scale to assesscognition after head injury. J Nerv Ment Dis 1979;167:675–84


21. Teasdale G, Jennett B: Assessment of coma and impairedconsciousness. A practical scale. Lancet 1974;2:81–4

22. Gale JL, Dikmen S, Wyler A, Temkin N, McLean A: Headinjury in the Pacific Northwest. Neurosurgery 1983;12:487–91

23. Derogatis LR, Melisaratos N: The brief symptom inventory:an introductory report. Psychol Med 1983;13:595–605

24. Willer B, Ottenbacher KJ, Coad ML: The community inte-gration questionnaire. A comparative examination. Am JPhys Med Rehabil 1994;73:103–11

25. Ware JE, Snow KK, Kosinski M: SF-36 Health Survey:Manual and Interpretation Guide. Lincoln, RI, Quality Met-ric Inc., 1993

26. Von Korff M, Jensen MP, Karoly P: Assessing global painseverity by self-report in clinical and health services re-search. Spine 2000;25:3140–51

27. Findler M, Cantor J, Haddad L, Gordon W, Ashman T: Thereliability and validity of the SF-36 health survey question-naire for use with individuals with traumatic brain injury.Brain Inj 2001;15:715–23

28. MacKinnon DP, Krull JL, Lockwood CM: Equivalence of themediation, confounding, and suppression effect. Prev Sci2000;1:173–81

29. Chiu TT, Lam TH, Hedley AJ: Correlation among physicalimpairments, pain, disability, and patient satisfaction in

patients with chronic neck pain. Arch Phys Med Rehabil2005;86:534–40

30. MacKinnon DP, Lockwood CM, Hoffman JM, West SG,Sheets V: A comparison of methods to test mediation andother intervening variable effeccts. Psychol Methods 2002;7:83–104

31. Martelli MF, Nicholson K, Zasler ND: Psychological ap-proaches to comprehensive pain assessment and manage-ment following TBI, in Zasler ND, Katz DI, Zafonte RD(eds): Brain Injury Medicine: Principles and Practice. NewYork, Demos, 2007, pp 723–40

32. Zasler ND, Horn LJ, Martelli MF, Nicholson K: Post-trau-matic pain disorders: medical assessment and management,in Zasler ND, Katz DI, Zafonte RD (eds): Brain InjuryMedicine: Principles and Practice. New York, Demos, 2007,pp 697–722

33. Rosenthal M, Christensen BK, Ross TP: Depression follow-ing traumatic brain injury. Arch Phys Med Rehabil 1998;79:90–103

34. Cook AK, Niven CA, Downs MG: Assessing the pain of peoplewith cognitive impairment. Int J Geriatr Psychiatry 1999;14:421–5

35. Edwards RR, Moric M, Husfeldt B, Buvanendran A,Ivankovich O: Ethnic similarities and differences in thechronic pain experience: a comparison of african american,Hispanic, and white patients. Pain Med 2005;6:88–98


Authors:Jenny P. Liao, MDAndrew J. Waclawik, MDBarend P. Lotz, MDSharhriar M. Salamat, MD, PhDBrad R. Beinlich, MDBenjamin R. Brooks, MD

Affiliations:From the Department of Neurology,University of Wisconsin MedicalSchool, Madison, Wisconsin.

Correspondence:All correspondence and requests forreprints should be addressed toJenny P. Liao, MD, Clinical AssistantProfessor of Neurology, Departmentof Neurology, CSC H6/574, Universityof Wisconsin Medical School, 600Highland Avenue, Madison, WI53792.

Disclosures:A portion of this manuscript waspresented in abstract form inNeurology 1997;48:A445.


DOI: 10.1097/PHM.0b013e3181588331

Myopathic Dropped Head SyndromeAn Expanding Clinicopathological Spectrum

ABSTRACT

Liao JP, Waclawik AJ, Lotz BP, Salamat SM, Beinlich BR, Brooks BR: Myo-pathic dropped head syndrome: an expanding clinicopathological spectrum. Am JPhys Med Rehabil 2007;86:970–976.

Objective: A number of neuromuscular conditions may lead to adropped head syndrome (DHS), with some patients developing a lateonset noninflammatory myopathy affecting only, or predominantly, neckextensor muscles (NEM). The cause, pathogenesis, and nosological clas-sification of this condition are unclear. To further investigate this condition,the authors evaluated the clinical, electrodiagnostic and pathologic find-ings in seven patients with a myopathic DHS.

Design: Analysis of clinical data, electrodiagnostic studies, and musclebiopsies of seven patients, including one set of identical twins, whodeveloped a very late onset myopathy with severe NEM weakness.

Results: Age of onset was 61–79 yrs, with the pair of identical twinsdeveloping NEM weakness within 1 yr of each other (ages 63 and 64,respectively). Seven patients developed weakness (six slight weaknessand one more severe) in muscles other than NEM. The group wascharacterized by the electromyography (EMG) showing a “myopathic”pattern in cervical paraspinal muscles (7/7), muscle biopsies with non-specific myopathic changes on histologic stains (7/7), marked abnormal-ities in NADH dehydrogenase–reacted sections (6/7), desmin-positivesarcoplasmic deposits (1/7), low carnitine levels by biochemical assays(2/7), and mitochondrial changes (3/7).

Conclusions: Myopathic DHS encompasses a wide spectrum of con-ditions that strongly affect NEM; however, as documented in the monozy-gotic twins, some patients may suffer from a distinct, genetically deter-mined form of late-onset restricted myopathy leading clinically to DHS.

Key Words: Dropped Head, Myopathy, Neck Extensor Muscles, Bent Spine


RESEARCH ARTICLE

Myopathy

A number of neuromuscular conditions, includ-ing neurogenic and myopathic entities, as well asdefective neuromuscular transmission, may presentwith isolated or predominately neck extensorweakness.1–10 Motor neuron disease, polymyositis,and myasthenia gravis are three conditions mostcommonly associated with severe weakness of neckextensors. The term dropped head syndrome(DHS) was introduced by Suarez and Kelly10 in1992 to designate patients with a late-onset non-inflammatory myopathy that affects only, or pre-dominantly, neck extensor muscles (NEM). Severalcase series of DHS have been reported.4,7,10–13 Sub-sequently the term isolated neck extensor myop-athy (INEM) was proposed by Katz et al.4,5 to em-phasize the distinct clinical presentation of thesepatients.

The cause and pathogenesis of INEM/DHS re-main unclear. It has been increasingly recognizedby neuromuscular specialists. The nosological clas-sification of INEM/DHS is controversial, and itlikely represents a heterogeneous condition. Someauthors have included in the DHS spectrum disor-ders such as parkinsonism,14 which initially maypresent with neck extension weakness, adding tothe nosological confusion.

We performed a retrospective review of sevenpatients, including a pair of monozygotic twins,who presented with noninflammatory myopathieswith predominant NEM weakness. Our findingsreveal that in all cases, the weakness and myo-pathic changes seen on muscle biopsies extendedbeyond neck extensor muscle group. It seems thatmyopathic dropped head syndrome is an expandingclinicopathological spectrum.

PATIENTS AND METHODSSeven patients were reviewed, including a set

of monozygotic twins, who were followed in ourneuromuscular clinic spanning the years 1993–2002, and who presented with severe NEM weak-ness. All patients underwent serial neurologicalexaminations, laboratory studies (including creat-ine kinase and acetylcholine receptor antibody test-ing), electrodiagnostic testing (including 2-Hz re-petitive nerve stimulation and needle EMG) andmuscle biopsy.

Open muscle biopsy sites included the trape-zius muscle in two patients, deltoid muscle infour patients, and quadriceps muscle in one pa-tient. The muscle specimens were frozen in liq-uid nitrogen before being cross-sectioned andstained with hematoxylin and eosin, Gomoritrichrome, nicotinamide adenine dinucleotide de-hydrogenase-tetrazolium reductase (NADH-Tr),adenosine triphosphatase (at pH 4.3, 4.6, and 9.4),acid phosphatase, nonspecific esterase, periodic

acid Schiff, oil red O, and myophosphorylase, perstandard protocols.15 Where appropriate, succinatedehydrogenase and cytochrome-c oxidase–reactedsections were also prepared. Quantitative carnitinelevels in muscle were obtained in three patients(patients 1A, 1B, and 2). Immunolabeling fordesmin was obtained in one patient (patient 4).Electron microscopy (EM) was performed in fivepatients (patients 1A, 1B, 2, 3, 4).

RESULTSClinical Presentation

The clinical and laboratory findings of ourseven patients are summarized in Table 1. The ageof onset was between 58 and 79 yrs old. Patientstypically reported insidious onset of NEM weaknessfor several weeks to months. The time from symp-tom onset to diagnosis of DHS was 1–7 yrs. Thedegree of NEM weakness varied from 1/5 to 5�/5on the Medical Research Council (MRC) scale. Allof our patients experienced at least some degree ofweakness in other muscle groups besides the NEM;these frequently included the neck flexors, proxi-mal arm muscles, or proximal leg muscles. Thecreatine kinase (CK) values ranged from normal tomildly elevated (up to twice the upper limit ofnormal value).

Patients 1A and 1B were monozygotic twinswho developed severe NEM weakness (Fig. 1) atages 64 and 63, respectively. They have been fol-lowed in our neurology clinic for 12 yrs with rela-tive stabilization of their NEM weakness. Patient 2had, in addition to NEM weakness, mild involve-ment of shoulder and pelvic girdle muscles. Patient3 initially presented with NEM weakness but,within 2 yrs, developed cogwheel rigidity in herlimbs, a right upper-limb resting tremor, shufflinggait, dysarthria, dysphagia, and limitation in herextraocular eye movements. She was later diag-nosed with multisystem atrophy. Patient 4 initiallycarried the diagnosis of limb girdle muscular dys-trophy; he was noted to have accompanying severeweakness of his hip flexors in the range of 1/5 to 2/5on the MRC scale. Patients 5 (Fig. 2) and 6 devel-oped NEM weakness during several weeks, whichthen stabilized. No patients had any electrodiag-nostic evidence of a neuromuscular transmissiondefect by 2-Hz repetitive nerve stimulation in sev-eral muscle groups, and all patients tested normalfor thyroid function and acetylcholine receptor an-tibody studies. There was also absence of ocularinvolvement (except patient 3, who developed lim-itations of extraocular movements late in thecourse of her disease as part of her diagnosis ofmultisystem atrophy), speech abnormality, or bul-bar signs in our patients on serial neurologicalexaminations.

December 2007 Myopathic Dropped Head Syndrome 971

Magnetic resonance imaging (MRI) of the cer-vical spine was obtained in all the patients. Asidefrom what was felt to be age-related degenerativejoint disease changes and/or mild disk bulges,there was no significant cervical stenosis, cordcompression, bony structural abnormalities, ormass lesions.

Electrodiagnostic DataAll the patients had nerve conduction study

(NCS) and EMG performed on them at least once.

FIGURE 1 Patients 1A and 1B (identical twins). Bothpatients were barely able to hold theirheads up against gravity (neck extensionstrength 3/5 on the MRC scale) and even-tually required neck collars for support.

FIGURE 2 Patient 5. Severe neck extension weak-ness, with dropped head appearance.

TABLE 1 Summary of clinical and laboratory findings

Patient SexAge atOnset

Age atTime of

Diagnosis

NeckExtensionWeaknessa

Other Muscle GroupWeakness

CK (Normal0–175 U/liter)

MuscleBiopsied

1A F 64 65 4 to 5�/5 4� to 5�/5 neckflexion

352 Trapezius

1B F 63 65 4 to 4�/5 5�/5 neck flexion,4� to 5�/5 hipflexion

181 Trapezius

2 F 79 84 2 to 3/5 4/5 shoulder girdle,4�/5 neck flexionand pelvic girdle

118 Deltoid

3 F 61 62 1/5 4�/5 neck flexionand arm abduction

369 Deltoid

4 M 67 74 2/5 4/5 neck flexion andshoulder girdle, 1to 2/5 hip flexion

44 Quadriceps

5 M 58 59 3/5 4 to 4�/5 neckflexion, 4 to 4�/5arm abduction,5�/5 hip flexion

306 (0–215) Deltoid

6 F 63 65 3 to 4�/5 4/5 neck flexion, 4�/5 neck lateralrotation, 4�/5arm abduction,4� to 5�/5 elbowflexion/extension

569 (0–250) Deltoid

a Medical Research Council Scale.CK, creatine kinase.

972 Liao et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

Two-hertz repetitive nerve stimulations on theNCS did not reveal any decremental responses tosuggest a neuromuscular transmission defect. TheEMG showed active denervation (as evidenced byfibrillation changes and positive waves) in five ofseven patients (patients 1A, 1B, 2, 4, and 6) andmyopathic changes (short duration, small ampli-tude, and polyphasic motor units) in the cervicalparaspinal muscles in seven of seven patients. Ofthe proximal upper-limb muscles tested, myo-pathic motor unit potentials were detected in thetriceps and deltoid muscles of five of seven patients(patients 1A, 1B, 4, 5, and 6). Four of seven patients(patients 2, 3, 5, and 6) exhibited myopathicchanges in the trapezius and infraspinatus mus-cles. Patient 4, who had severe weakness of his hipflexors (1 to 2/5 on the MRC scale), was found tohave myopathic motor units in his iliopsoas andanterior tibialis muscles. Of note, not all patientsreceived uniform screening EMG studies; rather,the examiner directed his study according to thepatient’s clinical presentation.

Muscle PathologyThe trapezius and deltoid muscles were the

most commonly biopsied muscles. One patient(patient 4) underwent a quadriceps muscle biopsy.All muscle biopsies revealed chronic myopathicchanges (see Figs. 3 and 4) with large variation infiber sizes, internalized nuclei, split fibers, andreduced oxidative enzyme activity on NADH-Trstaining. There was no evidence for inflammatorychanges. In two patients (patients 1A and 1B), theGomori trichrome–reacted sections revealed someof the muscle fibers with increased subsarcolem-mal accumulations of mitochondria (ragged redfibers), and one of the patients (patient 3) had

abnormally shaped mitochondria that containeddense bodies on EM. The muscle biopsies from themonozygotic twins (patients 1A and 1B) revealedsome increase in the number of small lipid drop-lets. Their muscle tissue tested low for total andfree carnitine levels (about half of normal for pa-tient 1A and a fifth of normal for patient 1B). Themuscle biopsy from patient 4 contained accumula-tions of sarcoplasmic material that tested positivefor desmin (see Fig. 5). Five of the seven musclebiopsies (patients 1A, 2, 4, 5, and 6) revealed mildneurogenic atrophy in addition to the myopathicchanges.

DISCUSSIONWe agree with previously reported postulates

that at least a subgroup of patients with a myo-pathic DHS most likely represent a unique form ofrestricted late-onset myopathy affecting predomi-nantly NEM.4,5 However, our experience with pa-tients diagnosed with myopathic DHS indicatesthat in some patients, in addition to NEM weak-ness, there may be coexisting mild involvement ofother muscle groups, especially in the shouldergirdle region. This may be easily overlooked clini-cally when the NEM weakness presents in such adramatic fashion and the mild degree of weaknessin other muscle groups frequently becomes diffi-cult to appreciate. Also, muscle biopsies of theNEM are fraught with technical difficulties as thecervical paraspinal muscle group lies deep underthe thick layer of the trapezius. Therefore, alterna-tive muscle biopsy sites usually include the trape-zius, deltoid, or biceps. Electrodiagnostic studieswith EMG and muscle biopsies of these musclesadjacent to or close to the cervical paraspinalsfrequently reveal that the myopathic process (notrelated to inflammatory, metabolic, or advanceddystrophic process) is not purely limited to the

FIGURE 3 Muscle biopsy from patient 1B. Note in-creased variation in fiber sizes, hypertro-phic and splitting fiber (asterisk), degen-erating fibers (arrow), and internalizednuclei (arrowheads). Modified Gomoritrichrome stain �300.

FIGURE 4 Patient 1A. Marked abnormality in thedistribution of oxidative enzyme activity.Many fibers seem “mottled” (arrows).NADH-Tr stain �380.


NEM and represents a more diffuse muscle condi-tion. This was not only observed in our case seriesbut also in some patients from previous reports onDHS.4,10

During the past decade, scattered reports havesurfaced linking several disorders (aside from theclassic progressive neuromuscular conditions suchas polymyositis, motor neuron disease, or myasthe-nia gravis) with DHS. These include parkinsonism,chronic inflammatory demyelinating polyneurop-athy, inclusion body myositis, and other nonin-flammatory myopathies such as nemaline rod my-opathy, desmin storage myopathy, mitochondrialmyopathy, facioscapulohumeral dystrophy, adultacid maltase deficiency, and carnitine deficiencymyopathy.1–5,14,16 Metabolic conditions, such ashypothyroidism and hyperparathyroidism, havealso been described in association with DHS.3–5,7,8

Most recently, Rowin et al.17 have reported threepatients who received remote (12–36 yrs ago) ra-diation treatments for Hodgkin’s lymphoma whosubsequently presented with DHS. It is imperativethat clinicians be vigilant in their evaluations forthese conditions as the full clinical picture may bemasked for some time. Within our case series, onepatient was later diagnosed with multisystem atro-

phy, two patients (a set of twins) had low musclecarnitine levels, and one patient’s muscle biopsyshowed sarcoplasmic accumulations of materialthat stained positive for desmin. The desmin accu-mulations may be nonspecific but consistentwith a myofibrillar myopathy.18,19 The mitochon-drial changes seen in our three patients may onlyreflect the aging process, rather than a primarymitochondrial myopathy.7,14,20,21

Because all of the patients reported so far havebeen elderly, one cannot exclude mechanical strainas possibly playing a role in the DHS.4,5,14,21 Manyof the patients have documented cervical degener-ative joint disease, silent radiculopathies, and ky-phosis that may compromise stability, tissue elas-ticity, and musculoskeletal support of the NEM.These underlying age-related changes may furthercompound and exacerbate the NEM weakness ingenetically susceptible individuals. Although theexact association between these conditions andmyopathic DHS has not been well delineated, oneshould consider that DHS is a more heterogeneousdisorder than previously realized.

Work-up for patients with DHS should be tai-lored according to their clinical presentation, butroutine laboratory studies should include CK,erythrocyte-sedimentation rate, C-reactive protein,acetylcholine receptor antibodies, and thyroidstudies. Recently introduced to clinical practice,muscle-specific typrosine kinase antibodies (MuSKAb) may also be useful in detecting atypical sero-negative myasthenia gravis patients presentingwith DHS, because electrodiagnostic studies maynot be sensitive enough to rule out a neuromus-cular transmission defect. In a publication bySanders et al.,22 MuSK Ab were positive in 12 of 32patients with generalized seronegative myastheniagravis. Many of these patients exhibited atypicalfeatures; for instance, repetitive nerve stimulationwas normal in four patients, jitter was normal insix patients, and EMG revealed myopathic motorunits in five patients. Three of the 12 patientspresented initially with NEM weakness, and if itwere not for the presence of MuSK Ab, they verywell could have been classified as having DHSconsistent with INEM. Although all our patientstested negative for acetylcholine receptor antibody(AchR Ab), MuSK Ab was not yet commerciallyavailable to test for any of the patients in ourgroup; thus, atypical myasthenia gravis was notexcluded in all of our patients.

Electrodiagnostic testing may detect myo-pathic change, usually limited to the NEM. Cervicalspine MRI should be obtained to evaluate for com-pressive or spondylitic lesions that may contributeto the DHS. The MRIs of the cervical spine in ourpatients were not done specifically to evaluate theparaspinal musculature; rather, they were ordered

FIGURE 5 Patient 4. Foci of abnormal sarcoplasmicmaterial depositions (arrows). A, ModifiedGomori trichrome; B, desmin stain. Both�375.


to ensure that there was no significant cervicalstenosis, cord compression, bony structural abnor-malities, or mass lesions. There have been reportsin the literature of edema,4 atrophy,23 or fatty re-placement24 of the cervical paraspinal musclesnoted on MRI studies. In selected cases, patientsmay undergo muscle biopsies of clinically or EMG-affected muscles to confirm presence of myopathicchanges and to help exclude inflammatory or met-abolic conditions. Some authors have performedcervical paraspinal muscle biopsies.4,10,14,23,25–27

However, as already stated, these muscle biopsiesmay be technically difficult. Therefore, alternativebiopsy sites (such as the trapezius, deltoid, or bi-ceps) may be considered, on the basis of clinicaland electrodiagnostic findings.

There are no established effective therapies forpatients with myopathic INEM. There are anec-dotal reports of treatments with steroids, azathio-prine, carnitine, coenzyme Q10, and vitamin ther-apies,5–7,25,28,29 but the overall consensus is thatthese treatments are of dubious utility. The fewcases that have reported improvement were pa-tients with inflammatory changes in their musclebiopsies,5,25,29 and they were treated with oral ste-roids, sometimes followed by azathioprine. One canargue that because DHS is felt to be a noninflam-matory myopathic process, the inflammation notedon the muscle biopsies of these patients may actu-ally represent focal polymyositis mimicking asDHS/INEM. Patient 5 was placed on a short trial oforal steroids (unknown dose and duration) by hislocal physician, without apparent improvement.Six of our seven patients were placed on at least a trialof carnitine supplementation at different doses (rang-ing from 990 mg to 3 g total per day) for at least 6mos, and although several of them did report im-proved muscle strength on carnitine supplementa-tion, we could not detect any objective improvement.The monozygotic twins in particular were given car-nitine because their muscle carnitine levels were low,but we suspect that the carnitine deficiency may bean epiphenomenon of muscle degeneration ratherthan representing a primary carnitine deficiency my-opathy, because there is evidence that carnitine defi-ciency does not become a clinically relevant factor inlimiting fatty acid oxidation in muscle until carnitinelevels fall below 5% of normal values.30,31 Eventhough the twins have been maintained on carnitinefor 12 yrs with stable courses, there have been nocontrolled studies or clear evidence to date recom-mending effective treatment modalities for thisgroup of patients. Cervical collars or neck braces,along with physical therapy, may be the most prac-tical supportive treatment options we have to offerthese patients.

We suspect that a subgroup of patients withINEM/DHS may have a genetically determined my-

opathy, particularly given that our monozygotictwins developed this syndrome within 1 yr of eachother. They both shared a very similar clinicalcourse and muscle pathology findings. They arestill actively observed in our institution and, de-spite progressive NEM weakness requiring the useof soft neck collars, the twins have not developedsignificant involvement of other muscle groups,aside from what has been indicated in Table 1. Thispossibility of a genetic factor in the INEM patho-genesis in some cases may be further supported bysome familial cases reported in the bent spinesyndrome,5,24,26 a condition with weakness of thetruncal extensor muscles that is felt to be withinthe same spectrum as DHS.27 Identification ofmore familial cases and further evaluations withlinkage analysis may help elucidate the clinical andgenetic spectrum of this unique condition.

REFERENCES1. Finsterer J: Dropped head syndrome in mitochondriopathy.

Eur Spine J 2004;13:652–6

2. Hoffman D, Gutman L: The dropped head syndrome withchronic inflammatory demyelinating polyneuropathy. MuscleNerve 1994;17:808–10

3. Hund E, Heckl R, Goebel HH, Meinck H: Inclusion bodymyositis presenting with isolated erector spinae paresis.Neurology 1995;45:993–4

4. Katz JS, Wolfe GI, Burns DK, Bryan WW, Fleckenstein JL,Barohn RJ: Isolated neck extensor myopathy: a commoncause of dropped head syndrome. Neurology 1996;46:917–21

5. Katz JS, Barohn RJ: Paraspinous myopathies: dropped headand bent spine syndromes, in Katirji B, Kaminski HJ, Pres-ton DC, Ruff RL, Shapiro BE (eds): Neuromuscular Disor-ders in Clinical Practice. Boston, Butterworth-Heinemann,2002, pp 1181–6

6. Narayanaswami P, Bertorini TE: The dropped head syn-drome. J Clin Neuromusc Dis 2000;2:106–12

7. Rahim F, Gupta D, Bertorini TE, LeDoux MS: Dropped headpresentation of mitochondrial myopathy. J Clin NeuromuscDis 2003;5:108–14

8. Gourie-Devi M, Nalini A, Sandhya S: Early or late appear-ance of “dropped head syndrome” in amyotrophic lateralsclerosis. J Neurol Neurosurg Psychiatry 2003;74:683–6

9. Lomen-Hoerth C, Simmons ML, DeArmond SJ, Layzer RB:Adult-onset nemaline myopathy: another cause of droppedhead. Muscle Nerve 1999;22:1146–50

10. Suarez GA, Kelly JJ: The dropped head syndrome. Neurol-ogy 1992;42:1625–7

11. Goh KJ, Wong KT, Tan CT: Myopathic dropped headsyndrome: a syndrome of mixed aetiology. J Clin Neurosci2000;7:330–6

12. Chaouat D, Belange G: Dropped head syndrome. Rev RhumEngl Ed 1999;66:430–3

13. Lange DJ, Fetel MR, Lovelace RE, Rowland LP: The floppyhead syndrome [abstract]. Ann Neurol 1986;20:33

14. Askmark H, Eeg-Olofsson KE, Johansson A, Nilsson P,Olsson Y, Aquilonius S: Parkinsonism and neck extensormyopathy. Arch Neurol 2001;58:232–7

15. Dubowitz V: Muscle Biopsy: A Practical Approach, ed 2.London, Bailliere Tindall, 1985

16. Padberg GW, Lunt PW, Koch M, Fardeau M: Diagnosticcriteria for fascioscapulohumeral muscular dystrophy. Neu-romusc Disord 1991;1:231–4


17. Rowin J, Cheng G, Lewis SL, Meriggioli MN: Late appear-ance of dropped head syndrome after radiotherapy forHodgkin’s disease. Muscle Nerve 2006;34:666–9

18. Nakano S, Engel AG, Waclawik AJ, Emslie-Smith AM, BusisNA: Myofibrillar myopathy with abncormal foci of desminpositivity. I. Light and electron microscopy analysis of 10cases. J Neuropathol Exp Neurol 1996;55:549–62

19. De Bleecker JL, Engel AG, Ertl BB: Myofibrillar myopathywith abnormal foci of desmin positivity. II. Immunocyto-chemical analysis reveals accumulation of multiple otherproteins. J Neuropathol Exp Neurol 1996;55:563–77

20. Rifai Z, Welle S, Kamp C, Thornton CA: Ragged red fibers innormal aging and inflammatory myopathy. Ann Neurol1995;37:24–9

21. Mendell JR: Mitochondrial myopathy in the elderly: exag-gerated aging in the pathogenesis of disease. Ann Neurol1995;37:3–4

22. Sanders DB, El-Salem K, Massey JM, McConville J, VincentA: Clinical aspects of MuSK antibody positive seronegativeMG. Neurology 2003;60:1978–80

23. Okamiya S, Ito H, Saito T, Kowa H, Torii J: A case of

isolated neck extensor myopathy with parkinsonism. Rin-sho Shinkeigaku 1997;37:393–6

24. Yoshimura DM, Greenberg SA, Lazer RB: Paraspinal mus-cular dystrophy: a new entity [abstract]. Neurology 1996;46(suppl):A309

25. Baquis GD, Moral L, Sorrell M: Neck extensor myopathy: amitochondrial disease [abstract]. Neurology 1997;48:A443

26. Serratice G, Pouget J, Pellisier JF: Bent spine syndrome.J Neurol Neurosurg Psychiatry 1996;60:51–4

27. Swash M: Dropped-head and bent-spine syndromes; axialmyopathies? Lancet 1998;352:758

28. Bahnof R: The dropped head syndrome: rehabilitation ofcervical focal myositis. Disabil Rehabil 1999;21:563–5

29. Biran I, Cohen O, Diment J, Peyser A, Bahnof R, Steiner I:Focal, steroid responsive myositis causing dropped headsyndrome. Muscle Nerve 1999;22:769–71

30. Stanley CA, Deleeuw S, Coates PM, et al: Chronic cardio-myopathy and weakness or acute coma in children with adefect in carnitine uptake. Ann Neurol 1991;30:709

31. Griggs RC, Mendell JR, Miller RG: Evaluation and Treat-ment of Myopathies. Contemporary Neurology Series. Phil-adelphia, F.A. Davis Company, 1995, pp 270–1


Authors:Elaine Biddiss, PhDTom Chau, PhD

Affiliations:From the Bloorview ResearchInstitute, Toronto, Canada (EB, TC);and Institute of Biomaterials andBiomedical Engineering, University ofToronto, Toronto, Canada (EB, TC).

Correspondence:All correspondence and requests forreprints should be addressed to TomChau, Bloorview Research Institute,150 Kilgour Road, Toronto, ON, M4G1R8. Canada.

Disclosures:Supported by the Natural Science andEngineering Research Council,Canada.


DOI: 10.1097/PHM.0b013e3181587f6c

Upper-Limb ProstheticsCritical Factors in Device Abandonment

ABSTRACT

Biddiss E, Chau T: Upper-limb prosthetics: critical factors in device abandon-ment. Am J Phys Med Rehabil 2007;86:977–987.

Objective: To investigate the roles of predisposing characteristics,established need, and enabling resources in upper-limb prosthesis useand abandonment.

Design: A self-administered, anonymous survey was designed to ex-plore these factors. The questionnaire was available online and in paperformat and was distributed through healthcare providers, communitysupport groups, and one prosthesis manufacturer. Two hundred forty-twoparticipants of all ages and levels of upper-limb absence completed thesurvey.

Results: Of participants, 20% had abandoned prosthesis use. Predis-posing factors, namely, origin of limb absence, gender, bilateral limbabsence, and, most importantly, level of limb absence, proved influentialin the decision not to wear prostheses. Enabling resources such as theavailability of health care, cost, and quality of training did not weigh heavilyon prosthesis rejection, with the exception of the fitting time frame and theinvolvement of clients in the prosthesis selection. Conversely, the state ofavailable technology was a highly censured factor in abandonment, spe-cifically in the areas of comfort and function. Perceived need emerged asa predominant factor in prosthesis use.

Conclusions: Future research should focus on continued developmentof more comfortable and functional prostheses, particularly for individualswith high-level or bilateral limb absence. Improved follow-up, repair, andinformation services, together with active involvement of clients in theselection of prostheses meeting their specific goals and needs, is recom-mended.

Key Words: Upper Extremity, Limb Prosthesis, Prosthesis Fitting, Prosthesis Design,Rehabilitation

December 2007 Factors in Prosthesis Abandonment 977

RESEARCH ARTICLE

Amputee

From cosmetic hands, to body-powered hooks,to externally powered devices that flatter their nat-ural counterpart in attempts of mimicry, prosthe-sis design endeavors to address the varying needsand desires of individuals with upper-limb absence.Nevertheless, an estimated 20% of individuals withupper-limb absence reportedly do not use pros-thetic devices.1 Nonuse may be a lifestyle choicebased on personal needs, values, and perspec-tives.2–6 It may result from medical conditions thatpreempt prosthesis fitting.3,7–9 Or, nonuse may bea consequence of external factors such as limitedinsurance coverage or availability, and inadequatetechnology.10,10–12 Defining reasons for prosthesisnonuse is pertinent to health professionals, re-searchers, and administrators involved in the careof individuals with upper-limb absence, the designof improved prosthetics, and the evaluation of clin-ical strategies.

In a recent review,13 literature on the motivat-ing factors in prosthesis use and abandonment wasexplored using the Anderson behavioral model forhealthcare use.14 Under this framework, prosthesisacceptance is modeled as a function of the predis-posing characteristics (e.g., gender, level, or originof limb absence), the established need, and theenabling resources (e.g., healthcare services). Thereview identifies the following topics as warrantingfurther exploration13:

● The potential roles of specific factors (e.g., originof limb absence, incidence of pain, fitting timeframe) in determining prosthesis use/nonuse.

● The consequences of prosthesis use/nonuse onoverall quality of life and health.

● The interaction and relative importance of pre-disposing characteristics, established need, andenabling resources in prosthesis rejection.

Also of importance in forming a realistic pic-ture of prosthesis use and nonuse is the inclusionof nonwearers who, with one notable exception,6

are often underrepresented because of the ten-dency to sample only those actively involved withrehabilitation centers.6,15–17

The specific objective of this study was, there-fore, to elucidate the motivations for prosthesis useand/or abandonment by a broad examination of thepredisposing characteristics, established need, andenabling resources in a population sampled bothfrom hospital- and community-based support net-works.

METHODSDesign

The six-part survey, described in Table 1, wasdeveloped in consultation with several individuals

with upper-limb absence and a number of promi-nent researchers and clinicians in the field (seeAcknowledgments).

The first four sections were completed by allparticipants. Section Vi, exploring reasons for pros-thesis abandonment, was completed by nonusersonly, and section Vii, regarding prosthesis use andsatisfaction, was completed by current or past us-ers who wished to comment on previous prostheticexperiences. The anonymous survey required20–30 mins to complete and was available online(www.prismlab.org/survey.htm) or in hard copy bycontacting the authors.

Sample Population and SamplingStrategies

All individuals with upper-limb absence wereeligible to participate in this study regardless of (1)level of limb absence, (2) origin of limb absence, (3)bilateral/unilateral limb absence, (4) user status, or(5) types of prostheses used. Parents/guardianswere asked to complete the survey for childrenunder the age of 12. The survey was available inEnglish, Spanish, French, and Dutch and was cir-culated through a number of online supportgroups (i.e., Arm-Amp, I-CAN, Stumps R Us,UpperEx) and healthcare providers (i.e., BloorviewKids Rehab, Canada; Shriners Hospital for Childrenin Los Angeles, CA; Sint Maartenskliniek, the Neth-erlands). In addition, the survey was promoted onthe Otto Bock, Inc., Web site, a prominent manu-facturer of upper-limb prostheses. The samplegroup was self-selected.

Distribution and Data CollectionTen months of data were collected via the

online survey. Various measures, as recommendedfor Web surveys,18 were implemented to monitorsampling and online data collection:

a. To mitigate nonresponse attributable to spamfiltering, incorrect mailing addresses, or mis-trust, local collaborators were enlisted to recruitparticipants. Respondents were asked to specifythe organization through which they were re-cruited.

b. Where possible, prenotification (e.g., mailed fly-ers) and reminders (e.g., through e-mail or tele-phone call) were undertaken to increase re-sponse rates.

c. A well-established, experienced Web provider(www.vovici.com) was employed to host the sur-vey and to limit any system incompatibilities,such as those introduced by browser settings.

d. The survey was made available in paper format,to reduce sampling bias.

978 Biddiss and Chau Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

TABLE 1 Breakdown of questionnaire content

Section TopicNo. and Type of

Questions Relevant Factors

I Participant demographics 7, categorical Level of limb loss1, Likert Length of residua2, numerical Origin of limb loss

Dominant-hand involvementGenderFrequency of associated medical conditions

and painYear of birthYear of amputation

II Personal experiences andlifestyle

1, categorical Occupation status

6, Likert Activity level, work/schoolActivity level, leisureFrequency of exposure to sand,

dirt, grease, etc.Quality of lifeImpact of limb loss on quality of lifeAcceptance of limb loss

III Healthcare experiences 2, Likert Satisfaction with healthcare servicesprovided

1, multiple choice Satisfaction with healthcareinformation provided

1, categorical Frequency of health visits related tolimb loss/prostheses

Country of healthcare provisionIV Personal views on

prostheses1, ordinal Prosthesis design priorities

1, Likert Perceived need for a prosthesis1, multiple choice User status

Vi (Nonusers) Past prosthesis use andreasons for nonwear

1, categorical Past prosthesis use

1, Likert Views on future prosthesis use3, multiple choice Reasons for nonwear3, numerical Challenges encountered in daily life2, open ended

Vii (users) Prosthesis use andsatisfaction

3, categorical Time of fitting

11, Likert Past prostheses3, ordinal Current prostheses14, multiple choice Views on future prosthesis use4, numerical Primary prosthesis4, open ended Model/manufacturer of primary prosthesis

Technical characteristics of primary prosthesis(e.g., proportional control, control strategy)

Cost of prosthesis and repairs/maintenanceFrequency of maintenancePersonal involvement in prosthesis selectionSelection factorsFrequency of wearActivities for which wornFunctional useSatisfaction (i.e., appearance, function, control,

maintenance, comfort, and cost)Detailed design prioritiesUse of secondary prosthesesReasons for nonwearChallenges encountered in daily life

VI Qualitative experiences 1, multiple choice Suggestions for future prosthesis development3, open ended Suggestions for improvements in healthcare

provisionAdditional comments/feedback


e. The Internet protocol (IP) address was used toidentify and filter repeat respondents.

f. To identify possible false respondents, data werescreened for completeness and internal consis-tency using information patterns (e.g., series ofdates, detailed descriptions of limb absence, andprosthesis components).

g. Participant demographics were meticulouslycollected and reported in an effort to contextu-alize the sample population and identify possi-ble biases.

Data AnalysisSPSS 15.0 software was used for statistical

analyses. Nonuser and user groups were comparedvia the �2 test (with Yates correction where appli-cable) or the Student’s t test for nominal andinterval data, respectively. Differences in ordinalratings (i.e., satisfaction, design priorities, qualityof life, etc.) were evaluated using the Mann–Whit-ney U test (for two groups) or the Kruskal–Wallistest (for more than two groups). Mean rankingswere determined and compared using the Fried-man nonparametric ANOVA for repeated measures.Other statistical tools included frequency distribu-tions and three measures of central tendency (i.e.,mean, median, mode) as appropriate to the distri-bution of data in question.

EthicsEthical approval for this study was obtained

from Bloorview Kids Rehab and the University ofToronto.

RESULTSStudy Population Characteristics

A total of 266 respondents completed the anon-ymous survey of which 242 were included in thisstudy (i.e., responses were excluded on the followingbasis: repeated IP address and similarity of data [n �3]; incomplete or inconsistent of data [n � 7]; ineli-gible limb absence [n � 5]; under the age of 12 [n �9]). Survey participants were recruited through reha-bilitation centers (52%), online support groups(39%), the Otto Bock Web site (3%), or through anindependent Internet search or family/friends (6%).Of submissions, 80% were electronic and 20% wereby hard copy. The completion rate of the onlinesurvey (i.e., the number of surveys submitted dividedby the number of times the survey was accessed) was40%. Previous response rates for surveys addressingupper-limb prosthesis design have ranged from24%19 for large-scale mailing campaigns to 69%15 forsampling strategies using more localized and directedcontact.

Levels of limb absence varied, with 16% of thepopulation reporting limb absence at a level distal

to the wrist, 54% at the transradial level, 21% atthe transhumeral level, and 7% at shoulder level orhigher. Of participants, 15% had bilateral limbabsence. The occurrence of congenital limb ab-sence was higher in the pediatric population (91%)than in the adult population (41%) (P � 0.001). Ofindividuals with acquired limb absence, 54% hadlost their dominant hand. Fifty-one percent of par-ticipants were male, and 60% were adults (i.e., 19yrs or older), with an average age of 43 � 15 yrs(19–80 yrs). The average age of the pediatric group(n � 97) was 9.5 � 6 yrs (1–18 yrs) and, conse-quently, 70% of pediatric responses were submit-ted by parents/guardians. Respondents were pri-marily from Canada (35%), the United States(43%), and Europe (17%).

Of participants, 14% had never worn a pros-thesis. Two factors characterized individuals whohad never worn a prosthesis: (a) limb absence distalto wrist was more frequent than in the generalpopulation (P � 0.001), and (b) the mean age ofchildren was younger than in the general popula-tion (P � 0.01), whereas that of adults was older(P � 0.002). According to detailed comments onpatterns of use, 28% of participants who had beenfitted with a device were categorized as prosthesisrejecters (i.e., used a prosthesis once a year or less),whereas 64% were considered frequent wearers(i.e., either full-time or part-time consistent wear).Of note, 45% of participants recruited from thecommunity rejected prosthetic devices as com-pared with 21% of those contacted through reha-bilitation centers or hospitals (P � 0.002). Full-time users wore the prosthesis for an average of13 � 4 hrs on a typical workday and 10 � 5 hrs ona day off. Part-time users reported prosthesis wearfor 7 � 4 hrs per workday and 5 � 4 hrs per day off.

The following will focus on comparisons be-tween prosthesis rejecters and frequent wearerswith regards to predisposing factors, establishedneed, and enabling resources.

Predisposing Factors

Level of limb lossLevel of limb absence emerged as a primary

predisposing factor in prosthesis acceptance. Spe-cifically, individuals with limb absence proximal tothe elbow (high level) or to the wrist (low level)were more likely to reject the device than werethose with transradial limb absence (P � 0.001).

Origin of limb absenceThe relationship between level of limb absence

and prosthesis rejection persisted for both congen-ital and acquired limb absence. Rejection rates forindividuals with congenital limb absence weregreater for low-level (63%) or high-level (65%)


limb absence and less for transradial (21%) (P �0.02). For acquired limb absence, prostheses wererejected by 0% of individuals with low-level ampu-tations, 16% with transradial, and 39% with high-level amputations (P � 0.001). Rejection rates werecomparable for individuals with transradial limbabsence, regardless of origin. Individuals with ac-quired amputations at a low or high level seemedless likely to reject prostheses than those withcongenital limb absence; however, more data arerequired for verification.

GenderOverall, females (39%) were more likely to

reject a prosthesis than were males (23%), (P �0.022). For males and females with congenital limbabsence at any level, no significant differences inprosthesis use were observed. However, femaleswith acquired limb absence were more likely toreject prostheses than were males at both the tran-sradial (P � 0.042) and at high levels (P � 0.001);not enough data were available to assess low-levellimb absence. Females with high-level, acquiredlimb absence in particular, rejected prostheses in80% of cases as compared with 15% of males (P �0.001). A high rate of prosthesis acceptance wasobserved in males with acquired limb absence(91%) as compared with males with congenitallimb absence (60%) (P � 0.001). The oppositetrend was evident in females, for whom 38% withacquired limb absence accepted prostheses, ascompared with 70% with congenital limb absence(P � 0.011).

Bilateral limb absenceIn general, rejection rates for bilateral and

unilateral limb absence did not differ significantly,irrespective of level of limb absence. Rejectionrates for unilateral and bilateral limb absence werecomparable in individuals with acquired amputa-tions. However, individuals with congenital, bilat-eral limb absence had significantly higher rates ofrejection (75%) in comparison with those withunilateral limb absence (28%) (P � 0.004).

Length of residuaThe length of residua was not a factor in pros-

thesis rejection at either the transradial or thetranshumeral level.

Dominant-hand amputationAmputation of the dominant hand was not

associated with prosthesis acceptance or rejection.

Medical factorsDiscomfort, ranging from predominantly mild

to severe, was reported by participants as follows:residual limb pain (32%), phantom pain (32%),

skin irritation (46%), blisters (23%), and upper-body pain (44%). Residual limb pain, phantompain, and upper-body pain were more prevalent inindividuals with acquired limb absence (P � 0.001)and in individuals with high-level limb absence(P � 0.025). Skin irritation and blisters were moreprevalent in frequent wearers than in prosthesisrejecters (P � 0.001). No significant differences inresidual limb, phantom, or upper-body pain wereobserved between prosthesis rejecters and frequentwearers, whereas the latter were more likely toexperience skin irritation and/or blisters (P �0.001) as consequences of prosthesis use. In gen-eral, reports of discomfort and pain were low andthe majority did not experience upper-body pain,skin irritation, blisters, residual limb, or phantompain at any point during a typical week.

AgeNo differences in prosthesis rejection were ob-

served between adults and children with congenitallimb absence, irrespective of gender or level of limbabsence. The mean age of prosthesis rejecters andfrequent wearers was also not significantly differ-ent. However, when grouped by different lifestages, as presented in Figure 1, significant differ-ences in rates of rejection were observed (P �0.02), irrespective of origin of limb absence. Rejec-tion rates peaked markedly in three age groups,from 4 to 10 yrs, from 24 to 35 yrs, and for thosegreater than 65 yrs. These differences are likelyrelated to lifestyle and functional needs, as will bediscussed in the subsequent section.

Established NeedParticipants were asked to rate their perceived

need for a prosthesis on a scale from 1 (not at allneeded for daily life) to 7 (absolutely essential fordaily life). Perceived need varied significantly forfrequent wearers and prosthesis rejecters (P �0.001), with median ratings of 6 and 1 for wearersand rejecters, respectively. No difference in per-ceived need was observed between users of differentprosthesis types.

Participants were asked to comment on anyactivities found to be challenging in everyday life.Of the 178 individuals who responded to this ques-tion, 16% had not encountered any insurmount-able challenges in their daily life, whereas 10% ofindividuals found most activities, particularly bilat-eral tasks, to be challenging. Almost all individualswho reported a great number of challenges madeuse of a prosthesis to aid in their daily life, whereasfrequent wearers and prosthesis rejecters wereequally represented in the group not experiencingchallenges.

Self-reported activity levels (i.e., use of upperlimbs in recreational and job/school activities) and


typical operating environments (i.e., exposure towater, grease, dirt, etc.) were comparable for fre-quent wearers and prosthesis rejecters. Rates ofunemployment and disability leave were low (num-bering 3% each) and not statistically different forwearers and rejecters. Figure 2 presents rejectionrates for various occupations statuses. Of note,students were more likely to reject prostheses thanwere full-time workers (P � 0.021).

Enabling FactorsHealthcare provision and the state of technol-

ogy are potential enabling factors in prosthesisacceptance and will be explored in this section.

Health careSatisfaction with health care in all areas, spe-

cifically fitting (P � 0.001), follow-up (P � 0.001),

FIGURE 1 Rates of prosthesis rejection observed between adults and children with congenital limb absence,grouped by different life stages.

FIGURE 2 Prosthesis-rejection rates for various occupation statuses.


repair (P � 0.001), training (P � 0.007), and in-formation provision (P � 0.009), were significantlylower for prosthesis rejecters. Prosthesis rejecterswere less satisfied with the information providedwith respect to prosthesis technology (P � 0.001),sources of funding (P � 0.01), use of multipleprostheses (P � 0.001), level of expectations set(P � 0.001), and the overall knowledge and expe-rience of healthcare providers (P � 0.001). Bothprosthesis rejecters and frequent wearers were in-terested in receiving better information on non-prosthetic options (i.e., strategies for accomplish-ing activities without use of a prosthesis) andresources for peer support. With these two excep-tions, frequent wearers were generally satisfiedwith all other aspects of health care. Prosthesisrejecters reported neutral feelings (i.e., neither sat-isfied nor dissatisfied) for all other aspects of healthcare with the exceptions of training and fitting,which elicited median ratings of satisfied. Prosthe-sis wearers were in much more frequent contactwith healthcare providers with regard to upper-limb absence than were rejecters (P � 0.001), withappointments occurring at a median frequency ofevery 7–12 mos, whereas 51% of prosthesis reject-ers had not been in contact with healthcare pro-viders for 6 yrs or more.

Of note, satisfaction with overall health care(P � 0.017) and the knowledge/experience ofhealthcare providers (P � 0.022) was rated signif-icantly lower by bilateral wearers in comparisonwith unilateral. Prosthesis wearers with acquiredlimb absence were less satisfied with all areas ofhealth care and information provision (P � 0.05)than were those with congenital limb absence. In-dividuals with high-level limb absence were lesssatisfied than those with transradial limb absencewith information provided on technology (P �0.015), nonprosthetic options (P � 0.039), re-sources for peer support (P � 0.008), and training(P � 0.027) and fitting (P � 0.012) services.

Satisfaction levels of participants from thethree major hospitals involved in this study werecomparable in all aspects with the exception ofinformation on funding (P � 0.003), which waslower, and information on nonprosthetic options(P � 0.049) which was higher, for one hospital.Satisfaction with health care and information pro-vision was significantly lower for individuals re-cruited from community-based support groups incomparison with the hospitals (P � 0.005).

The fitting time frame emerged as an impor-tant factor in prosthesis acceptance for individualswith congenital limb absence. Prosthesis rejecterswere fitted within a median of 3.9 yrs, with aninterquartile range (IQR) of 2–6.6 yrs, whereasfrequent wearers were fitted more quickly, within11 mos (IQR: 5 mos to 1.5 yrs). A similar trend was

apparent for individuals with acquired limb ab-sence. Rejecters were fitted at a median 6 mos afteramputation, with an IQR of 3 mos to 1 yrs, whereaswearers were fitted within a median of 3 mos (IQR:2–5 mos). At the time of follow-up, pediatric wear-ers had worn prostheses for an average of 8 � 5 yrs,whereas adult wearers had worn them for 21 � 15yrs. There was no statistically significant differencebetween the times of follow-up for prosthesis wear-ers vs. rejecters.

When asked to rate their involvement in theselection of their primary prosthesis on a scalefrom 1 (no choice) to 7 (entirely my choice), fre-quent wearers reported a median rating of 5,whereas rejecters reported a significantly lower de-gree of involvement of 2 (P � 0.001). Prosthesisrejection and involvement in prosthesis selectionwere not related to the type of device selected, norwas prosthesis rejection.

Prosthesis technologyProsthesis rejecters were significantly less sat-

isfied with all aspects of prosthesis design, includ-ing appearance (P � 0.014), comfort (P � 0.001),function (P � 0.001), ease of control (P � 0.001),reliability (P � 0.001), and cost (P � 0.034). Thedistribution of prosthesis types used by bothgroups was comparable in terms of the shapes ofterminal devices used (i.e., hand or hook) and themode of actuation (i.e., passive, body powered, orelectric). Frequency of maintenance activities (i.e.,fitting, replacement of components, and repairs)was also not significantly different between fre-quent wearers and prosthesis rejecters. Of prosthe-sis rejecters, 74% stated that they might reconsiderprosthesis use if technological improvements weremade at a reasonable cost.

Relative Importance of FactorsRespondents were asked to specify the impor-

tance of a variety of factors in their choice not towear a prosthesis (a) permanently for the case ofnonusers, and (b) on a typical day when the pros-thesis is not worn for the case of current users. Thepercentages of participants who considered eachfactor to be of some importance in the decision notto wear a prosthesis and the median ratings arepresented in Table 2. Evidently, prosthesis rejectersdiscontinue use largely because of a lack of func-tional need, discomfort, and impediment to sen-sory feedback. Discomfort is also a primary factorfor occasional nonuse in frequent wearers in con-junction with reasons of necessity (i.e., the pros-thesis must be removed for certain activities likeswimming or sleeping, for medical reasons, or forrepair). Excessive weight was deemed more impor-tant by prosthesis rejecters (P � 0.001). Heat wasconsidered an important factor in prosthesis non-


wear in both groups, whereas medical factors wereconsidered to be somewhat important by both fre-quent wearers and prosthesis rejecters in the deci-sion not to wear, as was need for repairs. With thelatter three exceptions, prosthesis rejecters ratedall other factors of nonwear with greater impor-tance than frequent wearers (P � 0.001). Quality ofinformation and training services, along with otherenabling factors, such as availability of health careand prostheses, and cost, were generally not con-sidered important in the decision to reject prosthe-ses, although the latter was noted by approximately50% of individuals as a minor factor in the deci-sion. Technological factors pertaining to prosthesisdesign were paramount.

DISCUSSIONSummary of Key Findings

Rejection rates, as determined in this study,were comparable with a meta-analysis of previousliterature conducted in this area.1 Table 3 providesa summary of the key findings with respect to thepredisposing factors, enabling resources, and es-tablished need. Established need and satisfactionwith available prosthesis technology emerged as

the predominant factors in prosthesis use and re-jection. Individuals who felt that a prosthesis washelpful in their daily activities made use of thedevice, whereas those who felt it to be a hindranceor unreasonably uncomfortable/difficult to use, didnot. It is hypothesized that the decision to wear orreject a prosthesis is resolved in a manner to bestmeet personal priorities and needs. Although not aprimary factor, more than 50% of those surveyedconsidered availability or cost to play at least aminor role in the decision not to wear a prosthesis.

Merits and LimitationsProbably the most interesting attributes of the

study presented above are (a) investigation of pre-disposing factors through exploration of high-or-der correlations; (b) the ability to assess the relativeimportance of predisposing factors, enabling fac-tors, and established need on prosthesis use andabandonment; and (c) the opportunity to tap intothe opinions and experiences of nonusers recruitedoutside of the rehabilitation center.

A number of Web-based surveys have beenconducted in the past with regards to health careon a large scale (i.e., �100 participants)20–22 and

TABLE 2 Factors in the decision not to wear a prosthesis (a) permanently for prosthesis rejectersor (b) on a specific day for prosthesis wearers, based on self-reported ratings ofimportance from 0 (not at all a factor) to 3 (most important factor)

Factors in Nonwear

Prosthesis Rejecters Frequent Wearers

Percentage ofRespondentsa

MedianRating

Percentage ofRespondentsa

MedianRating

Just as or more functional without it* 98 3 60 1More comfortable without it* 95 3 66 1Too difficult or tiring to use* 88 2 39 0Too heavy* 88 2 65 1Too hot 88 2 77 2More sensory feedback without it* 85 2 44 0Inconvenience* 93 2 53 1Lifestyle 80 2 N/A N/ADissatisfaction with prosthetic technology 70 1.5 N/A N/AAppearance of the prosthesis* 70 1 33 0Medical factors (i.e., skin irritation, blisters, etc.) 55 1 64 1Stopped working and needs repair 49 0 56 1Cost 48 0 N/A N/AAvailability of prostheses 48 0 N/A N/AAvailability of healthcare services 51 1 N/A N/ALack of information about prosthetic options 28 0 N/A N/ALack of training 28 0 N/A N/ASomeone else made the decision 15 0 N/A N/AMoral, cultural, or religious reasons 8 0 N/A N/AMust be removed (i.e., for sleeping, swimming) N/A N/A 71 2Mood N/A N/A 51 0Fear of damage N/A N/A 35 0

a The percentage of respondents who considered the factor to be of some importance in the decision not to wear a prosthesis.* Factors that were rated significantly more important (P � 0.001) by prosthesis rejecters than by frequent wearers.N/A: As indicated, some factors (e.g. availability, fear of damage, etc.) were not applicable to both prosthesis rejecters and

frequent wearers


prosthetics on a smaller scale (i.e., �100 partici-pants).23,24 Internal consistency and test–retest re-liability between a large variety of questionnairesadministered via the Internet and by other means(i.e., telephone, mail, etc.) have been demonstrat-ed.22,25 However, it is important to recognize thepossible sampling biases inherent to the resultspresented herein.

1. The sample population was largely self-selected, asis often the case in consumer-based surveys, mak-ing it difficult to assess the extent to which theopinions expressed are reflective of the populationas a whole. To quantify this potential limitation,rates of prosthesis rejection in samples recruitedfrom the three primary pediatric hospitals werecompared. Response rates for the three hospitalswere approximately 40%, 55%, and 100%. No sig-nificant differences in rates of prosthesis rejectionor satisfaction with overall health care were ob-served, irrespective of the rate of self-selection.

2. It is expected that the study population based inCanada was, in general, younger, with greater

use of electric prostheses than the general pop-ulation, reflecting the expertise of BloorviewKids Rehab in myoelectric prostheses and pedi-atric care. Age (i.e., pediatric or adult) and de-vice type were not factors in prosthesis use orabandonment in either Canada or the UnitedStates.

3. To quantify possible sampling biases introducedby the use of an online survey, we compared thedemographic distribution of this study’s elec-tronic respondents from the United States (n �

92) with that of a large-scale epidemiologicstudy (n � 2477), also conducted in the UnitedStates in 1996.19 No statistically significant dif-ferences with respect to age, prevalence of tran-sradial limb absence, or origin of limb absencewere observed. The prevalence of electric handsand body-powered hooks was not statisticallydifferent. In this study, 51% of participants weremale, as compared with 63% in the Atkins etal.19 study (P � 0.02). Previous research has

TABLE 3 Summary of key findings

Factor Key Findings

Predisposing characteristicsLevel of limb absence Higher rejection rates associated with low- and high-level limb absenceOrigin of limb absence Origin of limb absence not a factor in rejection for individuals with

transradial limb absenceIndividuals with congenital amputations at a low or high level more

likely to reject prostheses than those with acquired limb absenceGender Females with acquired, high-level limb absence more likely to reject

prostheses than malesLow rate of prosthesis rejection for males with acquired limb absenceNo gender differences for individuals with congenital limb absence

Bilateral limb absence Individuals with congenital, bilateral limb absence more likely to rejectprostheses than those with unilateral limb absence

Length of residua Length of residua not associated with prosthesis rejectionDominant-hand amputation Dominant-hand amputation not associated with prosthesis rejectionMedical factors Phantom and residual limb pain and upper-body pain more prevalent in

individuals with acquired or high-level limb absence, but notassociated with prosthesis rejection

Skin irritation, blisters, and upper-body pain more prevalent in frequentwearers than in prosthesis rejecters

Age 4–10 and 24–35 age groups associated with higher rates of rejectionEstablished need

Perceived need for prosthesis Prosthesis rejection associated with lower perceived needOccupation status Prosthesis rejection higher for students than full-time workersActivity levels Self-reported activity levels not associated with prosthesis rejection

Enabling resourcesHealth care Satisfaction with overall health care lower for prosthesis rejecters

Fitting at an older age for children with congenital limb absence isassociated with prosthesis rejection, as is a longer time gap betweenamputation and fitting for individuals with acquired limb absence

Low perceived involvement in the selection of a prosthesis is associatedwith prosthesis rejection

Prosthetic technology Dissatisfaction with prosthesis technology highly associated withprosthesis rejection


found no association between gender and Inter-net access.26 It is possible, because of the po-tential bias of Internet-based surveys to excludelow-income individuals,27 that concerns regard-ing high costs of prostheses are actually under-estimated in this study. Future research in thisarea is needed.

The results of this study must be considered inthe context of these limitations and should not begeneralized to the population en masse.

CONCLUSIONSProsthesis use is highly related to established

need and enabling resources, particularly the state ofavailable technology. Future research should focuson continued development of more comfortable andfunctional prostheses, particularly for individualswith high-level or bilateral limb absence. Efforts byhealthcare providers to actively involve clients in theselection of a prosthesis most suited to their personalgoals and needs should be escalated. Individuals withlimb absence are particularly interested in resourcesfor peer support and are desirous to know all of theiroptions, including nonprosthetic. Better support,particularly with regard to information provision, isneeded for specialty groups including individualswith bilateral, acquired, or high-level limb absence.Efforts to ensure the availability of quality prosthesesand health care should continue.

ACKNOWLEDGMENTSThe authors would like to sincerely thank the

following for their feedback in the survey develop-ment: Hanna Heger, Dr. Peter Kyberd, Dr. DickPlettenberg, Virginia Wright, Rafael Reyes, JohnKrenzel, Julie Shaperman, Sheila Hubbard, Dr.Monique Gignac, Dr. Colin Macarthur, Randall Al-ley, Karen Tam, and Nora Fayed. For their promo-tion of the survey, we thank: Margriet Poelma,Joanne Shida, Shawn Swanson, Karen Lundquist,Lisa Artero, Vera van Heijningen, Tim Pauley, Dr.Michael Devlin, Karen Litman, Karen Roberts,Fiona Carnegie, Rosemary Franklin, John Billock,the Prosthetics and Orthotics Care Company, De-sign Prosthetic Appliance Company, CustomProsthetic Services Ltd., and Atlantic ProstheticsInc. We also extend our gratitude to Dan Sorkin,Joyce Baughn, Wayne Renardson, Eric Westover,Brian Wade, and Lorraine Peacock for their re-cruitment of participants from community-basedsupport groups. Thanks to Toshiba for their en-dorsement. This work was supported in part bythe Natural Sciences and Engineering ResearchCouncil of Canada, and the Canada ResearchChairs Program.

REFERENCES1. Biddiss E, Chau T: Upper extremity prosthesis use and

abandonment: a survey of the last 25 years. Prosthet OrthotInt 2007;31:236–57

2. Scotland TR, Galway HR: A long-term review of childrenwith congenital and acquired upper limb deficiency. J BoneJoint Surg (Br) 1983;65:346–9

3. Wright TW, Hagen AD, Wood MB: Prosthetic usage in majorupper extremity amputations. J Hand Surg (Am) 1995;20:619–22

4. Hubbard S, Kurtz I, Heim W, Montgomery G: Poweredprosthetic intervention in upper extremity deficiency, inHerring J, Birch J (eds): AAOS/Shrine Symposium: TheLimb Deficient Child. Rosemont, Ill, American Academy ofOrthopardic Surgeons, 1997, pp 417–31

5. Durance JP, O’Shea BJ: Upper limb amputees: a clinicprofile. Int Disabil Stud 1988;10:68–72

6. Melendez D, Leblanc M: Survey of arm amputees not wear-ing prostheses: implications for research and service. JAssoc Child Prosthet Orthot Clin [serial online]. 1988;23.Available at: http://www.acpoc.org/library/1988_03_062.asp.Accessed November 5, 2005

7. Gaine WJ, Smart C, Bransby-Zachary M: Upper limb trau-matic amputees. Review of prosthetic use. J Hand Surg (Br)1997;22:73–6

8. Datta D, Selvarajah K, Davey N: Functional outcome ofpatients with proximal upper limb deficiency—acquiredand congenital. Clin Rehabil 2004;18:172–7

9. Davidson J: A survey of the satisfaction of upper limbamputees with their prostheses, their lifestyles, and theirabilities. J Hand Ther 2002;15:62–70

10. Bigelow J, Korth M, Jacobs J, Anger N, Riddle M, Gifford J:A picture of amputees and the prosthetic situation in Haiti.Disabil Rehabil 2004;26:246–52

11. Lake C, Miguelez J: Comparative analysis of microproces-sors in upper limb prosthetics. J Prosthet Orthot [serialonline]. 2003;15. Available at: http://www.oandp.org/jpo/library/2003_02_048.asp. Accessed November 5, 2005

12. Bhaskaranand K, Bhat AK, Acharya KN: Prostheticrehabilitation in traumatic upper limb amputees (anIndian perspective). Arch Orthop Trauma Surg 2003;123:363–6

13. Biddiss E, Chau T: The roles of predisposing characteristics,established need, and enabling resources on upper extrem-ity prosthesis use and abandonment. Disabil Rehab AssistTech 2007;2:71–84

14. Anderson RM: Revisiting the behavioral model and access tomedical care: does it matter? J Health Soc Behav 1995;36:1–10

15. Kyberd P, Davey J, Morrison J: A survey of upper limbprosthesis users in Oxfordshire. J Prosthet Orthot [serialonline]. 1998;10. Available at: http://www.oandp.org/jpo/library/1998_04_085.asp. Accessed November 5, 2005

16. Fraser CM: An evaluation of the use made of cosmetic andfunctional prostheses by unilateral upper limb amputees.Prosthet Orthot Int 1998;22:216–23

17. Fraser C: A survey of users of upper limb prostheses. Br JOccup Ther 1993;56:166–8

18. Hartford K, Carey R, Mendonca J: Sampling bias in an inter-national Internet survey of diversion programs in the criminaljustice system. Eval Health Prof 2007;30:35–46

19. Atkins D, Heard D, Donovan W: Epidemiologic overview ofindividuals with upper-limb loss and their reported re-search priorities. J Prosthet Orthot 1996;9:2–11

20. Ayantunde AA, Welch NT, Parsons SL: A survey of patientsatisfaction and use of the Internet for health information.Int J Clin Pract 2007;61:458–62

21. Legters K, Barber Verbus N, Kitchen S, Tomecsko J, UrbanN: Fear of falling, balance confidence and health-relatedquality of life in individuals with postpolio syndrome. Phys-iother Theory Pract 2006;22:127–35


22. Ritter P, Lorig K, Laurent D, Matthew K: Internet versusmailed questionnaires: a randomized comparison. J MedInternet Res 2004;6:e29

23. Murray CD, Fox J: Body image and prosthesis satisfaction inthe lower limb amputee. Disabil Rehabil 2002;24:925–31

24. Murray CD: The social meanings of prosthesis use. J HealthPsychol 2005;10:425–41

25. Graham AL, Papandonatos GD, Bock BC, et al: Internet- vs.telephone-administered questionnaires in a randomizedtrial of smoking cessation. Nicotine Tob Res 2006;8:S49–57

26. Ono H, Zavodny M: Gender and the Internet. Soc Sci Q2003;84:111–21

27. Statistics Canada. Canadian Internet use survey. Available at:http://www.statcan.ca/Daily/English/060815/d060815b.htm. Ac-cessed March 8, 2007


Authors:Barbara A. Crane, PhD, PT, ATPMargo B. Holm, PhD, OTR/LDouglas Hobson, PhDRory A. Cooper, PhDMatthew P. Reed, PhD

Affiliations:University of Hartford, PhysicalTherapy Department, College ofEducation, Nursing and HealthProfessions, West Hartford,Connecticut (BAC); Departments ofOccupational Therapy (MBH) andRehabilitation Science andTechnology (DH, RAC), School ofHealth and Rehabilitation Sciences,University of Pittsburgh, Pittsburgh,Pennsylvania; University of MichiganTransportation Research Institute,Ann Arbor, Michigan (MPR).

Correspondence:All correspondence and requests forreprints should be addressed toBarbara Crane, University of Hartford,Dana 410C, 200 Bloomfield Ave, WestHartford, CT 06117.

Disclosures:This research was conducted at theSchool of Health and RehabilitationScience, Department of RehabilitationScience and Technology, University ofPittsburgh. The study was supportedby National Institute on DisabilityRehabilitation and Research, grant#H133E990001.


DOI: 10.1097/PHM.0b013e3181583ed9

A Dynamic Seating Intervention forWheelchair Seating Discomfort

ABSTRACT

Crane BA, Holm MB, Hobson D, Cooper RA, Reed MP: A dynamic seatingintervention for wheelchair seating discomfort. Am J Phys Med Rehabil 2007;86:988–993.

Objective: The objective of this study was to examine the effectivenessof a new user-adjustable wheelchair seating system designed to relievediscomfort for long-duration wheelchair users.

Design: This objective was carried out using the newly developed Toolfor Assessing Wheelchair disComfort (TAWC) as the primary outcomemeasure. Two wheelchair users each tested two different designs andfeedback from the wheelchair users regarding the first design was used toguide development of the second design. A single-subject research meth-odology was used, allowing long-duration (up to 2 wks per test) evaluationof the wheelchair seating systems and comparison of subject discomfortlevels with those experienced during a baseline period using their ownwheelchairs. The experimental wheelchair seating systems employed ex-isting automotive seating with embedded pneumatic bladders that allowedadjustment of the seat and back-support characteristics. The test wheel-chair also had tilt, recline, and elevating leg rests.

Results: The two subjects completed limited periods of testing with thefirst design, both finding poor results with either stable or increased levelsof discomfort. Subject feedback was used to redesign the wheelchair seat.After redesign, both subjects tested the second design and found itsubstantially more comfortable.

Conclusions: The selected research methodology was a very positivemethod for a progressive wheelchair seating design and the seconddesign provided improved comfort for both users when compared withthat experienced using their own wheelchairs and the first test wheelchair.Future research of this type of user-controlled technology is recom-mended.

Key Words: Wheelchair Seating, Wheelchair, Discomfort, Disability


RESEARCH ARTICLE

Wheelchair

Long-duration wheelchair users (i.e., greaterthan 8 hrs of use per day) with intact sensationfrequently experience problematic levels of discom-fort.1–3 Discomfort has been shown to negativelyinfluence consumer satisfaction,4 cause decreasedquality of life,5 and lead to problems related towheelchair propulsion ergonomics6 and adoptionof poor sitting postures7—all of which may impaireveryday function and the ability to remain seatedin a wheelchair.8 Yet, few researchers have inves-tigated the nature and causes of wheelchair seatdiscomfort, or possible solutions to this problem.1,9

However, the automotive seating industry has donesignificant product design research to meet thecomfort needs of long-duration drivers. Thereforewe decided to try and incorporate automotive tech-nologies into wheelchair seating. An additional ad-vantage of this would be the ability to providelower-cost product options for wheelchair users,because these products are already commerciallyavailable in a much larger market than the wheel-chair seating market.

Anecdotal experience by clinicians suggeststhat giving users greater control of the seatingsystem increases satisfaction, but no systematicstudies have supported this observation. Precedingwork in this research program developed a vali-dated methodology for assessing the affect ofwheelchair technology on sitter comfort. The cur-rent paper describes the application of this meth-odology to the development and refinement of anadvanced wheelchair seating system. The system,which is based on off-the-shelf commercial com-ponents, was intended to be a test bed for technol-ogies rather than a commercial system. As such,the findings reported here may encourage develop-ers to invest in more configurable seating systemsfor wheelchair users. The purpose of this study wasto examine the effectiveness of a this wheelchairseating system designed to relieve discomfort forlong-duration wheelchair users.

METHODSThis study involved an iterative process of test-

ing two prototype powered wheelchair seating de-vices constructed on a Permobil powered wheel-chair base. Although the initial seating systemdesign was based on laboratory-based research per-formed before this study, it was not clear to theresearchers that the initial prototype would actu-ally prove to be an effective intervention. The ulti-mate goal of designing an effective dynamic seatingintervention to address the comfort needs of thispopulation of wheelchair users was highly depen-dent on the outcomes of each design and testingphase. In this study, the initial prototype wastested, but when initial results indicated poor ac-

complishment of the ultimate comfort-relatedgoal, the study design allowed flexibility to removethe initial prototype from further testing and rede-sign the seating according to the feedback receivedfrom the subjects.

The Permobil powered wheelchair used fortesting included powered chair tilt, powered backrecline, powered elevating leg rests, and a poweredseat lift—all currently available wheelchair seatingfeatures. An automotive bucket-style seat wasmounted on the Permobil wheelchair base. Thisseat consisted of a foam contoured seat and backwith cloth upholstery. Air bladders were embeddedin both the seat and back cushions underneath thefoam cushions. Four air bladders were added to theseat, and three bladders were added to the back (seeFig. 1). These air bladders were connected to bat-tery-powered pumps, and the wheelchair user hadcontrols to inflate or deflate each bladder to adjustthe support characteristics of the cushion. Thewheelchair controls were adjusted until the partic-ipant was satisfied and able to independently oper-ate all of the seating and wheelchair controls.

The Tool for Assessing Wheelchair disComfort(TAWC),10,11 a previously developed and validatedtool, was the main outcome measure, allowingquantification of subject discomfort in a real-life

FIGURE 1 Arrangement of air bladders in the testwheelchairs.

December 2007 Wheelchair Seating Discomfort 989

testing process following a single-system, withinsubjects research design. Data from the TAWCwere used to guide modifications of the wheelchairseating system and to quantify its effectiveness inreducing discomfort. After testing of the first pro-totype, the seat of the system was redesigned, andthe foam structure was replaced with a custom-made, four-quadrant Roho air cushion on the basisof feedback from the participants that the automo-tive seat was too firm. In addition to the TAWCdiscomfort scores, subjects monitored their skindaily for any changes in skin integrity for safety. Ifany redness was detected after sitting in the testwheelchair, the subject was instructed to stop us-ing the wheelchair and contact the researchers.The total amount of time sitting in the wheelchaireach day was also monitored via a log completed bythe subject each day.

An ABCA single-subject design test protocolwas used. Phase A was the baseline phase, duringwhich the subjects used their own wheelchairs.Phase B involved using the prototype wheelchairwith only the traditional powered wheelchair fea-tures operational. The traditional features usedwere power seat tilt, power back recline, powerelevating leg rests, and power seat lift. Phase Cinvolved using the prototype wheelchair with thetraditional wheelchair features as in Phase B, withthe new air bladder features operational. Subjectswere randomly assigned to either an ABCA or anACBA test order to minimize possible interferenceof an order effect.

Two wheelchair users were recruited for thisstudy. The participants reported using poweredwheelchairs for an average duration of 8 hrs/day.They had severe motor impairment and intact sen-sation on their buttocks and lower extremities andexperienced discomfort associated with sitting intheir wheelchairs. Neither of the participants hadexperienced skin breakdown on their seating sur-faces (buttocks or posterior thighs) within the yearbefore enrollment. Both participants were usingPermobil power wheelchairs before the study. Thefirst participant was using a Permobil standingframe power wheelchair when he was initially re-cruited, but he later received a Permobil powerwheelchair with power tilt and recline features.Participant 1 used a high-profile Roho seat cush-ion, and participant 2 used standard Permobilseating including mildly contoured seat and backcushions.

The TAWC was used to quantify sitting discom-fort experienced by the test subjects under all testconditions. This tool is divided into sections thatresult in two discomfort scores.10 The General Dis-comfort Assessment score (GDA) contains eightstatements related to discomfort and five state-ments related to comfort, rated on a seven-point

Likert scale. The GDA score has a possible range of13–91, with higher scores representing increasedlevels of discomfort. GDA scores were recorded foreach 4-hr period of time spent sitting. The Discom-fort Intensity Score (DIS) allows subjects to ratelevel of discomfort in eight body areas and in thebody as a whole. The DIS score has a range of 8–99.A score of 8 indicates no discomfort in any part ofthe body, and a score of 99 indicates a maximumamount of discomfort in eight body areas and inthe body as a whole. The reliability and validity ofthe TAWC have been established and reported in anearlier publication.11

Before data collection, participants signed in-formed consent documents approved by the Uni-versity of Pittsburgh’s institutional review board.For the baseline measures discussed previously,participants completed daily logs and TAWC ques-tionnaires. The researcher contacted each partici-pant several times per week to answer any ques-tions and clarify procedures as needed. Participantswere also encouraged to call or page the researcherif there were any problems or question regardingeither the documentation or the test wheelchair.The researcher visited with each participant weeklyto transition the participant from one interventionto the next and to collect data.

Both traditional graphic visual analyses12 andspecialized semistatistical and statistical proce-dures designed for use in single-subject design,12,13

were used to analyze the effectiveness of the seat-ing interventions. The comfort effects of three seat-ing systems were compared. The baseline phases(A1 and A2) (participants’ own chairs) were com-pared with a test chair equipped with traditionalpowered seating options (B1 and B2) and the sametest chair equipped with traditional powered seat-ing options and new user-adjustable seating com-ponents (adjustable air bladder systems) (C1 andC2). The two intervention systems were also com-pared with each other.

All discomfort-related data were manually en-tered into an SPSS data file, then summary datawere transferred into a Microsoft Excel spreadsheetand graphs were developed for performing visualanalyses. Semistatistical and statistical proceduresused included celeration line analyses and theTryon C statistic.14 Before any visual or statisticalanalysis, the Bartlett test of the lag-1 autocorrela-tion coefficients12 was performed on each phase ofdata collected for each participant. When serialdependency of the data were found, the C statisticalone was relied on for indications of a significantintervention effect, because the C statistic remainseffective even with data that are serially depen-dent.14

990 Crane et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

RESULTSBoth subjects completed partial tests of chair

1. The chair was removed early in both cases be-cause of negative outcomes evidenced by the dis-comfort scores and development of reddened areason the buttocks, a preestablished stopping point forsafety. Discomfort scores for chair 1 testing aresummarized in Table 1. Both subjects also testedchair 2 after the redesign process. This testingallowed comparison of the two chair designs as wellas examining the effectiveness of the new design.Table 2 contains chair 2 summary results.

Subject 1This subject used a Permobil Chairman model

powered wheelchair with a high-profile Roho cush-ion for the A phases of his first trial. He completedphases A, C, and A when testing the first wheelchairdesign. He had been randomized to an ACBA de-sign, but he experienced redness under one of hisischial tuberosities after 5 days of phase C testingwith this first prototype, and therefore the test wasended for safety reasons. He did not progress tophase B; instead, he returned to his own wheel-chair for his return to baseline phase (phase A). HisGDA and DIS discomfort scores changed littlewhile he used the prototype wheelchair. The scoreswere slightly lower, but they were not particularlystable, and none of the semistatistical tests indi-cated a significant difference when using the testwheelchair. The C statistic analysis indicated sig-nificantly greater discomfort levels on both theGDA and DIS measures when he was using the firsttest wheelchair.

After the redesign of the test wheelchair, hewas contacted and asked whether he would beinterested in testing the redesigned wheelchair. Heagreed, but because he had just obtained a newpersonal wheelchair, he underwent a new trial,including new baseline-phase data collection. Hisnew wheelchair was also a Permobil power wheel-chair base, but instead of a standing feature, hisnew wheelchair had power seating options includ-ing power seat tilt and power back recline. He wasassigned the same testing schedule: ACBA, for con-sistency with his first round of testing.

Once again, his mean GDA and DIS discomfortscores changed little across phases. Trend linesillustrated increasing levels of discomfort duringall phases of testing. However, the slopes of thetrend lines were slightly decreased during phases Band C. On return to baseline, mean discomfortincreased and the trend reversed in slope, indicat-ing increasing discomfort when he returned tousing his own wheelchair. The DIS data C statisticsindicated significantly less discomfort during phaseB (z � 2.55) and significantly more discomfort

during phase C (z � 2.78) and his final baselinephase (z � 2.37). This indicates lower levels ofdiscomfort with the test wheelchair with tradi-tional power seating, but no further decrease indiscomfort with the addition of the new, user-adjustable seat features.

Subject 2The second subject was randomized to an

ABCA testing order. During his baseline phases, heused a Permobil power wheelchair with the stan-dard Permobil foam seat and back cushions. Hisown wheelchair had power seat tilt, power backrecline, and power elevating leg rests. His firstchair trial was interrupted during phase B becauseof complaints of increased discomfort and difficultyusing chair 1 in his home environment.

This participant’s mean discomfort scores in-dicated greater discomfort during phase B with thefirst chair than during his baseline phase (phase A).

TABLE 1 Chair 1 testing: means and standard deviations ofsubject discomfort scores

Subject

Phase

A1 B1 C1 A2

GDAscore

1 57.1 (6.4) * 52.7 (4.2) 49.6 (4.8)2 39.6 (9.3) 54.4 (3.9) * 40.1 (10.4)

DISscore

1 33.8 (5.1) * 27.4 (2.8) 24.9 (1.9)2 12.7 (1.0) 14.8 (1.3) * 12.9 (0.9)

GDA, General Discomfort Assessment score: 13–91, with higher scores indi-cating greater discomfort; DIS, Discomfort Intensity Score: 8–99, with higherscores indicating greater discomfort.

* No data collected for these phases.

TABLE 2 Chair 2 testing: means and standard deviations ofsubject discomfort scores

Subject

Phase

A1 B2 C2 A2

GDAscore

1 50.2 (3.0) 49.5 (2.2) 50.0 (3.2) 51.7 (1.9)2 39.6 (9.3) 36.0 (8.8) 43.6 (10.2) 40.1 (10.4)

DISscore

1 21.4 (1.6) 19.5 (1.0) 21.3 (2.1) 20.6 (1.2)2 12.7 (1.0) 12.5 (1.4) 12.9 (0.2) 12.9 (0.9)

GDA, General Discomfort Assessment score: 13–91, with higher scores indi-cating greater discomfort; DIS, Discomfort Intensity Score: 8–99, with higherscores indicating greater discomfort.


Trends in discomfort levels indicated stable levelsof discomfort during both baseline phases and in-creasing levels of discomfort during phase B. DISscore trends followed a similar pattern as thosefound with GDA scores with increasing discomfortlevels during phase B (see Fig. 2). The result of theC statistic testing of DIS scores for phase B indi-cated a significantly higher discomfort level (z �2.53) with this intervention.

After redesign of the test wheelchair, this par-ticipant agreed to complete his trial using chair 2.Because his own wheelchair had not changed, thetest wheelchair was reintroduced, and he com-pleted phases B and C and then returned to usinghis own wheelchair for his final baseline phase.Celeration line testing of the GDA scores indicatedgreater discomfort during phase C than duringphase B. Overall, GDA scores were quite variable,but with a consistent pattern each day: lower scoresin the morning, and higher scores at the end ofeach day. DIS scores for this participant wereslightly less variable overall, yet they still exhibiteda similar pattern. Celeration line testing of thesescores indicated lower discomfort during phases Band C, but greater discomfort during phase C whencompared with B. Figure 2 illustrates the DIS scoremeans and trends. C statistic testing of the DISscores from phase B with the first wheelchair and

phase B with the second wheelchair was significant(z � 2.69), indicating a significantly lower discom-fort level with chair 2. This was indicative of im-proved comfort with chair 2 when compared withChair 1 with the same features.

DISCUSSIONThe newly developed TAWC subjective assess-

ment tool and associated clinical measures wereused in an iterative testing process to develop aprototype wheelchair seating system designed tominimize discomfort of wheelchair users. Chair 1,the first prototype, underwent initial testing byboth participants. The TAWC measurements dem-onstrated that the chair 1 was not providing theexpected benefits. Because of the results from theinitial testing, the wheelchair seating system wasredesigned and then retested as chair 2. The sameparticipants tested chair 2 after this redesign, andthe TAWC was again used to assess the effective-ness of the new intervention.

One concern with using a within-subjects de-sign in research is autocorrelated data. Because ofthis concern, all data were tested for autocorrela-tion. In these two subjects only data from twophases resulted in autocorrelated data. For theseresults, the C statistic was relied on for testing ofsignificant results. The remainder of the data were

FIGURE 2 Participant 2: means (upper panel) and trends (lower panel) for DIS scores.

992 Crane et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

not autocorrelated, allowing for both visual andsemistatistical analysis with all included methods.

The results of this testing must be consideredin light of the limitations of this particular re-search. The difficulties involved in introducing anovel wheelchair design to these participants can-not be overlooked. The wheelchair base itself,which was not being studied, may have causedenough difficulty for the subjects to interfere withthe ability to truly examine the effects of the seat-ing intervention. The fit and function of the wheel-chair seat also was suboptimal for the participants.Because of the close relationship between fit andcomfort, this may have caused suboptimal levels ofcomfort for the subjects unrelated to the dynamicfeature effects.

Recommendations for future study include (1)studying “standalone” seating products intendedfor use in the participants’ own wheelchairs; (2)reducing the TAWC scoring to twice per day—onceat 6 hrs and once at bed time—because of therelatively low levels of discomfort reported withshorter sitting durations; (3) reducing the data-collection labor involved on the part of the partic-ipants (perhaps by automating more of the data-collection process) and extending the phaselengths to 8–10 days per phase to attain greaterstability of trends; and (4) increasing the numberand pattern of alternating phases—for example,studying one intervention at a time and using anABABA design.

CONCLUSIONThe goal of this research was to assess the

effectiveness of a new, user-adjustable wheelchairseating system in alleviating discomfort. A secondarygoal was to use the subject data in further develop-ment of the wheelchair seating design. These goalswere accomplished through this within-subjectsmethodology. The first wheelchair design, developedafter short-duration laboratory testing, did not meetthe goal of reducing seating discomfort with long-duration sitting. Therefore, this wheelchair was re-called and redesigned. On construction of the newdesign, the wheelchair was reintroduced and retestedby these two participants. This method of testingwas very helpful in designing a wheelchair that wasable to meet the goal of enhancing comfort of usersfrom the target population. The within-subjectsdesign allowed for in-depth investigation of theeffectiveness of a user-adjustable seating system forlong-duration wheelchair users with intact sensa-tion, using three measures of discomfort.

ACKNOWLEDGMENTSThe authors would like to thank Mark Fried-

man, PhD and Steven Stadelmeier, MSc for theirtechnical assistance in the design and constructionof both of the test wheelchairs (test chair 1 and testchair 2) used for this study. The following manu-facturers supplied equipment used for this re-search. The power wheelchair base was donated byPermobil, USA, and the specialized seat cushionused in the test chair 2 design was provided by TheRoho Group. In addition, I would like to thank thewheelchair users who participated in this study fortheir great commitment of time and dedication tothis complex data-collection process.

REFERENCES1. Monette M, Weiss-Lambrou R, Dansereau J: In search of a

better understanding of wheelchair sitting comfort anddiscomfort. Paper presented at: RESNA annual conference,Long Beach, CA, 1999

2. Shaw G, Taylor S: A survey of wheelchair seating problemsof the institutionalized elderly. Assist Technol 1991;3:5–10

3. Scherer MJ: Users desires for wheelchairs. Rehab Manag1996:121–3

4. Weiss-Lambrou R, Tremblay C, LeBlanc R, Lacoste M,Dansereau J: Wheelchair seating aids: how satisfied areconsumers? Assist Technol 1999;11:43–53

5. Herzberg S: Positioning the nursing home resident: anissue of quality of life. Am J Occup Ther 1993;47:75–7

6. DiGiovine MM, Cooper RA, Boninger ML, Lawrence B, Van-Sickle DP, Renschler A: User assessment of manual wheel-chair ride comfort and ergonomics. Arch Phys Med Rehabil2000;81:490–4

7. Engstrom B: Ergonomic Seating: A True Challenge—Seat-ing and Mobility for the Physically Challenged, Risks andPossibilities When Using Wheelchairs, vol 1, ed 2. Stock-holm, Posturalis Books, 2002

8. Crane B, Hobson D: The importance of comfort to wheel-chair users—a preliminary study. Paper presented at: 18thInternational Seating Symposium; March 7–9, 2002; Van-couver, Canada

9. Shaw G: Wheelchair seat comfort for the institutionalizedelderly. Assist Technol 1992;3:11–23

10. Crane B, Holm MB, Hobson D, Cooper RA, Reed MP,Stadelmeier S: Development of a consumer-driven Wheel-chair Seating Discomfort Assessment Tool (WCS-DAT). IntJ Rehabil Res 2003;27:85–90

11. Crane B, Holm MB, Hobson D, Cooper RA, Reed MP,Stadelmeier S: Test-retest reliability, internal item consis-tency, and concurrent validity of the Wheelchair SeatingDiscomfort Assessment Tool. Assist Technol 2005;17:98–107

12. Ottenbacher KJ: Evaluating Clinical Change: Strategies forOccupational and Physical Therapists, vol 1. Baltimore,Williams and Wilkins, 1986

13. Franklin RD, Allison DB, Gorman BS (eds): Design andAnalysis of Single-Case Research. Mahwah, Lawrence Erl-baum Associates, 1996

14. Tryon WW: A simplified time-series analysis for evaluat-ing treatment interventions. J Appl Behav Anal 1982;15:423–9


CME Objectives:On completion of this article, the readershould be able to (1) define thebiochemical rationale for the use ofcyproheptadine in the treatment ofintrathecal baclofen withdrawalsyndrome, (2) identify the signs andsymptoms of intrathecal baclofenwithdrawal, and (3) formulate aneffective treatment plan for thetreatment of intrathecal baclofenwithdrawal syndrome.

Level: Advanced.

Accreditation: The Association ofAcademic Physiatrists is accredited bythe Accreditation Council forContinuing Medical Education toprovide continuing medical educationfor physicians. The Association ofAcademic Physiatrists designates thiscontinuing medical education activityfor a maximum of 1.5 credits inCategory 1 of the Physician’sRecognition Award of the AmericanMedical Association. Each physicianshould claim only those credits that heor she actually spent in the educationactivity.

Disclosures: Disclosure statementshave been obtained regarding theauthors’ relationships with financialsupporters of this activity. There are noapparent conflicts of interest related tothe context of participation of theauthors of this article.0894-9115/07/8612-0994/0American Journal of Physical Medicine& RehabilitationCopyright © 2007 by LippincottWilliams & Wilkins

DOI: 10.1097/PHM.0b013e31815b5c58

Cyproheptadine for PediatricIntrathecal Baclofen WithdrawalA Case Report

ABSTRACT

Saveika JA, Shelton JE: Cyproheptadine for pediatric intrathecal baclofen with-drawal: a case report. Am J Phys Med Rehabil 2007;86:994–997.

Intrathecal baclofen withdrawal syndrome is a known complication of intrathecalbaclofen pumps. Its origin is postulated as an independent form of a serotonergicsyndrome occurring from loss of �-aminobutyric acid B receptor–mediatedinhibition of serotonin. Prodromal symptoms include pruritis, a return of deeptendon reflexes, and increased spastic hypertonia. Previous reports have docu-mented use of cyproheptadine in treatment of this syndrome in adults withpositive results. We present the case of a 14-yr-old child with cerebral palsy whodeveloped pruritis and worsening spastic hypertonia 18 mos after pump implan-tation. She had been previously treated with 520 �g/day of intrathecal baclofen.Progression of her symptoms was successfully arrested by the administration ofboth oral and intrathecal baclofen and 6 mg of oral cyproheptadine every 6 hrsfor 1 day. We postulate that cyproheptadine should be considered a valuableadjuvant therapy for treatment of suspected intrathecal baclofen withdrawalsyndrome.

Key Words: Baclofen, Cyproheptadine, Withdrawal Symptoms, Rehabilitation


CME ARTICLE ● 2007 SERIES ● NUMBER 5

Spasticity

Authors:Joseph A. Saveika, MDJean E. Shelton, MD

Affiliations:From the Department of Physical Medicine andRehabilitation, Eastern Virginia Medical School,Norfolk, Virginia.

Correspondence:All correspondence and requests for reprints should beaddressed to Joseph A. Saveika, MD, Eastern VirginiaMedical School, Department of Physical Medicine andRehabilitation, Hofheimer Hall Suite 646, 825 FairfaxAvenue, Norfolk, VA 23507.

The value of intrathecal baclofen (ITB) therapyin spastic hypertonia of cerebral or spinal originhas been well established.1,2 Unfortunately, ITBdelivery has been associated with a number ofcomplications ranging in severity from inconve-nient to life threatening. The most significant ofthese include pump and catheter malfunctions,withdrawal syndromes, and infection necessitatingremoval of hardware.3

In ITB administration, one of the most con-cerning complications is interruption of medica-tion administration. This interruption is a com-mon end result of pump malfunction, catheterblockage, or catheter dissociation. Interruption ofITB administration has been shown to lead to awithdrawal syndrome. This syndrome has a widerange of severity, from increased spastic hypertoniaand pruritis to seizures and audiovisual hallucina-tions. In its most severe form, ITB withdrawalsyndrome presents a constellation of symptomsdescribed as mimicking autonomic dysreflexia,neuroleptic–malignant syndrome, or malignanthyperthermia.4 The associated morbidity and mor-tality of ITB withdrawal syndrome has been welldocumented. Given the resemblance of ITB with-drawal syndrome to the above syndromes, the stan-dard of treatment has thus far been the adminis-tration of baclofen and benzodiazepines, with somearguments for the administration of dantrolene.4–7

The most recent development in the treatmentof ITB withdrawal syndrome was proposed byMeythaler and associates,8 who propose the use ofcyproheptadine for treatment of ITB withdrawal.Their logic was based on the similarity of ITBwithdrawal syndrome to serotonergic syndromes,as in overdoses of selective serotonin reuptake in-hibitors. They postulated that ITB withdrawal syn-drome may actually be a form of serotonergic syn-drome caused by downregulation of �-aminobutyric acid (GABA) B receptors, and that addition

of a serotonergic antagonist would prove of value.This concept led to the successful administration ofcyproheptadine prophylactically in patients dis-playing early signs and symptoms of ITB with-drawal at the authors’ institution.

Given the increasing use of ITB therapy forspasticity in the pediatric patient, ITB withdrawalsyndrome is likely to become more common, and aneed for treatment of the syndrome has arisen.9 Wepresent the first published use of cyproheptadine inITB withdrawal syndrome in a pediatric patient,and we propose that its use in addition to theadministration of baclofen and benzodiazepinesmay prove of benefit to the pediatric physiatristwho manages children receiving ITB therapy.

CASE REPORTA 50-kg, 14-yr-old female with a history of

cerebral palsy and spastic tetraplegia was previ-ously treated successfully with ITB for 18 mos. Sheunderwent an uneventful pump refill in an outpa-tient setting. Her reservoir returned less than 1 mlof baclofen, but no low-volume alarm was noted.She returned home without incident. Approxi-mately 8 hrs later, she developed severe pruritisand increased spasticity. She was taken to theemergency room of a pediatric hospital, where shewas evaluated and found to have a low-grade fever(37.8°C), tachycardia (pulse 123 beats per minute),hyperreflexia, clonus, and increased spasticity. Ag-itation and feelings of fear were noted. Her lower-extremity modified Ashworth scale was increasedfrom 1� to 3, her deep tendon reflexes were in-creased from 0 to 3�, and her right ankle displayedseven to eight beats of clonus, increased from two.Her left ankle displayed no clonus.

In the emergency department, her pump wasinterrogated and found to be functioning normallyat her previously programmed dose of 520 �g/day,simple continuous administration. Radiographicevaluation of pump and catheter were withoutchange from prior studies, with pump and catheterin place and catheter tip terminating at the supe-rior endplate of the sixth thoracic vertebrae. Shewas given 20 mg of baclofen orally with 6 mg ofcyproheptadine (one and one half 4-mg tablets) fora diagnosis of presumed ITB withdrawal. Symp-toms abated dramatically within 1 hr. Her agitationdiminished, her tachycardia and fever resolved, andher spasticity, hyperreflexia, and increased rightankle clonus likewise returned to her original base-line. After telephone consultation with technicalrepresentatives of the pump manufacturer, a diag-nosis of pump stall was reached, and a 50-�g in-trathecal bolus of baclofen was administered pro-phylactically via pump. She was admitted forovernight observation with orders for 20 mg of oral

December 2007 Intrathecal Baclofen Withdrawal Treatment 995

baclofen every 6 hrs and 6 mg of cyproheptadineevery 6 hrs for symptoms of ITB withdrawal.

On the morning after admission, she was againevaluated and found to have remained afebrileovernight. Her lower-extremity spasticity was a 1�on the modified Ashworth scale; she displayed onlytwo beats of clonus at the right ankle. Her deeptendon reflexes could not be elicited. Her pumpresidual volume was changed from 2 to 3 ml toavoid recurrences of ITB withdrawal. After com-pleting 24 hrs of observation, she was dischargedhome with a prescription for oral baclofen andcyproheptadine for use if necessary after telephoneconsultation with a physician.

DISCUSSIONBaclofen is a selective ligand of the bicuculline

insensitive GABAB receptors present on primary af-ferent terminals and highly concentrated in laminaeI to IV of the dorsal horn of the spinal cord. Onbinding, it inhibits the influx of calcium, which pre-vents the release of the excitatory neurotransmittersglutamate and aspartate. Receptor downregulationwas shown in mouse models by Hwang and Wilcox.10

In humans, oral baclofen is routinely used as a skel-etal muscle relaxant for spasticity. Central nervoussystem depression limits the maximal allowable oraldose. The concept of intrathecal administration is todeliver the medication directly to its site of action inthe spinal cord, allowing for much smaller doses andminimizing side effects. Unfortunately, the introduc-tion of ITB therapy also introduced new complica-tions, related both to the hardware implantation andmedication administration.

Intrathecal baclofen therapy has been relatedto both baclofen overdose and withdrawal. Over-dose signs and symptoms include flaccidity, hy-poreflexia, respiratory depression, apnea, seizures,coma, autonomic instability, hallucinations, hypo-thermia, and cardiac conduction abnormalities.11

Management includes supportive care and removalof as much baclofen as possible from the hardware.If not contraindicated, physostigmine may be usedto reduce central side effects such as somnolenceand respiratory depression.12

The use of benzodiazepines and dantrolenesodium in the treatment of ITB withdrawal syn-drome has been drawn from experience in thetreatment of neuroleptic malignant syndrome(NMS).8 Both NMS and ITB withdrawal syndromedo share some similarities in presentation. Themost overlap has been observed in autonomicsymptoms, with fever, tachycardia, and hypoten-sion documented in cases of each,4 and as observedin our patient. However, NMS is thought to becaused by dopamine blockade or reduced transmis-sion to the thalamus. Intrathecal baclofen with-drawal syndrome, on the other hand, has been

linked to serotonin syndrome. Meythaler and asso-ciates8 have advanced the theory that GABA inhib-its the release of serotonin through GABAB recep-tors at the level of the brainstem. The serotonergicpathway seems to adjust to this long-term inhibi-tion. With abrupt discontinuation of this inhibi-tion, as in our patient, the system seems to swingback toward release of excessive serotonin. Thismanifests in the patient as a form of serotoninsyndrome, forming the basis for the use of cypro-heptadine in its treatment.

Although ITB withdrawal syndrome is perhapsthe most feared complication of ITB therapy, nodefined treatment algorithm exists for this entity inthe pediatric population. Current pediatric treat-ment standards are based on experiences in theadult population and consist of administration ofbaclofen, either orally or intrathecally, and benzo-diazepines.4 In our case, the use of oral cyprohep-tadine in conjunction with oral baclofen allowedsufficient time for pump interrogation and admin-istration of intrathecal baclofen bolus, without pro-gression of the patient’s withdrawal symptoms. Al-though it is possible that the symptom abatementwas attributable entirely to the administration oforal baclofen, one would not expect such a pro-found response to oral baclofen in a patient receiv-ing ITB. In addition, the cyproheptadine providedexcellent subjective symptomatic relief of pruritis.Pruritis is considered a frequent but inconsistentsymptom of ITB withdrawal syndrome. Althoughnot pathognomic, it helps differentiate betweenITB withdrawal and aggravation of spasticity stem-ming from other causes.13 In our experience, theappearance of pruritis and its relief by the admin-istration of cyproheptadine helped guide clinicaldecision making to proceed with intrathecal bolusof baclofen, despite a normal diagnostic interroga-tion of the patient’s intrathecal pump.

Our success in this case has led to a change inthe standard education of the family of the ITBpump implantation patient. Previous education didnot emphasize pruritis, but we now provide infor-mation regarding the symptoms of ITB withdrawalsyndrome, with particular emphasis on return ofspasticity and the appearance of pruritis. All pa-tients are now discharged with prescriptions forcyproheptadine, in addition to oral baclofen. In theevent of subjective increase of spasticity or pruriticsymptoms, patient families are given instructionsto immediately call their physician. If instructed,they administer oral baclofen and cyproheptadineand proceed to the emergency room of our insti-tution. We continue to provide families with infor-mation to distribute to medical staff of other hos-pitals if they are unable to travel to our institutionor are traveling themselves.

996 Saveika and Shelton Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

As illustrated in the case report, prompt initi-ation of cyproheptadine in conjunction with thestandard administration of oral and intrathecal ba-clofen proved successful in averting the progres-sion of ITB withdrawal syndrome in our patient.We recommend that its use in children receivefurther study, given the prevalence of implantableITB pumps in this patient population.

REFERENCES1. Parke B, Penn RD, Savoy SM, Corcos D: Functional out-

come after delivery of intrathecal baclofen. Arch Phys MedRehabil 1989;70:30–2

2. Albright AL, Gilmartin R, Swift D, et al : Long-term intra-thecal baclofen therapy for severe spasticity of cerebralorigin. J Neurosurg 2003;98:291–5

3. Steinbock P, O’Donnel M: Baclofen infusion for spasticcerebral palsy. Clin Neurosurg 2000;47:440–57

4. Coffey RJ, Edgar TS, Francisco GE, Graziani V, MeythalerJM, Ridgley PM: Abrupt withdrawal from intrathecalbaclofen: recognition and management of a potentially life-threatening syndrome. Arch Phys Med Rehabil 2002;83:735–41

5. Douglas AF, Weiner HL, Schwartz DR: Prolonged intrathe-cal baclofen withdrawal syndrome. Case report and discus-

sion of current therapeutic management. J Neurosurg2005;102:1133–6

6. Green LB, Nelson VS: Death after acute withdrawal ofintrathecal baclofen: case report and literature review. ArchPhys Med Rehabil 1999;80:1600–4

7. Khorasani A, Peruzzi WT: Dantrolene treatment for abruptintrathecal baclofen withdrawal. Anesth Analg 1995;80:1054–6

8. Meythaler JM, Roper JF, Brunner RC: Cyproheptadine forintrathecal baclofen withdrawal. Arch Phys Med Rehabil2003;84:638–42

9. Zuckerbraun NS, Ferson SS, Albright AL, Vogeley E: Intra-thecal baclofen withdrawal: emergent recognition and man-agement. Pediatr Emerg Care 2004;20:759–64

10. Hwang SA, Wilcox GL: Baclofen, �-aminobutyric acidB re-ceptors and substance P in the mouse spinal cord. J Phar-macol Exp Ther 1989;248:1026–33

11. Shirley KW, Kothare S, Piatt JH, Adirim TA: Intrathecalbaclofen overdose and withdrawal. Pediatr Emerg Care2006;22:258–61

12. Darbari FP, Melvin JJ, Piatt JJ, Adirim TA, Kothare SV:Intrathecal baclofen overdose followed by withdrawal: clin-ical and EEG features. Pediatr Neurol 2005;33:373–7

13. Ben Smail D, Hugeron C, Denys P, Bussel B: Pruritis afterintrathecal baclofen withdrawal: a retrospective study. ArchPhys Med Rehabil 2005;86:494–7

MEDIA REVIEW

Physical Medicine and Rehabilitation Board Review: Palm Editionby USBMIS. Developed from the book by Sara J.Cuccurullo, MD. Published by Demos Medical Pub-lishing, New York. $119.00. Available online at: www.usbmis.com/store.

This program is a digitized version of the Physical Med-icine and Rehabilitation Board Review book. It allowsthe loading of a condensed version of the book onto apersonal digital assistants (PDA). It is available for Palmdevices as well as Windows-based devices.

Overall, this is a good program. It was easy to use andeasy to load onto a PDA. Once on the PDA, navigation isfairly easy. The information seems accurate, although thetext is hard to read on a Palm device. Palm devices witholder Palm operating system software may have problemsreading some of the files. The text can get distorted, andsome letters may be changed to symbols. There also seemto be problems running the program off a peripheral stor-age source such as a memory stick or memory card.

Although the program is largely useful, some sub-jects could use more elaboration. For instance, the or-thotics and prosthetic section would be better if a quickreference guide were added, including the K1–4 levelsand what they mean, how to measure leg-length discrep-ancy, how to perform the Thomas test, and how tomeasure the popliteal angle. Pictures of the differenttypes of sockets, knees, and feet would also be useful. Inthe EMG section, pictures of the different waveformswould be helpful. A chapter about interventional physia-try would also be beneficial. Wound care is an importanttopic that does not receive much coverage.

This program was evaluated by several residents. Allagreed that it would be helpful for board review. Thebook version was felt to be better, but the PDA versionwas considered good as a supplement to be used while onthe go. Their biggest complaint was that the programwas hard to read because of the size of the text. Severalresidents were interested in purchasing the program.

Rating: �� Lam Nguyen, DOSan Antonio, Texas

December 2007 Media Review 997

CME Self Assessment Exam QuestionsCME Article Number 5: Saveika, et al.

1. Baclofen is a ligand of which receptor?A. GABAA

B. GABAB

C. BicucullineD. Tetrodotoxin

2. �-aminobutyric acid B (GABAB) has been postulated toinhibit which neurotransmitter?A. SerotoninB. DopamineC. GutamateD. Aspartate

3. Which of the following is a symptom of intrathecal baclofenwithdrawal?A. PruritisB. BradycardiaC. HyporeflexiaD. Psychomotor retardation

4. Which of the following is most likely to interrupt delivery ofintrathecal baclofen?A. Pump malfunctionB. Pump migrationC. Systemic infectionD. Localized trauma

5. Treatment of intrathecal baclofen withdrawal should includeall of the following EXCEPT:A. BaclofenB. CyproheptadineC. BenzodiazepinesD. Clonidine

INSTRUCTIONS TO OBTAINCATEGORY 1 CME CREDITS:

1. Read the Designated CMEArticles in this issue.

2. Read the following CME Self-Assessment Exam Questions.

3. Photocopy and complete theCME Self-Assessment ExamAnswering Sheet and CMEEvaluation.

4. Send the completed AnsweringSheet and Evaluation to:Bradley R. Johns, ManagingEditor, CME Department-AAP,American Journal of PhysicalMedicine & Rehabilitation,7240 Fishback Hill Lane,Indianapolis, IN 46278

AMERICAN JOURNAL OF PHYSICAL MEDICINE& REHABILITATION

Vol. 86, No. 12 • December 2007

This is an adult learning experience and there is no requirement for obtaininga certain score. The objective is to have each participant learn from the totalexperience of studying the article, taking the exam, and being able to immedi-ately receive feedback with the correct answers. For complete information,please see “Instructions for Obtaining Continuing Medical Education Credit” atthe front of this issue.

Every question must be completed on the exam answering sheet to beeligible for CME credit. Leaving any item unanswered will make void theparticipant’s response. This CME activity must be completed and postmarked byDecember 31, 2008. The documentation received will be compiled throughoutthe calendar year, and once a year in January, participants will receive acertificate indicating CME credits earned for the prior year of work. This CMEactivity was planned and produced in accordance with the ACCME Essentials.


CME SELF-ASSESSMENT EXAM

AMERICAN JOURNAL OF PHYSICAL MEDICINE& REHABILITATION

Please photocopy this form and complete the informationrequired for each CME Activity.

Journal Issue Month and YearVolume Number Issue NumberCME Article NumberCME Article Author’s Name

Circle the appropriate answers.

1. A B C D

2. A B C D

3. A B C D

4. A B C D

5. A B C D

The answers to any essay questionsmust be typed or computer printedon a separate piece of paper and at-tached to this page.

After finishing this exam:

1. Check your answers with the cor-rect answers on page 999.

2. Complete the CME Evaluation andCertification on the following pageand mail to Bradley R. Johns, Man-aging Editor, CME Dept.-AAP, Amer-ican Journal of Physical Medicine &Rehabilitation, 7240 Fishback HillLane, Indianapolis, IN 46278.

3. This educational activity must becompleted and postmarked by De-cember 31, 2008. AAP Membersmay complete and submit thisCME Answering Sheet and the fol-lowing CME Evaluation and Certi-fication page online through themembers-only section of the AAPweb page at www.physiatry.org.

December 2007 CME Self-Assessment Exam 999

STANDARDIZED CME SELF-ASSESSMENT EXAM ANSWERING SHEET

Poor Satisfactory OutstandingWas the article consistent with the stated objectives? 1 2 3

Did reading this article prepare you to achieveits stated objectives?

1 2 3

Is reading this article likely to enhance yourprofessional effectiveness?

1 2 3

Was the article format conducive to learning? 1 2 3

Suggestions for future topics:?

I, certify that I have met the criteria for CMEcredit by studying the designated materials, by responding to the self-assessment questions, by reviewing thoseparts of the article dealing with any question(s) answered incorrectly, and by referring to the supplementalmaterials listed in the references.

This educational activity is designated for 11⁄2 category 1 CME credits.

Indicate total credits claimed: (maximum of 11⁄2 credits)

Signature of Participant Date

Are you a member of the AAP? � Yes � No

Do you have an individual subscription to the Journal? � Yes � No

If you are not an AAP member or a Journal subscriber, have you enclosed payment of$15 with your exam? � Yes � No

Payment Options: � Check (Payable to AAP) � VISA � Mastercard

Card # Exp. date Amount $

Signature

Please provide your name as it should appear on your certificate of credit and also provide your mailing address.

Name Position

Institution

Street Address

City, State, Zip

Phone Fax Email

Office Use

CME EVALUATION & CERTIFICATION

1000 CME Self-Assessment Exam Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

Authors:Vishwa S. Raj, MDDiana H. Rintala, PhD

Affiliations:From the Department of PhysicalMedicine & Rehabilitation, BaylorCollege of Medicine, Houston, Texas(VSR, DHR); The Institute forRehabilitation and Research,Houston, Texas (VSR); and Michael E.DeBakey Veterans Affairs MedicalCenter, Houston, Texas (DHR).

Correspondence:All correspondence and requests forreprints should be addressed toVishwa S. Raj, MD, 1100 BlytheBoulevard, Charlotte, NC 28203.


DOI: 10.1097/PHM.0b013e3181583f44

Perceived Preparedness for PhysiatricSpecialization and Future Career Goalsof Graduating Postgraduate Year IVResidents during the 2004–2005Academic YearABSTRACT

Raj VS, Rintala DH: Perceived preparedness for physiatric specialization andfuture career goals of graduating postgraduate year IV residents during the2004–2005 academic year. Am J Phys Med Rehabil 2007;86:1001–1006.

Objective: The purpose of this study was to evaluate trends amongpostgraduate year (PGY) IV physiatry residents, at the time of graduationfrom residency, in terms of their perceived experiences in the core clinicalareas, confidence with procedural subspecialization, choice in careerspecialization, and desire to pursue clinical fellowship.

Design: Surveys were distributed to 386 PGY IV residents in physiatryat the end of the 2004–2005 academic year.

Results: Ninety-three residents (24%) completed responses in a confi-dential manner. Residents who were generally more confident in core clinicalareas, as defined by the Self-Assessment Examination, and specialty prescrip-tion writing also believed themselves to be more prepared to practice thesetopics in their careers. Overall levels of confidence and perceived prepared-ness correlated positively with months of training and negatively with the beliefin the need for postresidency fellowship training to incorporate these areasinto clinical practice. Positive correlations also existed among perceived levelsof preparedness in performing various physiatric procedures. Statisticallysignificant differences in levels of confidence and preparedness existedamong geographic regions when evaluating core physiatric subject matter.Fifty-six percent of residents who responded planned to pursue fellowshiptraining, and a majority of residents intended to perform interventional pro-cedures and musculoskeletal medicine in their practices.

Conclusions: These results provide insight into how trainees perceivetheir current clinical education. With validation of measures for confidenceand preparedness, this survey may be useful as an adjunct resource forresidency programs to evaluate their trainees.

Key Words: PM&R Residency Training, Residency Education, Resident Career Goals,Regional Residency Training, PM&R Procedures, Resident Fellowship Goals

December 2007 Preparedness for Physiatry 1001

RESEARCH SURVEY

Residency Training

As the field of physical medicine and rehabilita-tion evolves, many residency programs have noteda trend for recent graduates to pursue fellowshiptraining and specialization.1 However, data evalu-ating physiatric residency education and trends inresident career choices is sparse. Though correla-tions have been shown between quartile ranking onthe Self-Assessment Examination (SAE) and pass-ing the American Board of Physical Medicine andRehabilitation (ABPMR) Part 1 Board CertificationExamination,2 it is unclear how residents perceivetheir current education in terms of learning theclinical core areas and preparing for future special-ization. During the 1987–1988 academic year, asurvey of chief residents from 70 board-accreditedphysical medicine and rehabilitation residency pro-grams found great variability in the total months ofresidency devoted to various clinical specialties inphysiatry.3 A 1993 survey found that trainees be-lieved they needed more clinical and didactic expe-rience in industrial medicine, sports medicine, andtherapeutic injections.4 During the past few years,however, residents and recent graduates haveshown interest in musculoskeletal medicine, soft-tissue disorders, therapeutic injections, and nerveblocks.5 This study was conducted to evaluatetrends among postgraduate year (PGY) IV physiatryresidents, at the time of graduation from residency,in terms of their perceived experiences in the coreclinical areas, confidence with procedural subspe-cialization, choice in career specialization, and de-sire to pursue clinical fellowship.

METHODOLOGYPhysiatric residency programs accredited by the

Accreditation Council for Graduate Medical Educa-tion (ACGME) were identified using the ACGME Webpage (www.acgme.org) in March 2005. On the basis ofinformation provided, each program coordinator wassent a package in May 2005 and was asked to distrib-ute the contents among graduating PGY IV residentsin their respective program. Enclosed for each resi-dent were an introductory letter, a questionnaire, anda postage-paid, preaddressed unmarked envelope toprotect confidentiality. Residents were asked to re-turn the surveys by mail within 4 wks of receiving them.

The questionnaire was designed to addressclinical areas as outlined by the SAE and estab-lished by the ABPMR for the 2004–2005 academicyear (Table 1). In the first section, residents wereasked to rate their levels of confidence in evaluat-ing physiatric diagnoses, and their perceived levelsof preparedness to enter a career in specific areas.The response options were 1 � very unsure/veryunprepared, 2 � somewhat unsure/somewhat un-prepared, 3 � neutral, 4 � somewhat confident/somewhat prepared, and 5 � very confident/very

prepared. They were asked about the need for fel-lowship training to continue a career path afterresidency (1 � yes, 0 � no), and the numbermonths of training devoted to specific disciplinesduring their residency. The second section ad-dressed perceived levels of preparedness for per-forming specialized procedures (Table 1). The thirdsection consisted of open-ended questions to deter-mine plans for fellowship training and career goals.Finally, residents rated their confidence levels inrecommending workplace modifications, prescrib-ing prosthetic or orthotic devices, and prescribingphysiatric therapeutics. Though residents did notprovide identifying markers to protect confidenti-ality, they were asked to provide the name of theircurrent residency training program to help catego-rize data according to geographic regions (EastCoast, Midwest, South, and West Coast). Using theStatistical Package for the Social Sciences (SPSS)computer software, the data were analyzed for (a)general descriptive information, (b) relationshipsamong various factors, and (c) differences amonggeographic region.

RESULTSOf 386 surveys distributed, 93 responses were

returned, yielding an overall response rate of 24%,

TABLE 1 Clinical areas for evaluation asdefined by the Self AssessmentExamination (2005) and specializedphysiatric procedures

Brain disorders (including stroke and traumaticbrain injury)

Electrodiagnosis (including electromyography andnerve-conduction studies)

Industrial rehabilitation (including occupationaland performing arts medicine)

Joint and connective tissue disorder (includingrheumatological disorders)

Medical rehabilitation (including cancer, cardiac,general, and pulmonary rehabilitation)

Musculoskeletal medicine (not includinginterventional physiatry)

Neuromuscular disorders (including motor neurondisease, movement disorders, and multiplesclerosis)

Pediatric rehabilitationPhysiatric therapeutics (including modalities and

exercise prescription)Prosthetics and orthoticsSpinal cord medicineAcupunctureBotulinum toxin and/or phenol injectionsInterventional physiatric procedures (including

epidural injections, facet blocks, and medialbranch blocks)

Osteopathic manual therapyPeripheral joint aspiration and/or injections

1002 Raj and Rintala Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

with ranges of 14%–34% per region (Table 2).Ninety-one residents responded that they had spenttheir entire residency within the same program,one resident responded that he or she had not, andone resident provided no response. Overall, therewere significant correlations between: prepared-ness and confidence (r � 0.983; P � 0.001); con-fidence and months of training within specific clin-ical areas (r � 0.844; P � 0.001); preparedness andmonths of training (r � 0.849; P � 0.001); andpreparedness and need for fellowship training (r ��0.655; P � 0.03).

Respondents were most confident and believedthemselves to be most prepared in the areas ofmusculoskeletal medicine and electrodiagnosis(Table 3). They were least confident and preparedin pediatric rehabilitation. Correlations rangingfrom 0.726 to 0.936 (P � 0.001) existed betweenlevels of confidence and preparedness in all corerehabilitation specialties. Negative associationswith belief in a need for fellowship training werenoted for both levels of confidence (r � �0.216 to�0.553; P � 0.04) and levels of preparedness (r ��0.209 to �0.518; P � 0.05) in specific areas,except that results for medical rehabilitation werenot statistically significant. For most core rehabil-itation specialties, significant relationships existedbetween levels of confidence and total months oftraining (r � 0.268–0.471; P � 0.02); however,

results for brain disorders, electrodiagnostics, andindustrial rehabilitation were not statistically sig-nificant. Significant relationship also existed be-tween levels of preparedness and total months oftraining (r � 0.246–0.519; P � 0.01), except thatresults for electrodiagnostics were not significant.

With regard to writing physiatric therapeuticprescriptions, creating prosthetic and orthotic (P&O)prescriptions, and recommending workplace modifi-cations, there were associations between levels ofconfidence in the core clinical areas (r � 0.547–0.839; P � 0.001) and levels of preparedness forspecialty prescription writing in that clinical area(r � 0.514–0.915; P � 0.001; Table 4). Months oftraining correlated positively with confidence in writ-ing P&O prescriptions and prescriptions for physiat-ric therapeutics (r � 0.428 and 0.463, respectively;P � 0.001). Confidence in writing P&O prescriptionsand recommending workplace modifications werenegatively related to belief in the need for fellowshiptraining in the respective core specialty (r � �0.388and �0.291, respectively; P � 0.01).

Residents who believed they were better pre-pared in performing acupuncture also believedthey were better prepared in performing interven-tional physiatric procedures (r � 0.277; P � 0.007)and osteopathic manual manipulation (r � 0.301;P � 0.003). Resident levels of perceived prepared-ness in both interventional physiatry and periph-

TABLE 2 Response rate by geographic region

RegionContributionto Study, %

Numberof Responses

TotalPossible Responses

ResponseRate, %

East 41.9 39 170 23South 28.0 26 76 34Midwest 14.0 13 95 14West 16.1 15 45 33Totals 100 93 386 24

TABLE 3 Mean responses regarding core clinical areas

Clinical Area Confidence PreparednessBelief for Need

for Fellowship, %Months Overall

of Training

Brain disorders 3.98 3.72 32 4.61Electrodiagnostics 4.28 4.18 14 6.16Industrial rehabilitation 3.45 3.35 24 2.39Joint and connective tissue disorders 3.69 3.52 23 3.40Medical rehabilitation 3.83 3.79 11 5.42Musculoskeletal 4.41 4.23 23 7.22Neuromuscular disorders 3.51 3.35 39 3.96Pediatric rehabilitation 3.19 3.02 73 2.77Physiatric therapeutics 4.14 4.11 3 5.16Prosthetics/orthotics 3.67 3.59 12 3.32Spinal cord injury 4.16 3.96 38 3.91


eral joint aspiration/injections were significantlycorrelated with preparedness in performing phenoland/or botulinum toxin injections (r � 0.256 and0.359, respectively; P � 0.02). Finally, a relation-ship was noted between levels of preparedness forinterventional physiatry and peripheral joint aspi-ration/injections (r � 0.375; P � 0.001).

Analyses of variances revealed statistically sig-nificant geographic differences in confidence levelsand preparedness levels for certain subject matter.Specifically, East Coast residents were more confi-dent than West Coast residents in joint and con-nective tissue disorder (means: East Coast: 3.92;West Coast: 3.13; P � 0.004). Southern residentswere more confident than West Coast residents inprescribing P&O (means: South: 3.92; West Coast:3.07; P � 0.01), and Southern residents were moreconfident than East Coast residents in spinal cordinjury medicine (means: South: 4.60; East Coast:3.76; P � 0.001). Residents in both the South andEast Coast believed they were more prepared inprescribing P&O than did West Coast residents(means: South: 3.96; East Coast: 3.62; West Coast:2.87; P � 0.03). Southern residents believed theywere more prepared than did East Coast residents(means: South: 4.36; East Coast: 3.59; P � 0.006)in spinal cord injury medicine.

When asked to note future career paths, mostresidents indicated they planned to pursue inter-ventional procedures and musculoskeletal medi-cine (Table 5). Fifty-two respondents (56%) indi-cated that they planned to pursue a fellowship aftergraduation from residency. Of the residents sur-veyed, 95.3% of those pursuing careers in interven-tional procedures practice (significance: two sided,�0.001, with forward exact test) planned to pursuefellowships to continue their training.

DISCUSSIONThe results of this study indicate that resident

confidence and preparedness generally correlated

with total months of training in core physiatric dis-ciplines and specialty prescription writing. Traineeswho were more confident also believed themselves tobe better prepared to enter careers in those clinicalareas, and those individuals who believed they werebetter prepared generally felt that postresidency fel-lowship training was not necessary for professionalclinical work.

Surveys of confidence and preparedness havebeen used by various medical specialties to evaluatephysician competence and clinical experience. Alongitudinal cohort study conducted in 2000 mea-sured overall preparedness of pediatric residents incore clinical areas, and evaluated these same indi-viduals after 5 yrs of clinical practice.6 In general,it was found that the residents’ perceptions of pre-paredness at the time of graduation were similar toretrospective assessments 5 yrs later. With respectto subspecialties in which there were differences inperceived preparedness, practicing physicians feltthey may have been more prepared than they ini-tially thought they were at the time of residencycompletion. Similarly, a 2005 survey of recentlygraduated obstetric–gynecology residents foundthat overall levels of confidence in managing clin-ical problems remained constant after graduationin core areas, although it was noted that periodicreview of graduates’ perception of training could beused to adjust residency curriculum relative tofuture clinical needs.7 Self-assessed measures ofconfidence in outpatient and inpatient competen-cies, as well as procedural ability, have already been

TABLE 4 Mean perceived preparedness forspecialty prescriptionsand procedures

Specialty AreasAverage

Preparedness

Physiatric therapeutic prescriptions 4.04Peripheral joint injections 4.00Botox/phenol injections 4.00Prosthetic/orthotics prescriptions 3.52Workplace modification prescriptions 3.28Epidural injections and nerve blocks 2.70OMT 2.19Acupuncture 1.61

TABLE 5 Career paths and interests ofgraduates in 2004–2005

Career Paths andInterestsGraduates Plannedto Pursue

Number ofResponses

Percentage ofRespondents

Interventionalprocedurespractice

43 46

Musculoskeletalpractice

32 34

General practice 17 18Outpatient practice 9 9.7Sports practice 9 9.7Inpatient practice 6 6.5Spinal cord injury

practice6 6.5

Electromyographypractice

4 4.3

Traumatic braininjury/strokepractice

2 2.2

Private practice 1 1.1

The sum of the results is greater than 100% becauserespondents had multiple responses.


implemented to evaluate specific areas during fam-ily practice residency,8 and have even been used asan internal gauge for these residencies to evaluatetheir overall effectiveness in teaching ACGME corecompetencies to their trainees.9 Finally, thoughresident perception of comfort is multifactorial, a2006 survey of internal medicine residents foundthat self-perceived levels of comfort in performingclinical procedures was associated with the numberof procedures performed.10

Despite well-documented criteria required bythe ACGME for clinical education of residents, ithas been noted that some difficulty exists in assess-ing and measuring overall resident competence,and identifying appropriate methods of evaluationfor residency programs.11 The most recent pro-gram requirements for residency education estab-lished by the residency review committee of theACGME state that residents should have the follow-ing clinical experiences before graduation: at least12 mos of direct responsibility for the completemanagement of inpatients; at least 12 mos in thecare of outpatients, with significant emphasis onpatients with musculoskeletal disorders; and approx-imately 200 electromyographic consultations.12 Be-yond these requirements, residency programs have agreat deal of the flexibility to design their curriculawhen incorporating the clinical competencies re-quired for a broadly based and complete physiatricclinical exposure. This flexibility may account for thegreat variability between residency programs in de-signing their clinical and didactic schedules.

In addition to the SAE, scheduled resident andprogram director meetings, and annual program in-ternal reviews, the measures used in this study mayprovide a supplemental resource to help monitor theprogress of trainees. With further studies, correla-tions may be shown between resident self perceptionand success on board certification or subspecialtycertification examinations. Quantitative analysiscould also be conducted to determine the appro-priate number of patient encounters or months ofclinical exposure, as they relate to reaching appro-priate confidence and preparedness levels for pro-fessional success. As an adjunct method of evalua-tion, these data may assist residency programswhen designing clinical and didactic componentsto focus on the core rehabilitation disciplines andnewly evolving clinical and procedure-based expe-riences that trainees will need for future subspe-cialization.

Similar analyses could be used to evaluate theprocedure-based curriculum in residency pro-grams. This study indicates that there are positivecorrelations between perceived preparedness levelsamong different physiatric procedures. It is unclearwhether perceived preparedness in one proceduralarea tends to generalize to perceived preparedness

in other procedural areas or whether individualswith an interest in procedural based medicine seekout these procedures. However, according to thedata collected relevant to career paths, procedure-based practice will play an important role in theclinical work of future physiatrists. Further analy-sis of resident exposure to these procedures mayhelp determine optimal exposure during residencyto provide clinical competence after graduation.Furthermore, such information may provide fur-ther insight as to what types of procedures willyield the most effective results in preparing a res-ident for clinical practice. This seems to be ofparticular importance given the percentage of res-idents from this study that plan to incorporateprocedures in their future work.

Residency education and experiences may sig-nificantly differ according to the prevalence of cer-tain clinical diagnoses within various geographicareas. If a residency program’s clinical experienceis biased secondary to its geography, it may nothave appropriate resources to provide exposure inother facets of physiatry. This, in turn, could accountfor variable levels of confidence and preparednessbetween regions. With further geographic informa-tion, residency programs may be able to actively ad-just clinical experiences to provide a well-roundedphysiatric education.

The questions posed in this survey yield inter-esting results regarding resident self perception ofclinical skills, time devoted to clinical areas duringresidency education, and career and fellowshipplans for graduating residents. However beforegeneralizing the results of this study, potentialsources of error must be examined. A major threatto the external validity of this study was the rela-tively low response rate of 24%. Because of thepotential for response bias, caution must be usedwhen interpreting the data. More than 40% of therespondents to this survey stated they planned topursue interventional physiatry practice after grad-uation, and more than 30% planned to pursuemusculoskeletal medicine. It is possible that resi-dents with these career plans were more likely torespond to the survey. There is also potential biasfor the results of this study to be skewed accordingto the viewpoints and experiences of individualspursuing interventional, as well as musculoskele-tal, based practice. Similar considerations for biasmust be made when evaluating the geographic datain this study, because a majority of respondentshailed from the East Coast and South. Thoughproducing graduates with increased beliefs of com-petence may be a desired outcome of residencyeducation,13 confidence and perceived prepared-ness in this study have not been validated withmore objective measures, such as the SAE orABPMR Part I Certification Examination. In addi-


tion, months of training are traditionally used byresidency programs to determine adequate residentexposure to clinical areas of training, but it is notclear whether months of training, vs. specific patientencounters, is the appropriate method to quantita-tively evaluate clinical experience. Finally, the use ofopen-ended questions to describe fellowship plansand career choices yielded a diverse array of re-sponses. A possible source of error was the subse-quent interpretation and assignment of these re-sponses into more traditional categorizations.

To counter the issue of low response rate, arepeat study should be conducted to further in-crease both regional and national response rates.With the assistance and cooperation of residencyprogram directors, graduating residents couldcomplete the survey either in paper form or withan Internet-based questionnaire. The survey couldbe administered at the time of the resident exitinterviews for graduation to ensure that the datacollected reflect the most comprehensive evalua-tion of experiences. With confidential coding ofindividual residents completing the surveys, corre-lations could then be established between theirsubjective answers of perceived confidence and pre-paredness with objective scores on either the SAEor ABPMR Part I Certification Examination. Fur-thermore, the use of a more detailed survey withclosed-ended questions could eliminate the ambi-guity of open-ended responses regarding futurecareer plans.

In conclusion, this survey presents interestingdata pertinent to the evaluation of resident experi-ences and career choices. Future studies are neededwith larger sample sizes to generalize results to theoverall graduating resident population, specificallywith respect to overall levels of confidence and pre-paredness in various physiatric specialties, geo-graphic differences between resident perceptions, andfuture career plans. With validation of confidence and

preparedness measures, similar studies may provideresidency programs an additional resource to evalu-ate their trainees.

REFERENCES1. DeLisa JA, Jain SS, Yablon SA: Resident interest in physical

medicine and rehabilitation fellowship. Results of a survey.Am J Phys Med Rehabil 1991;70:290–3

2. Fish DE, Radfar-Baublitz L, Choi H, Felsenthel G: Correla-tion of standardized testing results with success on the2001 American Board of Physical Medicine and Rehabilita-tion part 1 board certification examination. Am J Phys MedRehabil 2003;82:686–91

3. Anderson JM, Felsenthel G: Residency training in physicalmedicine and rehabilitation I: clinical and didactic experi-ence. Arch Phys Med Rehabil 1990;71:372–5

4. Kerrigan DC, Janes WW, Martin WA, Roe TJ: Physical med-icine and rehabilitation residents’ educational needs assess-ment. Arch Phys Med Rehabil 1993;74:687–90

5. Hart KA, Kevorkian G, Rintala DH: Continuing medicaleducation: interests of former and current residents of aphysical medicine and rehabilitation residency program.Am J Phys Med Rehabil 1999;78:561–70

6. Roberts KB, Starr TG, DeWitt S: Resident preparedness forpractice: a longitudinal cohort study. Ambul Pediatr 2002;2:132–5

7. Schneider KM, Monga M: Former-resident self-assessment:a tool for residency program curriculum evaluation in ob-stetrics and gynecology. J Reprod Med 2005;50:821–3

8. Edwards F, Frey K: The future of residency education:implementing a competency-based educational model. FamMed 2007;39:116–25

9. Frey K, Edwards F, Altman K, Spahr N, Gorman RS: The‘collaborative care’ curriculum: an educational model ad-dressing key ACGME core competencies in primary careresidency training. Med Educ 2003;37:786–9

10. Huang GC, Smith CC, Gordon CE, et al: Beyond the com-fort zone: residents assess their comfort performing inpa-tient medicine procedures. Am J Med 2006;119:e17–24

11. Hassett JM, Zinnerstrom K, Nawotniak RH, SchimpfhauserF, Dayton MT: Utilization of standardized patients to eval-uate clinical and interpersonal skills of surgical residents.Surgery 2006;140:633–8

12. Accreditation Council for Graduate Medical Education Website. Available at: http://www.acgme.org. Accessed August23, 2007

13. Butler DJ, Holloway RL, Gottlieb M: Predicting residentconfident to lead family meetings. Fam Med 1998;30:356–61


Authors:Katherine L. Salter, BARobert W. Teasell, MDNorine C. Foley, MScJeffrey W. Jutai, PhD

Affiliations:Department of Physical Medicine andRehabilitation, St. Joseph’sHealthcare London, London, Canada(KLS, RWT, NCF, JWJ); SchulichSchool of Medicine, University ofWestern Ontario, London, Canada(RWT, JWJ); and Lawson HealthResearch Institute, London, Canada(RWT, JWJ).

Correspondence:All correspondence and requests forreprints should be addressed toKatherine Salter, c/o Dr. R. Teasell,Dept. of Physical Medicine andRehabilitation, St. Joseph’sHealthcare London, Parkwood Site,801 Commissioner’s Road E.,London, Ontario, Canada, N6C 5J1.


DOI: 10.1097/PHM.0b013e3181587b3d

Outcome Assessment inRandomized Controlled Trials ofStroke Rehabilitation

ABSTRACT

Salter KL, Teasell RW, Foley NC, Jutai JW: Outcome assessment in randomizedcontrolled trials of stroke rehabilitation. Am J Phys Med Rehabil 2007;86:1007–1012.

The lack of a unified approach to outcome assessment in stroke rehabilitationlimits our ability to interpret evidence provided by randomized controlled trials(RCTs). The purpose of this review was to identify outcomes and assessmenttools reported in RCTs of stroke rehabilitation interventions as a first step towardconsistent assessment of outcomes. Given that the validity of research is linkedto reliability and validity of measurement, the relationship between the use ofpreviously developed outcome measures and the methodological quality of RCTswas explored.

Electronic literature searches identified RCTs examining stroke rehabilitationtherapies from 1968 to 2005. The Physiotherapy Evidence Database (PEDro)scale was used to assess methodological quality. Cited outcomes were recordedand assessment tools identified as previously published or study specific. Fourhundred ninety-one RCTs cited the assessment of 1447 outcomes using 489measurement tools. Two hundred fifty-four of these were previously published,and 235 were study specific. A core of 30 frequently cited tools was identified.The use of previously published assessment tools to evaluate primary studyoutcomes was associated with higher PEDro scores.

Significant heterogeneity in outcome assessment was demonstrated, although acore of 30 frequently cited tools could be identified. Appropriate evaluation andselection of outcome measures would enhance the methodological quality ofrandomized controlled trials.

Key Words: Outcome Assessment, Evidence-Based Medicine, Reliability, Validity,Stroke Rehabilitation, Measurement, Randomized Controlled Trials

December 2007 Outcomes Assessment in RCTs in Stroke 1007

REVIEW & ANALYSIS

Outcomes

Stroke has a broad impact on individuals, fam-ilies, and the communities in which they live. Nu-merous assessment tools have been created to as-sess the complex and heterogeneous consequencesof stroke and the effect of stroke rehabilitationinterventions on stroke outcomes. However, thereis no consensus regarding what and how outcomesshould be measured to create a comprehensivedescription of individual functioning and healthand to facilitate our ability to make comparisonsacross interventions and studies.1,2 The currentlack of a unified approach to assessment of out-come limits our ability to interpret the evidenceprovided by randomized controlled trials (RCTs) ina clinically meaningful way.3–5

Our ability to interpret study results is also lim-ited by the validity of the research itself, which, inturn, is limited by the reliability of outcome assess-ment.6 Ideally, researchers should use, whenever pos-sible, measurement tools that have been previouslyevaluated for use in similar subject populations andhave demonstrated acceptable reliability, validity andresponsiveness.6,7 The revised CONSORT guidelines8

recommend the use of previously developed and val-idated tools to improve quality of measurement andfacilitate between study comparisons.

As a first step toward developing consensusaround consistent assessment of stroke rehabilitationoutcomes, an examination of the outcome assess-ment tools that have been reported in the strokerehabilitation literature was conducted. Given thatthe interpretability and utility of research results isassociated with the validity of outcome measurement,the association between the use of previously devel-oped outcome measures and the methodologicalquality of RCTs was examined.

METHODSLiterature Search

The present study arose from the Evidence-Based Review of Stroke Rehabilitation.9 Search strat-egies included the following databases: MEDLINE,CINAHL, EMBASE, MANTIS, PASCAL, Psych-Infoand Sci Search from 1968 to 2005. The literaturesearch, study selection, and data abstraction strate-gies employed for that review process have been de-scribed in detail elsewhere.9,10 For the present study,randomized controlled trials (RCTs) examining pri-mary or secondary stroke prevention treatments wereexcluded, as were review articles, meta-analyses, orsystematic reviews. RCTs were included if they citedthe use of a measurement tool or scale to assess atleast one of the study outcomes. Studies assessingonly discrete, counted outcomes (e.g., mortality,length of stay, occurrence of a stroke event, etc.) wereexcluded. On the basis of these criteria, 491 RCTs

evaluating the effectiveness of interventions in strokerehabilitation were selected for inclusion.

Data Abstraction and QualityAssessment

Two abstractors, each blinded to the other’sresults, reviewed each article independently. Thereview process consisted of two parts. In the firstpart, data pertinent to the study were collected,including study methodology, cited outcomes, andmeans by which study outcomes were assessed.After this, the study was assigned a quality-ratingscore derived using the Physiotherapy EvidenceDatabase (PEDro) scale.

The PEDro scale, developed by the Centre forEvidence-Based Physiotherapy (CEBP) in Australia(www.pedro.fhs.uwyd.edu.au), was created to as-sess bibliographic details and abstracts of RCTs,quasi-randomized studies, and systematic reviewsin physiotherapy.11 Based on a set of core criteriafor quality assessment in randomized controlledtrials derived by expert consensus,11,12 the scaleconsists of ten items (Table 1). Each item receivesa rating of either yes (1) or no (0). Item ratings are

TABLE 1 The Physiotherapy EvidenceDatabase (PEDro) scale

1. Subjects were randomly allocated togroups (in a crossover study, subjectswere randomly allocated an order inwhich treatments were received)

2. Allocation was concealed3. Groups were similar at baseline

regarding the most importantprognostic indicators

4. There was blinding of all subjects5. There was blinding of all therapists who

administered the therapy6. There was blinding of all assessors who

measured at least one key outcome7. Adequacy of follow-up (could all of the

subjects originally randomized beaccounted for at the end of thestudy?). This item was modified fromthe original item in the PEDro scale,which stipulated measurement of themain study outcome in at least 85%of subjects)

8. Intention-to-treat analysis9. Results of between-group statistical

comparisons were reported for at leastone outcome

10. Study provides both point measures andmeasures of variability for at least onekey outcome

Eligibility criteria were included in the original PEDroscale. Because subject selection pertains more directly toexternal rather than internal study validity, it was not in-cluded in the present review.9,10

1008 Salter et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

summed to provide a total score. Unlike checklistssuch as the Downs and Black,13 the PEDro scaledoes not include evaluation of study outcomes and,therefore, provides an evaluation of methodologi-cal quality unbiased by the selection of outcomemeasures. Four of the PEDro items have beenempirically validated (randomization, concealed al-location, blinding, and adequacy of follow-up), andthe remaining items have face validity.11 Reportedintraclass correlation coefficients for interrater re-liability for the PEDro scale range from 0.5614 to0.91.15 A moderate correlation (r � 0.59) betweenPEDro scores and ratings of methodological qualityusing the Jadad scale has also been reported.16 Twoindependent raters, blinded to each other, assigneda PEDro score to each study. A third reviewer re-solved any scoring disagreements. A detailed descrip-tion of this application of the PEDro scale has beenpublished previously.9,10

For each study included in the review, all citedoutcomes were recorded. Primary outcomes wereidentified. For our purposes, an outcome was con-sidered to be primary if it was explicitly named asthe primary outcome by the study author(s) or if itwas listed first in the study methodology. Eachmeasurement tool cited was categorized as previ-ously published or study specific. Tools were clas-sified as previously published if prior publication ofthe tool along with published, evaluation(s) of re-liability or validity or responsiveness could be iden-tified from the information provided by the studyauthors. Study-specific tools were those developedfor the project at hand; they had no previouslypublished evidence of reliability, validity, and/orresponsiveness cited by the study authors.

Mean (�SD) and median (interquartile range)PEDro scores were calculated for studies citingpreviously published outcome measures and alsofor studies citing the use of study-specific tools. Inaddition, mean and median PEDro scores werecalculated for studies using previously published orstudy-specific tools to assess the primary studyoutcome. Given that the PEDro scale yields non-parametric data, median scores for studies usingpreviously published measures were comparedwith those citing the use of study-specific measuresusing the Mann–Whitney U test (a nonparametricalternative to the t test). An alpha level of �0.05was considered to be statistically significant. Allstatistical calculations were performed using SPSSfor Windows (version 12.0).

RESULTSOne thousand seven hundred twenty-one out-

come citations were recorded from 491 identifiedRCTs. Of these, 274 represented assessments ofclinical variables, which were described but madeno reference to the use of a specific measurement

scale. The remaining 1447 cited study outcomeswere assessed using 489 assessment tools. Of these,254 had been previously published, and 235(48.1%) were study specific.

Of the 254 previously published scales, only 30were cited more than ten times each. These 30 mostfrequently cited measurement tools accounted forapproximately 50.1% (725) of all outcome citations. Alist of the most frequently cited assessment instru-ments is provided in Table 2.

One hundred fifty-two studies cited the use ofone or more study-specific measures to assess out-comes. Of these, 66 cited the use of a study-specificmeasure to assess the primary study outcome.Mean (�SD) and median PEDro scores as well asmean ranks are provided in Table 3.

Although the median PEDro score in thegroup of studies including study-specific assess-ment tools is the same as the studies using only

TABLE 2 Frequently cited outcome measures

Most Frequently Cited Outcomes1968–2005

No. ofCitations

Barthel index22 123Timed walk assessments (varying

times and distances)23,2472

Fugl–Meyer assessment of strokerecovery25

58

Modified ashworth scale26,27 43Functional independence measure28 38Mini-mental state examination29 27Action research arm test30 25Frenchay activities inventory31 22Nottingham health profile32 22Medical outcomes study short

form-3633,3422

Nottingham extended activities ofdaily living35

21

Motor assessment scale36 21Hospital anxiety and depression

scale3717

Berg balance scale38,39 16Rankin handicap/modified Rankin/

Oxford handicap4016

Nine-hole peg test41 15Hamilton rating scale for

depression4215

Motricity index43 14General health questionnaire-2844 14Rivermead mobility inventory45 13Rivermead motor assessment46 13VAS–pain47,48 12Porch index of communicative

ability4912

Motor activity log50 12Brunnstrom scale51 11Caregiver strain index52 11Timed up and go53 10Functional ambulation categories54 10Scandinavian stroke scale55 10Weschler memory subtests56,57 10


previously published measures, both the meanPEDro scores and mean rank reflect a significantdifference in distribution toward lower PEDroscores in the group using study-specific measures(z � �2.51, P � 0.012). Among the subset ofstudies citing the use of study-specific measures,those studies using a previously published measureto evaluate the primary study outcome obtained ahigher median and mean PEDro score than did thegroup of studies using a study-specific measure forthis purpose. Use of a previously published mea-sure to assess primary study outcomes was signif-icantly associated with higher PEDro scores (z ��2.6, P � 0.009).

DISCUSSIONAs in previous examinations of RCTs in stroke

research, considerable heterogeneity of measure-ment was noted.18,19 Four hundred ninety-oneRCTs cited the use of 489 measurement tools in theassessment of 1447 study outcomes. This variabil-ity and inconsistency in the selection and assess-ment of outcomes limits our ability to interpretindividual study data and make appropriate com-parisons between trials of similar interventions.5,19

However, a core of 30 frequently cited, previouslypublished assessment tools was identified. Thesetools were used in the assessment of a wide rangeof patient outcomes along the continuum frombody structure and function assessments to health-related quality of life and could, together withrigorous examination of their psychometric prop-erties, provide a starting point around which aconsensus for outcome assessment in stroke reha-bilitation trials could begin.

Despite the availability of a wide selection ofpreviously published measures, almost one half(48%) of all measurement tools identified in thepresent review were study specific. Marshall et al.20

have identified the use of unpublished rating scalesas a major source of bias in randomized controlledtrials of nonpharmacological treatments of schizo-phrenia, such that trials using unpublished scaleswere significantly more likely to report significant

effects in favor of treatment when compared withcontrol conditions. The authors suggest severalpossible explanations for why unpublished scalescould be a source of bias: (i) unpublished scalestend to be used by small trials of poorer quality, (ii)selective reporting of data obtained from unpub-lished scales, eliminating nonsignificant results, or(iii) ad hoc adjustment to the content of unpub-lished scale to present the most favorable data.Although this bias has not been demonstrated inother branches of medicine, it is recommended thatinvestigators refrain from using unpublished assess-ment tools.20 Indeed, it is generally accepted thatreliable and valid measurement is essential in theproduction of meaningful and replicable data, and, tothat end, researchers should endeavor to use mea-sures that are known to be reliable and valid.6,7

The present study compared the methodolog-ical quality of trials citing the use of previouslypublished measures with those using study-specifictools. Although use of a previously published mea-sure, particularly in the assessment of the primarystudy outcome, was associated with superior meth-odological quality, the difference between groupswas approximately equivalent to a single point onthe PEDro scale. Total PEDro scores provide anindication of which trials are more likely to be validand contain sufficient statistical information to en-hance interpretability.11 A difference of a singlepoint may represent an important deficiency inmethodology, particularly if the item were consid-ered to be a significant source of bias, such asconcealed allocation, blinding of patients or ther-apists, or intention-to-treat analysis.

The apparent association between the use ofpreviously published measurement tools and themethodological quality of RCTs signals the impor-tance of outcome assessment to research. The se-lection of outcome measures may be as importanta component to the overall quality of an RCT as aremethods for blinding, concealed allocation, ade-quacy of follow-up, and other components of re-search design. The selection of appropriate, reli-able, and valid measurement tools should be more

TABLE 3 Mean and median Physiotherapy Evidence Database (PEDro) scores for studies using previouslypublished vs. study-specific measures

Mean SD Median IQRMeanRank

All studies (n � 491)Cited only previously published measures (n � 339) 6.04 1.4 6.0 2 256.5Included study-specific measures (n � 152) 5.70 1.5 6.0 2 222.5

Studies including study-specific measures (n � 152)Used previously published measure to assess primary outcome (n � 86) 5.97 1.4 6.0 2 84.47Used study-specific measure to assess primary outcome (n � 66) 5.35 1.6 5.0 2 66.12


formally recognized and consistently reported in astandardized manner, as recommended by Dijkerset al.6 and Moher et al.21

Limitations of the Present StudyIn this study, previously published outcomes

were recorded separately from and compared withstudy-specific measures. Previously published mea-sures were defined, for the purposes of the presentpaper, as those that had been published previouslyand had published evidence available with regard toreliability, validity or responsiveness. However, nospecific claims can be made regarding the mea-surement properties of an assessment tool basedsolely on prior publication within the researchliterature. Indeed, confirming whether these pre-vious publications provide evidence of adequatereliability, validity, or responsiveness, or whetherprevious evaluations of the psychometric or clini-metric properties of a given assessment tool hadbeen conducted within a population of individualswho had experienced stroke, was beyond the scopeof the present review. Similarly, with regard to thetools designated as study specific, the present re-view neither reported whether reliability/validitydata were presented nor evaluated whether anysuch data were appropriate or acceptable.

CONCLUSIONSOur understanding of the effectiveness of a

given rehabilitation intervention is, in part, deter-mined by our ability to measure the impact of a givenintervention on an individual’s functional ability andhealth. However, there is no consensus regardingwhat and how outcomes should be measured. Thisstudy identified significant heterogeneity of outcomeassessment in a review of 491 randomized controlledtrials of stroke rehabilitation interventions. Althoughthere were literally hundreds of measures used in theassessment of outcomes, a core of 30 well-knownmeasures accounted for more than one half of themeasurement citations found in RCTs examining theefficacy of stroke rehabilitation interventions.

Standardized evaluation of outcome measures,and the process by which measures are selected foruse, would undoubtedly enhance the methodolog-ical quality of RCTs in stroke rehabilitation re-search. It would also improve the confidence withwhich research results can be interpreted, repro-duced, and applied.

ACKNOWLEDGMENTSFunding was provided in part by the Heart and

Stroke Foundation of Ontario, the Ontario Ministry ofHealth and Long-term Care, Canadian Stroke Net-work, and the Toronto Rehabilitation Institute. Drs.Jutai and Teasell are members of the Canadian StrokeNetwork (http://www.canadianstrokenetwork.ca). Dr.

Jutai is also a member of the Consortium for AssistiveTechnology Outcomes Research (CATOR) (http://www.atoutcomes.com).

REFERENCES1. Geyh S, Kurt T, Brockow T, et al: Identifying the concepts

contained in outcome measures of clinical trials on strokeusing the international classification of functioning, dis-ability and health as a reference. J Rehabil Med 2004;44(suppl):56–62

2. Geyh S, Cieza A, Schouten J, et al: ICF core sets for stroke.J Rehabil Med 2004;44(suppl):135–41

3. Jutai J, Teasell R: The necessity and limitations of evidence-based practice in stroke rehabilitation. Top Stroke Rehabil2003;10:71–8

4. Stucki G, Ewert T, Cieza A: Value and application of the ICFin rehabilitation medicine. Disabil Rehabil 2003;25:628–34

5. Haigh R, Tennant A, Biering-Sorensen, Grimby G, et al: Theuse of outcome measures in physical medicine and rehabil-itation within Europe. J Rehabil Med 2001;33:278

6. Dijkers M, Kropp GC, Esper RM, Yavuzer G, Cullen N,Bakdalieh Y: Reporting on reliability and validity of out-come measures in medical rehabilitation research. DisabilRehabil 2002;24:819–27

7. Baker GA, Hesdon B, Marson AG: Quality-of-life and behav-ioral outcome measures in randomized controlled trials ofantiepileptic drugs: a systematic review of methodology andreporting standards. Epilepsia 2000;41:1357–63

8. The CONSORT Group. CONSORT guidelines: revised state-ment. Available at: www.consort-statement.org/statement/revisedstatement.htm#checklist. Accessed August 23, 2007.

9. Teasell R, Foley N, Salter K, Bhogal SK, Jutai J, SpeechleyM: Evidence-Based Review of Stroke Rehabilitation. Lon-don, Canada, 2006. Available at: www.ebrsr.com

10. Foley NC, Teasell RW, Bhogal SK, Speechley M: Strokerehabilitation evidence-based review: methodology. TopStroke Rehabil 2003;10:1–7

11. Moseley AM, Herbert RD, Sherrington C, Maher CG: Evi-dence for physiotherapy practice: a survey of the Physio-therapy Evidence Database (PEDro). Aust J Physiother2002;48:43–9

12. Verhagen AP, de Vet HC, de Bie RA, et al: The Delphi List:a criteria list for quality assessment of randomized clinicaltrials for conducting systematic reviews developed by Del-phi consensus. J Clin Epidemiol 1998;51:1235–41

13. Downs SH, Black N: The feasibility of creating a checklistfor the assessment of the methodological quality both ofrandomised and non-randomised studies of health care in-terventions. J Epidemiol Community Health 1998;52:377–84

14. Maher CG, Sherrington C, Herbert RD, Moseley AM, ElkinsM: Reliability of the PEDro scale for rating quality of ran-domized controlled trials. Phys Ther 2003;83:713–21

15. Foley NC, Bhogal SK, Teasell RW, Bureau Y, Speechley MR:Estimates of quality and reliability with the physiotherapyevidence-based database scale to assess the methodology ofrandomized controlled trials of pharmacological and non-pharmacological interventions. Phys Ther 2006;86:817–24

16. Bhogal SK, Teasell RW, Foley NC, Speechley MR: ThePEDro scale provides a more comprehensive measure ofmethodological quality than the Jadad scale in stroke reha-bilitation literature. J Clin Epidemiol 2005;58:668–73

17. Deleted in proof

18. Roberts L, Counsell R: Assessment of clinical outcomes inacute stroke trials. Stroke 1998;28:986–91

19. Duncan PW, Jorgensen HS, Wade DT: Outcome measures inacute stroke trials. A systematic review and some recom-mendations to improve practice. Stroke 2000;31:1429–38

20. Marshall M, Lockwood A, Bradley C, Adams C, Joy C, Fenton


M: Unpublished rating scales: a major source of bias inrandomised controlled trials of treatments for schizophre-nia. Br J Psychiatry 2000;176:249–52

21. Moher D, Cook DJ, Jadad AR, et al: Assessing the quality ofreports of randomised trials: implications for the conduct ofmeta-analyses. Health Technol Assess (Rockv) 1999;3:1–98

22. Mahoney FI, Barthel DW: Functional evaluation: the Bar-thel index. Maryland State Med J 1965;14:61–5

23. Butland RJ, Pang J, Gross ER, Woodcock AA, Geddes DM:Two-, six-, and 12-minute walking tests in respiratory dis-ease. Br Med J (Clin Res Ed) 1982;284:1607–8

24. Kosak M, Smith T: Comparison of the 2-, 6-, and 12-minutewalk tests in patients with stroke. J Rehabil Res Dev 2005;42:103–7

25. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind S:The post-stroke hemiplegic patient: I. A method for evalu-ation of physical performance. Scand J Rehabil Med 1975;7:13–31

26. Ashworth B: Preliminary trial of carisoprodal in multiplesclerosis. Practitioner 1964;192:540–2

27. Bohannon RW, Smith MB: Inter rater reliability of a mod-ified Ashworth scale of muscle spasticity. Phys Ther 1987;67:206–7

28. Granger CV, Hamilton BB, Keith RA, Zielezny M, SherwinFS: Advances in functional assessment for medical rehabil-itation. Top Geriatr Rehabil 1986;1:59–74

29. Folstein MF, Folstein SE, McHugh PR: A practical methodfor grading the cognitive state of patients for the clinician.J Psychiatr Res 1975;12:198

30. Lyle RC: A performance test for assessment of upper limbfunction in physical rehabilitation treatment and research.Int J Rehabil Res 1981;4:483–92

31. Holbrook M, Skilbeck CE: An activities index for use withstroke patients. Age Aging 1983;12:166–70

32. Hunt SM, McKenna SP, McEwen J, Backett EM, Williams J,Papp E: A quantitative approach to perceived health status:a validation study. J Epidemiol Community Health 1980;34:281–4

33. Ware JE Jr, Sherbourne CD: The MOS 36-item short-formhealth survey (SF36) I: conceptual framework and itemselection. Med Care 1992;30:473–83

34. McHorney CA, Ware JE Jr, Raczek AE: The MOS 36-itemshort form health survey (SF-36) II: psychometric and clin-ical tests of validity in measuring physical and mentalhealth constructs. Med Care 1993;31:247–63

35. Nouri FM, Lincoln NB: An extended activities of daily livingscale for stroke patients. Clin Rehabil 1987;4:123–8

36. Carr JH, Shepherd RB, Nordholm L, Lynne D: Investigationof a new motor assessment scale for stroke patients. PhysTher 1985;65:175–80

37. Zigmond AS, Snaith RP: The hospital anxiety and depres-sion scale. Acta Psychiatr Scand 1983;67:361–70

38. Berg KO, Wood-Dauphinee S, Williams JL, Maki B: Measur-ing balance in the elderly: preliminary development of aninstrument. Physiother Can 1989;41:304–11

39. Berg KO, Maki B, Williams JL, Holliday PJ, Wood-Dauphinee S: Clinical and laboratory measures of posturalbalance in an elderly population. Arch Phys Med Rehabil1992;73:1073–80

40. Rankin J: Cerebral vascular accidents in patients over theage of 60 II: prognosis. Scott Med J 1957;2:200–15

41. Mathiowetz V, Weber K, Kashman N, Volland G: Adultnorms for the nine-hole peg test of finger dexterity. OccupTher J Res 1985;5:24–37

42. Hamilton MA: A rating scale for depression. J Neurol Neu-rosurg Psychiatry 1960;23:56–62

43. Demeurisse G, Demol O, Robaye E: Motor evaluation invascular hemiplegia. Eur Neurol 1980;19:382–9

44. Goldberg DP, Hillier VF: A scaled version of the GeneralHealth Questionnaire. Psychol Med 1979;9:139–45

45. Collen FM, Wade DT, Robb GF, Bradshaw CM: The River-mead Mobility Index: a further development of the River-mead Motor Assessment. Int Disabil Stud 1991;13:50–4

46. Lincoln N, Leadbitter D: Assessment of motor function instroke patients. Physiotherapy 1979;65:48–51

47. Price DD, McGrath PA, Rafii A, Buckingham B: The valida-tion of visual analogue scales as ratio scale measures forchronic and experimental pain. Pain 1983;17:45–56

48. Price DD, Harkins SW: The combined use of visual analoguescales and experimental pain in providing standardized as-sessment of clinical pain. Clin J Pain 1987;3:3–11

49. Porch BE: The Porch Index of Communicative Ability.Volume 2. Revised Edition: Administration and Scoring.Palo Alto, Consulting Psychologists, 1973

50. Taub E, Miller NE, Novack TA, et al: Technique to improvechronic motor deficit after stroke. Arch Phys Med Rehabil1993;74:347–54

51. Brunnstrom S: Motor testing procedures in hemiplegia.J Am Phys Ther Assoc 1966;46:357

52. Robinson BC: Validation of a caregiver strain index. J Ger-ontol 1983;38:344–8

53. Podsiadlo D, Richardson S: The timed “up and go” test. ArchPhys Med Rehabil 1989;67:387–98

54. Holden MK, Gill KM, Magliozzi MR, Nathan J, Piehl-BakerL: Clinical gait assessment in the neurologically impaired:reliabiity and meaningfulness. Phys Ther 1984;66:1530–9

55. Scandinavian Stroke Study Group: Multicenter trial of he-modiluation in ischemic stroke—background and studyprotocol. Stroke 1985;16:885–90

56. Lezak MD: Neuropsychological Assessment. New York, Ox-ford University Press, 1983

57. Wade DT: Measurement in Neurological Rehabilitation.New York, Oxford University Press, 1992


Authors:John A. Schuchmann, MDBarry A. Browne, PharmD

Affiliations:Departments of Physical Medicineand Rehabilitation (JAS) andMedicine (BAB), Scott and WhiteClinic, Scott, Sherwood, and BrindleyFoundation, Temple, Texas; and TexasA&M University College of Medicine,College Station, Texas (JAS, BAB).

Correspondence:All correspondence and requests forreprints should be addressed to JohnA. Schuchmann, MD, Department ofPhysical Medicine and Rehabilitation,Scott & White Clinic, 2401 South31st St., Temple, TX 76508.


DOI: 10.1097/PHM.0b013e3181152152

Persistent Hiccups DuringRehabilitation HospitalizationThree Case Reports and Review of the Literature

ABSTRACT

Schuchmann JA, Browne BA: Persistent hiccups during rehabilitation hospital-ization: three case reports and review of the literature. Am J Phys Med Rehabil2007;86:1013–1018.

Persistent hiccups have been reported to occur occasionally during rehabilitationhospitalizations. Hiccups can interfere with patient participation and progress, andthis can lengthen and complicate the rehabilitation hospitalization. Chlorproma-zine, the only Food and Drug Administration–approved agent specifically indi-cated for hiccups, is often chosen as a first-line treatment. However, chlorprom-azine does not always provide favorable results.

The intent of this paper is to (1) review hiccups and implications of persistenthiccups in the rehabilitation setting, (2) provide additional evidence that chlor-promazine is often not the best treatment choice for hiccups in rehabilitativepatients, (3) report the effectiveness and favorable tolerability of modest doses ofgabapentin in a small case series of three patients, and (4) review potentiallyeffective treatment approaches for hiccups.

Key Words: Hiccups, Hiccoughs, Singultus, Gabapentin, Stroke, Postoperative

OVERVIEW OF HICCUPSHiccups1,2 (also hiccoughs or singultus) involve sudden, abrupt inspiration

resulting from sudden, involuntary contraction of the diaphragm. Closure of theglottis then halts the incoming air. The column of air strikes the closed glottisto produce the characteristic sound. Hiccups most often involve unilateralcontraction of the diaphragm (the left more frequently than the right).

The reflex arc for hiccups includes an afferent limb (including the vagusand phrenic nerves and the sympathetic chain), a central mediator, and anefferent limb (including the phrenic nerve, plus efferents to the glottis andinspiratory intercostal muscles). The reflex pathways are similar to those thatproduce coughing, sneezing, swallowing, and vomiting.

Hiccups do not seem to have any physiologic purpose or function afterbirth, but speculation exists that intrauterine hiccups permit training of thediaphragm without aspiration of amniotic fluid. Most cases of hiccups are ofshort duration, lasting less than 48 hrs. No gender difference seems to exist inthe occurrence of transient hiccups. Brief bouts of hiccups do not requiremedical intervention.

December 2007 Persistent Hiccups During Rehabilitation 1013

REVIEW & CASE REPORTS

Hiccups

Transient hiccups are benign in nature andoccur nearly universally in all individuals fromtime to time. Such events are often spontaneousbut can be related to gastric overdistention (suchas overeating, drinking carbonated beverages,aerophagia, and gastric insufflation during endos-copy), sudden changes in ambient or gastrointes-tinal temperature, alcohol use, tobacco use, suddenexcitement, or emotional distress.

Intractable hiccups occur more frequently inmen, and about 4000 hospitalizations per year oc-cur in the United States for hiccups. The longer theduration of hiccups, the more likely an organiccause exists. Hiccups occurring at night also havea greater chance of having an organic vs. a psycho-genic cause. Numerous potential causes of medi-cally significant hiccups have been described,1,2

most commonly including central nervous systemdisorders, lesions of the vagus and/or phrenicnerve, toxic–metabolic disorders, adverse eventsassociated with pharmacological agents, and psy-chological disorders. Hiccups are also not uncom-mon in the postoperative state.

The longer hiccups persist, the more difficulttreatment becomes. Persistent hiccups can prolonghospitalizations by contributing to debilitationwith diminished nutritional and fluid intake, aspi-ration, insomnia, mental stress, and even wounddehiscence.

REVIEW OF THE REHABILITATIONLITERATURE

Two published articles concerning hiccupswere identified in the rehabilitation literature.Nickerson et al.3 report that hiccups developed in a69-yr-old male with lateral medullary syndromeduring his acute care stay. He was given chlor-promazine, promethazine, and prochlorperazine,without success. On admission to rehabilitation,his main complaints were of hiccups at a rate of25/min, associated with nausea. At the rehabilita-tion facility, baclofen (5 mg orally, three times aday) was initiated, and the hiccups resolved within48 hrs.

Kumar and Dromerick4 describe three patientswith strokes in the pontine region who developedintractable hiccups while participating in rehabili-tation programs. Complications seen in these pa-tients included dysphagia requiring enteral feed-ing, aspiration pneumonia requiring intravenousantibiotics, weight loss, respiratory depression, andesophagitis. The authors report that the hiccupscontributed to extending the rehabilitation lengthsof stay to 27 days, compared with an average of 16.6days for their other stroke patients. The authorschose chlorpromazine as their first-line therapy;chlorpromazine provided benefit in two of thethree patients, but it led to drowsiness, which lim-

ited participation in rehabilitative therapies. Drugdiscontinuation was required. Hiccup resolutionwas ultimately achieved with baclofen in one case,haloperidol in the second, and carbamazepine inthe third.

The above reports highlight that hiccups canoccur in individuals with stroke usually affectingthe brainstem. Complications can occur and canprotract the rehabilitation length of stay. Chlor-promazine was poorly tolerated, and various otheragents (baclofen, haloperidol, and carbamazepine)provided resolution of the hiccups.

CASE REPORTSThree cases are reported in which persistent

hiccups interfered with rehabilitation program par-ticipation and in which prompt hiccup resolutionwas achieved with the use of gabapentin.

Case 1A 60-yr-old male was admitted to a medical

unit after the acute onset of ataxia, dizziness, anddecreased coordination in the right upper extrem-ity. Hiccups at a rate of 10–15/min began 15–20mins after onset of the neurological symptoms. Amagnetic resonance imaging scan showed the pres-ence of a lateral medullary infarct. The patient wasadmitted to rehabilitation 9 days after onset, stillwith hiccups. His hiccups caused persistent nau-sea, associated with emesis, with resultant poorfluid and nutritional intake. Blood urea nitrogenand creatinine had increased from 16 and 0.9 mg/dlat time of acute care admission to 47 and 1.5 mg/dl,respectively, during the sixth day of his stay on therehabilitation unit. Baclofen was initially usedwhile on the medical unit at a dose of 5 mg orallythree times a day, then increased to 10 mg orallythree times a day on the rehabilitation unit, with-out success. Chlorpromazine at a dosage of 25 mgorally three times a day was substituted for thebaclofen. The patient’s hiccups improved, but hedeveloped worsening dizziness and orthostatic hy-potension, precluding participation in rehabilita-tion therapies. Gabapentin (100 mg orally, fourtimes a day) was then substituted for the chlor-promazine, with prompt resolution of hiccups, diz-ziness, nausea, and emesis, and normalization oflaboratory studies; blood urea nitrogen and creat-inine values had decreased to 15 and 1.0 mg/dl,respectively, at a follow-up visit 3 wks later.

Comments on Case 1Hiccups previously have been described as as-

sociated with lateral medullary infarction (alsoknown as posterior inferior cerebellar syndrome,or Wallenberg syndrome). Park et al.5 present aseries of 51 patients with lateral medullary infarc-tion; magnetic resonance imaging scanning within

1014 Schuchmann and Browne Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

3 days of onset confirmed the diagnosis. Seven ofthese 51 patients developed hiccups in addition tothe more typical symptoms of vertigo, dizziness,nausea, vomiting, and dysphagia.

Pelin et al.6 describe four patients with lateralmedullary syndrome who also presented with hic-cups. These patients were treated with varied dosesand combinations of chlorpromazine, valproic acid,and baclofen. Full resolution of hiccups was ob-served in all patients.

Moretti et al.7 describe the use of gabapentinin 15 patients (nine men and six women; mean age,69.1 yrs) who experienced hiccups caused by lateralmedullary infarcts. The authors report success inall of the patients with use of gabapentin, as wasthe case with our patient. Two years after a stroke,one patient had a recurrence of hiccups that againresponded to gabapentin. Follow-up neurologic ex-amination in this patient remained stable, and re-peat magnetic resonance imaging demonstratedthe previous stroke, but no acute findings weredetected. Moretti described the following gabapen-tin protocol: 400 mg three times a day for 3 days,then 400 mg daily for 3 days, and then discontin-uation of gabapentin. No significant adverse effectswere reported from the gabapentin.

Case 2A 72-yr-old female was initially seen with com-

plaints of headache and clumsiness in the rightlower extremity. She was brought to the emer-gency department. Computed tomographic scan ofthe head was negative. Concerns were expressedabout the upper cervical spine (possible calcifica-tion vs. cavernous angioma vs. small area of hem-orrhage). She was scheduled for a follow-up clinicappointment.

Four days later, she developed increased weak-ness in the right upper and lower extremities, andshe fell in her bathroom. She was brought to theemergency department and was admitted. Coagu-lation studies showed an international normalizedratio of 3.5.

Her past history was significant for hyperten-sion, breast cancer, paroxysmal atrial fibrillation,and remote deep venous thrombosis with pulmo-nary embolism 20 yrs previously (after bilateralmastectomy for breast cancer). She had been onchronic anticoagulation with warfarin, and aGreenfield filter had been placed previously.

Computed tomography of the cervical spinesoon after admission demonstrated a hematomalocalized to the right side of the cervical spinal cord(C2 and above), extending to the pontomedullaryjunction. Anticoagulation was reversed with fresh-frozen plasma. After normalization of her interna-tional normalized ratio, she was admitted to reha-bilitation, with persistent right hemiparesis. A

cervical collar was used to stabilize her cervicalspine because of the presence of the hematoma andbecause she complained of dizziness with just min-imal neck movements. She slowly, steadily im-proved to the point where she was able to becomeindependent with ambulation and self-care activi-ties. She was discharged home to live alone after a22-day rehabilitation unit stay. Follow-up com-puted tomographic scan of the cervical spine beforedischarge demonstrated nearly complete resolu-tion of the hematoma.

Interestingly, at the time of admission to re-habilitation, the patient indicated that she was ex-periencing persistent hiccups occurring at a fre-quency of two to three per minute; she reportedthat the hiccups had been present for 3 days andhad begun about 4 days after the worsening of herright hemiparesis. She stated that the hiccups wereannoying to her and that they interfered with eat-ing and drinking activities. She was prescribedchlorpromazine (25 mg orally, three times a day),and the hiccups greatly improved; however, shecomplained of dizziness, and she developed ortho-static hypotension, which precluded participationin her rehabilitation program. The chlorpromazinewas discontinued, and gabapentin (100 mg orally,four times a day) was initiated, with total resolu-tion of the hiccups and dizziness. The gabapentinwas tapered and discontinued without any recur-rence of hiccups.

Comments on Case 2Hemorrhage within the spinal canal is a rare but

serious complication of anticoagulation therapy, de-scribed most often in individuals in the sixth or sev-enth decade of life. Most cases reported in the litera-ture seem to be spontaneous, but others are describedas secondary to trauma (epidural catheters, spinalmanipulation, snowboarding, nose blowing), small-cell lung cancer, metastatic hepatocellular cancer, orcoagulopathies. Patients described in many of thereported cases had been anticoagulated within thetherapeutic range. The diagnosis should be consid-ered if a patient taking anticoagulant medicationcomplains of local or referred spinal pain and hasassociated limb weakness, sensory deficits, or urinaryretention. Anticoagulation should be reversedpromptly. Emergency laminectomy and decompres-sion of the spinal cord are often recommended(Pullarkat et al.8 describe these measures as “manda-tory”), but cases in which the patient is already im-proving rapidly and steadily and is under close obser-vation could be treated nonoperatively.

Hemorrhage adjacent to the spinal canal issufficiently rare; only limited numbers of caseshave been described in the literature. This is thefirst case reported in which hiccups were present inassociation with such a hemorrhage. The location


of the hemorrhage—from the pontomedullaryjunction descending to C2—is similar in locationto the lesions seen with lateral medullary stroke,implicating similar neurological factors in produc-ing the hiccups.

Case 3A 65-yr-old type 2 diabetic male was admitted

to rehabilitation 6 days after bilateral total-kneearthroplasties. After surgery, he had redness andswelling around his left leg that were thought to beattributable to cellulitis. He had been maintainedon deep venous thrombosis prophylaxis. Doppler ul-trasound studies for deep venous thrombosis werenegative. The cellulitis was treated with cephalexin,with improvement in the redness, swelling, and pain.His overall rehabilitation course was slow because ofpain and decreased active knee motion. Hydrocodonewas used for pain control. He also had hypoalbumin-emia (2.6 g/dl) and required psychosocial encourage-ment and supplements to boost his nutrition.

The patient developed hiccups (3–5/min), whichpersisted during his last day on the surgical serviceand for his first day on rehabilitation; the hiccupsaggravated his postsurgical pain. Gabapentin (100 mgorally, three times a day) was initiated, with promptresolution of the hiccups. After 8 days on the reha-bilitation unit, he had improved to the point where hewas ready for hospital discharge with outpatient ther-apy. Hiccups did not recur.

Comments on Case 3Souadjian and Cain1 analyzed 220 patients

with intractable hiccups seen at the Mayo Clinic.Risks produced by intractable hiccups were re-ported as severe debilitation, dehydration, insom-nia, protracted mental stress, wound dehiscence,and, possibly, even death. Eighty-two percent ofthe patients were male. Forty patients (18% of thetotal group) were reported to have postsurgicalhiccups. Hiccups usually appeared 1–4 days post-operatively and were generally severe and inter-fered with sleep. Initial abortive measures failed in65% of cases, and pharmacological measures wereusually required to resolve the hiccups.

In our patient, an elderly diabetic male, gaba-pentin was chosen as the initial treatment for hishiccups. He had a prompt, complete resolution ofhiccups on low-dose gabapentin, and he toleratedthis medication without adverse effects.

REVIEW OF TREATMENT APPROACHESFOR PERSISTENT HICCUPS

A variety of folk remedies and physical maneu-vers have been described as helpful in abortinghiccups (see Table 1). These measures seem to bemost effective in transient hiccups and are lesseffective with persistent hiccups.

A careful history, thorough physical examina-tion, and studies, as appropriate, should be per-formed to help identify conditions or disorders thatmight be producing or perpetuating the hiccups.The list of potential causes of hiccups is long,including central and peripheral neurological le-sions, head and neck lesions, cardiothoracic andabdominal disorders, infectious diseases, toxic/metabolic causes, postoperative status, and psycho-genic causes.

For persistent hiccups, the patient should bequestioned about any medication that might beproducing or aggravating the hiccups. Medicationcessation or a drug “holiday” should be considered(if possible) if a potentially offending medication isidentified. A number of different medications havebeen implicated in producing or aggravating hic-cups (see Table 2).

Review of treatment options should take intoconsideration the potential precipitating or aggra-vating causes of the hiccups. For example, if hic-cups are related to gastroesophageal reflux disease,a proton pump inhibitor might be quite effective inrelieving symptoms.

Often, no obvious cause is identified to accountfor the hiccups. In these cases, pharmacologic treat-ment is often needed. Initially, monotherapy is rec-ommended, but occasionally a combination of agentsis required to treat persistent hiccups.

Chlorpromazine is the only medication specif-ically approved by the Food and Drug Administra-tion for the treatment of hiccups. Use for hiccupswas first described by Friedgood and Ripstein9 in

TABLE 1 Folk remedies and othernonpharmacological treatments ofhiccups

Breathing into a bagDrinking from opposite side of glassSwallowing granulated sugarDrinking or gargling ice waterForcible traction on tongueBiting on a lemonEating a spoonful of peanut butterCompression of the eyeballsCarotid massageRectal massageCatheter or swab stimulation of the

naso-oropharynxValsalva maneuverBreath holdingFrightNoxious odors (inhaling ammonia)Gastric lavageModified Heimlich maneuver: three

moderate-pressure thrusts at 10-secintervals

Sexual intercourse


1955. The authors reported a series of 50 patients(46 men and 4 women, 26–80 yrs of age) withintractable hiccups treated with intravenous,50-mg injections of chlorpromazine, with the doserepeated once, 2–4 hrs later. Successful hiccupresolution was reported in 41 of the 50 cases. Fivepatients had hiccups recurrence, whereas four pa-tients showed no response to the medication. Re-ported adverse effects included a “depressive andsedative reaction,” and dermatitis in one case. Theauthors also reported that older, more fragile pa-tients had a more profound response to the medi-cation. The dose was, therefore, decreased in el-derly patients. Chlorpromazine, if chosen forintravenous use, must be infused slowly for severalhours to prevent the development of hypotension.Oral chlorpromazine (25–50 mg, three to fourtimes a day, for 7–10 days) also has been shown tobe effective. More recent reports3 indicate that forelderly and potentially neurologically impairedrehabilitation patients, chlorpromazine is bestavoided because of the problematic adverse effectsof dizziness with orthostatic hypotension anddrowsiness, which can preclude participation inactive rehabilitation programs. Two of the patientsin this study experienced adverse responses tochlorpromazine, providing further evidence thatchlorpromazine is often not the most appropriatetreatment choice for hiccups in rehabilitative pa-tients.

Numerous other medications have been re-ported as effective for treating hiccups. Omepra-zole has been found helpful in patients in whomthe hiccups seem to be related to the presence ofgastroesophageal reflux disease. Baclofen and anti-convulsants (phenytoin, valproic acid, carbamaz-epine, and gabapentin) also can be effective in somecases. Occasionally, in resistant cases of hiccups,

combinations of these medications are necessary.Other treatment approaches for resistant cases in-clude hypnosis, acupuncture, cervical epiduralblock, phrenic nerve block, crushing or pacing, andgeneral anesthesia with positive pressure ventila-tion and muscle relaxants.

Gabapentin was found to be well tolerated andsymptomatically useful for our three patients.Gabapentin is a Food and Drug Administration–approved anticonvulsant that is also frequentlyused for neuropathic pain, including diabetic neu-ropathic pain and postherpetic neuralgic pain. Thespecific anticonvulsant mechanism of action is un-clear; the agent is structurally related to gamma-aminobutyric acid (GABA), but it does not interactwith GABA receptors, it is not metabolically con-verted into GABA or a GABA agonist, and it is notan inhibitor of GABA uptake or degradation. Som-nolence, dizziness, ataxia, and fatigue are the mostfrequent adverse effects.

Hernandez et al.10 report a case of a 71-yr-oldmale with Guillain–Barre syndrome and metastaticgastric neoplasm who presented with intractablehiccups. The neoplasm had infiltrated the anteriorpillar of the diaphragm at the time of surgery.Omeprazole (20 mg daily) was initiated along withchlorpromazine (25 mg, three times a day). Hic-cups, as well as emesis, persisted. Chlorpromazinewas discontinued and baclofen was added, initially5 mg three times a day, and progressively increas-ing to 15 mg three times a day. The patient stillfailed to improve. Gabapentin was then added at adosage of 300 mg three times a day. Within 24 hrs,the hiccups ceased. One week later, hiccups re-curred and resolved within 48 hrs after resumptionof gabapentin at a dosage of 400 mg daily.

Porzio et al.11 describe gabapentin as beinguseful in treating hiccups in patients with cancer.A 63-yr-old male with metastatic liver cancer de-veloped chronic hiccups, nausea, and fatigue. Met-oclopramide and dexamethasone prescribed by hisprimary care physician were ineffective. Chlor-promazine (25 mg intravenously, twice daily) easedthe hiccups but caused postural hypotension andsevere drowsiness. Chlorpromazine was discontin-ued, and gabapentin (300 mg, three times daily)was initiated. The gabapentin was well toleratedand provided effective relief from the hiccups. Twoother cases—a 43-yr-old male with pancreatic can-cer, and a 51-yr-old male with metastatic small-celllung cancer—were reported to obtain relief of per-sistent hiccups with 300–400 mg of gabapentinthree times a day. The authors caution about theuse of baclofen in elderly patients because of thefrequent adverse effects of sedation, insomnia, diz-ziness, weakness, ataxia, and confusion.

TABLE 2 Medications implicated as possiblycausing or aggravating persistenthiccups

Alpha methyldopaShort-acting barbituratesCorticosteroids, including dexamethasone and

intravenous methylprednisoloneBenzodiazepines, including diazepam and

chlordiazepoxideChemotherapeutic agents, including cisplatinumOpioids, including hydrocodoneNonsteroidal antiinflammatory drugsAntidopaminergic agents, including perphenazineAntibiotics, including beta-lactams, macrolides,

and fluoroquinolonesCardiovascular agents, including digoxinProgesteroneAnabolic steroids


SUMMARYMost cases of hiccups are transient and benign.

Home treatments are initially reasonable and are of-ten effective. The longer hiccups persist, the moreresistant to treatment they become. Intractable hic-cups can be associated with a number of pathologicconditions. A detailed history and physical examina-tion with appropriate testing should be performed onpatients who are experiencing intractable hiccups.

Chlorpromazine, despite Food and Drug Admin-istration approval for hiccups, is often poorly toler-ated and is best avoided for older rehabilitative pa-tients, especially those with debility, acute stroke, orneurological events. Baclofen can be helpful in pro-moting hiccup resolution, but it should be used cau-tiously in elderly patients because of the potentialadverse effects of sedation, insomnia, dizziness, weak-ness, ataxia, and confusion.

Gabapentin at modest doses, in this case seriesand others, has been found to provide prompt reliefof hiccups, and it is usually well tolerated. A shortcourse often can provide long-term relief. Mea-sures combining pharmacotherapy with more in-vasive treatment are occasionally indicated formore resistant cases of hiccups.

REFERENCES1. Souadjian JV, Cain JC: Intractable hiccup. Etiologic factors

in 220 cases. Postgrad Med 1968;43:72–7

2. Cymet TC: Retrospective analysis of hiccups in patients at acommunity hospital from 1995–2000. J Natl Med Assoc2002;94:480–3

3. Nickerson RB, Atchison JW, Van Hoose JD, Hayes D: Hic-cups associated with lateral medullary syndrome. Am J PhysMed Rehabil 1997;76:144–6

4. Kumar A, Dromerick AW: Intractable hiccups during strokerehabilitation. Arch Phys Med Rehabil 1998;79:697–9

5. Park MH, Kim BJ, Koh SB, Park MK, Park KW, Lee DH:Lesional location of lateral medullary infarction presentinghiccups (singultus). J Neurol Neurosurg Psychiatry 2005;76:95–8

6. Pelin Z, Bozluolcay M, Erkol G, Bugay G, Denktas H:Hiccups as a sign of lateral medullary syndrome: reports offour cases. Cerrahpasa J Med 2001;32:115–9

7. Moretti R, Torre P, Antonello RM, Ukmar M, Cazzato G,Bava A: Gabapentin as a drug therapy of intractable hiccupbecause of vascular lesion: a three-year follow up. Neurol-ogist 2004;10:102–6

8. Pullarkat VA, Kalapura T, Pincus M, Baskharoun R: In-traspinal hemorrhage complicating oral anticoagulanttherapy: an unusual case of cervical hematomyelia and areview of the literature. Arch Intern Med 2000;160:237–40

9. Friedgood CE, Ripstein CB: Chlorpromazine (thorazine) inthe treatment of intractable hiccups. J Am Med Assoc 1955;157:309–10

10. Hernandez JL, Pajaron M, Garcia-Regata O, Jimanez V,Gonzalez-Macias J, Ramos-Estebanez C: Gabapentin for in-tractable hiccup. Am J Med 2004;117:279–81

11. Porzio G, Aielli F, Narducci F, et al: Hiccup in patients withadvanced cancer successfully treated with gabapentin: re-port of three cases. N Z Med J 2003;116:U605


Authors:Xiulu Ruan, MDJohn Patrick Couch, MDRinoo Shah, MDFrank Wang, MDHai Nan Liu, MS

Affiliations:From Physicians’ Pain Specialists ofAlabama, Mobile, Alabama (XR, JPC);Guthrie Clinic, Sayre, Pennsylvania(RS); School of Medicine, Universityof South Alabama, Mobile, Alabama(FW); and Department of Urology,Qilu Hospital, Shandong University,Jinan, China (HNL).

Correspondence:All correspondence and requests forreprints should be addressed to XiuluRuan, MD, Physicians’ PainSpecialists of Alabama, 2001Springhill Ave., Mobile, AL 36607.


DOI: 10.1097/PHM.0b013e31815206c8

Persistent Hiccup Associated withIntrathecal Morphine Infusion PumpTherapy

ABSTRACT

Ruan X, Couch JP, Shah R, Wang F, Liu HN: Persistent hiccup associated withintrathecal morphine infusion pump therapy. Am J Phys Med Rehabil 2007;86:1019–1022.

Intraspinal drug-delivery therapy has been increasingly used in patients withintractable nonmalignant pain syndromes during the past two decades. Morphine,the only FDA-approved opioid for intrathecal administration, has been the prin-ciple agent for such therapy. Although intrathecal morphine infusion can produceprofound spinal analgesia, it may also cause some untoward side effects. Wedescribe the first case of persistent hiccup caused by intrathecal morphineinfusion therapy.

Key Words: Hiccup, Intrathecal Opioid Infusion Pump, Intrathecal Morphine, Intrathe-cal Hydromorphone

Hiccup (singultu, hiccough) is a common phenomenon that involves suddeninspiration, immediately followed by active closure of glottis.1 It is an involun-tary, reflex-like activity that begins with contraction of the diaphragm, quicklyterminated by the abrupt closure of the glottis. Hiccups are experienced at onetime or another by almost everyone and, when transient, are usually harmless.Despite a common occurrence, the cause, mechanism, and significance ofhiccups are still incompletely understood.2 Although the causes are oftenconsidered idiopathic, more than 100 known organic causes have been identi-fied.3 The most common causes are gastrointestinal, such as gastric distentionor gastroesophageal reflux disease.4 Metabolic derangements, as well as drugs,are also frequently implicated as probable causes.1,4

We report the first case of intractable hiccups presumably caused by intra-thecal morphine infusion.

CASE REPORTA 65-yr-old female, referred to our multidisciplinary pain center, with a 5-yr

history of severe low-back pain and left-sided leg pain from degenerative lumbardisc disease, lumbar spinal stenosis, and lumbar radiculitis. The patient also hada history of lumbar laminectomy with decompressive fusion at L4–L5 andL5–S1, about 2 yrs ago after she had undergone multimodality treatment,including medication trials, physical therapy, and spinal interventional treat-

December 2007 Persistent Hiccup and Intrathecal Morphine 1019

CASE REPORT

Hiccup

ment, including epidural steroid injections, lum-bosacral transforaminal steroid injections, lumbarfacet blocks, and lumbar facet radiofrequency rhi-zotomy, with only short-term efficacy. The patientdescribed her low-back pain as being about as both-ersome as her left-leg pain. Her pain level wasusually at 8–9/10 on a numeric pain scale of 0–10.Her low-back pain generally worsened by sittingand standing, whereas her left-leg pain, which shedescribed as starting from the posterior thigh, ex-tending down to the posterior calf, and then to thebottom of her left foot, significantly worsened bywalking. The only relieving factor was frequentlychanging positions. She had no significant pastmedical history or surgical history, except for lum-bosacral decompressive fusion about 2 yrs beforepresenting to our clinic. Her family history andreview of systems were noncontributory. Her med-ications included extended-release morphine (30mg, thrice daily), oxycodone/acetaminophen (10/325, thrice daily, as needed), gabapentin (800 mg,thrice daily), tizanidine (4 mg, twice daily), andmeloxicam (7.5 mg, twice daily). She had been onthese drugs for more than 6 mos. Her lumbar MRIwith and without contrast revealed broad-basedbulging at L3–L4, with degenerative changes of thefacet joints and bilateral neuroforaminal narrow-ing, more so on the left side. Postsurgical changeswere at L4–L5 and L5–S1 with bilateral laminec-tomy defects and compression plates with interpe-dicular screws at L4, L5, and S1.

Because she had failed all other conventionaltreatment, we considered her for intraspinal drug-delivery therapy. After a psychological evaluationconfirmed her candidacy, she underwent an outpa-tient controlled continuous epidural morphinetrial. A tunneled lumbar epidural catheter wasplaced at L3–L4, with the catheter tip advanced toL1 under fluoroscopic guidance. Satisfactory cath-eter placement was confirmed by epiduragram. Theproximal tip of the catheter was then tunneledsubcutaneously and was connected to a MicrojectPCEA pump (Codman, Raynham, MA) and a reser-voir bag containing 0.4 mg/ml of preservative-freemorphine. The pump was programmed to deliver abasal rate of 0.5 ml/hr. The bolus dose was 0.2 ml,with a 60-min lockout interval. The patient wasthen instructed how to use the pump and wasdischarged home. During the trial, the infusionrate was further increased to 1 ml/hr, with herbolus dose increased to 0.4 ml. The infusion triallasted 7 days and was beneficial in controlling herpain. The patient reported more than 80% painreduction, with improved distance for ambulation.She did experience initial, transient itching, whichcompletely resolved. She did not feel the need touse any extra boluses while on epidural morphineinfusion at 0.4 mg/hr (0.4 mg/ml at 1 ml/hr). She

did not experience any hiccup during this epiduralmorphine-infusion trial. She subsequently con-sented and was scheduled for permanent intrathe-cal morphine pump implantation. The intrathecalcatheter was inserted at the left paramedian L2–L3,with the catheter tip located at T12, as confirmedunder fluoroscopy. A nonprogrammable Codman3000 continuous-rate infusion pump was placed inthe right-middle quadrant below the right rib cageand above the right iliac crest. The intrathecalinfusion consisted of 1.0 mg/day of preservative-free morphine.

The patient started experiencing hiccups about12 hrs after the intrathecal pump implantation.She did experience satisfactory analgesia of herusual back and leg pain, and there was no otherintolerable side effect, except for the hiccups. Theseoccurred at a frequency of four to six per minute.The patient was instructed to try different maneu-vers, including holding her breath, drinking water,etc., without efficacy. The patient then tried chlor-promazine as needed, which did help her slightly.She was able to sleep at night, and her hiccupsbecame less forceful. She also noted that the inter-val between episodes of hiccups was increased. Yet,she still experienced episodic hiccups while takingchlorpromazine. Six days after the initial implan-tation, the decision was made to change her intra-thecal infusion from morphine to hydromorphoneat 0.3 mg/day. Her hiccups resolved completely 16hrs after intrathecal morphine had been changedto hydromorphone. She has not experienced recur-rence of her hiccups since switching to hydromor-phone. She has been on intrathecal hydromorphonefor almost 3 yrs now. Her current intrathecal hydro-morphone-infusion rate is 2.5 mg/day.

DISCUSSIONThe discovery of highly specific opioid recep-

tors in the central nervous system, especially thespinal cord in 1970s, made it possible for spinalapplication of opioids to obtain spinally mediatedantinociception. Intraspinal drug-delivery therapyhas been used increasingly since the 1980s in pa-tients with intractable nonmalignant pain whohave failed to respond to conventional treatment orwho could not tolerate systemic opioid therapy be-cause of side effects.5–9 By infusing small amounts ofopioid directly into cerebrospinal fluid in closeproximity to the receptor sites in the spinal cord,one can achieve spinally mediated analgesia, spar-ing some of the side effects caused by systemicopioids. In our own practice, most of our patientswho ended up requiring intrathecal opioid infusionpumps for better pain control were those withintractable low-back pain and/or leg pain fromfailed back surgery syndrome, with/or without ad-hesive arachnoiditis. Morphine, the only FDA-ap-

1020 Ruan et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

proved opioid for intrathecal administration, hasbeen found to be associated with numerous sideeffects.10,11 The most common ones are pruritus,nausea and vomiting, urinary retention, water re-tention, constipation, and sexual dysfunction.10–12

Less common ones include mental status change,central nervous system excitation, hyperalgesia,neurotoxicity, and respiratory depression.10,11 Tothe best of our knowledge, hiccups induced byintrathecal morphine infusion have never been re-ported. No record of any publications in English onhiccups associated with intrathecal opioid usecould be found after Medline search using hic-cup(s) and intrathecal opioid(s). Hydromorphoneis considered a safe, alternative opioid for long-term intrathecal use when morphine fails becauseof pharmacological side effects or inadequate painrelief.13,14 Systemic opioids have been implicatedin causing hiccups, although there is a paucity ofdata in the English literature. Wilcox15 has re-ported one case of a 63-yr-old man with severeneck pain from squamous cell carcinoma of righttonsil, experiencing persistent hiccups after takingslow-released oral morphine. Hiccups resolvedonly after cessation of oral morphine, with no re-current hiccups after substitution with a fentanylpatch and oxycodone. Throughout the period whenhe was experiencing hiccups, the patient had noother apparent side effects that would suggest anopioid overdose. Lauterbach16 has reported a caseof a 55-yr-old man who experienced persistent hic-cups for 4 days after receiving hydrocodone (5 mg,every 4 hrs) after dental procedures, resolving afterhydrocodone cessation. Bagheri et al.17 have shownthat opioids accounted for 6% of the 53 cases ofdrug-induced hiccups in France between 1985 and1997. However, because of the small sample size,6% of 53 cases would correlate to a total of threecases in 12 yrs in France.

Transient hiccups are common and usuallyharmless. Hiccups lasting more than 48 hrs areconsidered persistent; those lasting more than 2mos are considered intractable.18 Persistent or in-tractable hiccups can lead to fatigue, sleep distur-bance, dehydration, and wound dehiscence in thepostoperative period.4

The exact mechanisms of hiccups are still notcompletely understood.

It is believed that the hiccup reflex consists of anafferent and efferent limb. The afferent limb com-prises the phrenic and vagal nerves as well as thesympathetic chain arising from thoracic segmentsT6–T12.3,19 The central connection is thought toinvolve interactions among the brainstem; the mid-brain, including the respiratory center, phrenic nervenuclei, medullary reticular formation, and hypothal-amus; and unidentified areas in the cervical spinalcord between C3 and C5.20 Stimulation of either the

afferent or efferent limb of the reflex arc triggershiccups.

Multiple causes have been attributed to inducehiccups. The most commonly recognized ones aregastrointestinal, such as gastric distention or gas-troesophageal reflex disease. Multiple pharmaco-logical agents,3,19 including barbiturates, benzodi-azepines, methyldopa, and steroids, have beenreported to cause hiccups. More than 100 knownorganic causes of hiccups have been identified.21,22

Local anesthetics given epidurally have beenimplicated as a cause of hiccups. McAllister et al.2

report one case of recurrent persistent hiccupsafter epidural steroid and bupivacaine injections ina 65-yr-old man with lumbar spinal stenosis. Epi-dural bupivacaine, rather than the steroid, was feltto be the culprit for causing hiccups, because hic-cups occurred on three separate occasions afterlumbar epidural injections with bupivacaine andtriamcinolone, whereas hiccups did not occur aftera fourth lumbar epidural injection with the sameamount of triamcinolone, with normal saline in-stead of bupivacaine. Interestingly, to the contrary,intravenous and nebulizer lidocaine have beenused as a therapeutic modality for persistent orintractable hiccups.23,24 Lower thoracic or highlumbar epidural local anesthetics may cause sympa-thetic blockade, resulting in contracted bowel attrib-utable to parasympathetic dominance.25 However,this seems to be in contrast to the gastric distentionthat is thought to be frequently implicated in causinghiccups.4 Diaphragmatic function can also be affectedby epidural blockade. Increased electrical diaphrag-matic activity was noted with thoracic epidural anes-thesia for upper abdominal or thoracic surgery.26,27

The most likely explanation for thoracic epidural an-esthesia-related increases in diaphragmatic activityseems to be the disruption of the inhibitory reflex ofphrenic nerve motor drive, caused by either directdeafferentation of visceral sensory pathways or bydiaphragmatic load reduction as a result of increasedabdominal compliance.25

Slipman et al.28 report a case of persistenthiccups associated with transforaminal thoracicepidural steroid injection with 3 ml of betametha-sone and 1 ml of 1% lidocaine in separate occasionsin a 31-yr-old man with thoracic discogenic pain.The hiccups lasted 15 and 18 hrs, respectively, andwere unresponsive to conventional maneuvers, in-cluding drinking water, holding breath, and initialchlorpromazine (10 mg). The hiccups were finallyaborted after four doses of metoclopramide (10 mg,every 6 hrs).

The authors attribute the hiccups to the epi-dural steroid, rather than lidocaine, because theybelieve that the injectate flow was confirmed in theepidural space, dorsal to the vertebral body, andthat it did not, therefore, reach the sympathetic

December 2007 Persistent Hiccup and Intrathecal Morphine 1021

chain that lies along the anterolateral aspects ofT6–T12 vertebral bodies. However, lower thoracicor high lumbar epidural local anesthetics maycause sympathetic blockade,25 simply because sym-pathetic blockade may still occur with epidurallocal anesthetic, but at more a proximal anatomiclocation instead of at the sympathetic chain.

In our patient, hiccups occurred only afterintrathecal morphine infusion with no local anes-thetic infusion. She had been on oral morphine(extended-release morphine, 30 mg, thrice daily)and oxycodone/acetaminophen (10 mg, thricedaily, as needed) for more than 6 mos withoutexperiencing any hiccups. The 1-wk, epidural mor-phine-infusion trial went well, with good analgesiaand without any hiccups or other intolerable sideeffects. The hiccups persisted for 6 days despitetreatment, only to resolve after the patient wasswitched to an intrathecal infusion of hydromor-phone. It is not clear why intrathecal morphineinduces hiccups, whereas intrathecal hydromor-phone does not. Indeed, it is puzzling becausehydromorphone, a hydrogenated ketone analog,shares similar properties to morphine, except forbeing more lipophilic than morphine and aboutfive to seven times more potent than morphine.29

Because of ethical concerns, we did not replaceintrathecal morphine infusion with normal salineand rechallenge the patient later on with intrathe-cal morphine to further confirm causality of hic-cups by intrathecal morphine infusion.

In summary, we present a rare case of persis-tent hiccups after intrathecal morphine pumpplacement. The exact mechanism of intrathecalmorphine, not hydromorphone, inducing hiccupsneeds further investigation. It is important to rec-ognize this possible adverse event so that propertreatment can be initiated.

REFERENCES1. Launois S, Bizec JC, Whitelaw WA, et al: Hiccup in adults:

an overview. Eur Respir J 1993;6:563–75

2. McAllister RK, McDavid AJ, Meyer TA, Bittenbinder TM:Recurrent persistent hiccups after epidural steroid injectionand analgesia with bupivacaine. Anesth Analg 2005;100:1834–6

3. Rousseau P: Hiccups. South Med J 1995;88:175–81

4. Smith HS, Busracamwongs A: Management of hiccups inthe palliative care populations. Am J Hosp Palliat Care2003;20:149–54

5. Krames E: Intraspinal analgesia for nonmalignant pain, inWaldman SD: Interventional Pain Management, ed 2. Phil-adelphia, PA, Saunders, 2001, pp 609–619

6. Raphael JH, Southall JL, Gnanadurai TV, Treharne GI, KitasGD: Long-term experience with implanted intrathecal drugadministration systems for failed back syndrome and me-chanical low back pain. BMC Musculoskeletal Disord 2002;3:17

7. Deer T, Chapple I, Classen A, et al: Intrathecal drug delivery

for treatment of chronic low back pain: report from theNational Outcomes Registry for Low Back Pain. Pain Med2004;:6–13

8. Kumar K, Kelly M, Pirlot T: Continuous intrathecal mor-phine treatment for chronic pain of nonmalignant etiology:Long-term benefits and efficacy. Surg Neurol 2001;55:79–88

9. Anderson VC, Burchiel KJ: A prospective study of long-termintrathecal morphine in the management of chronic non-maligmant pain. Congress Neurol Surg 1999;44:289–300

10. Ruan XL: Drug-related side effects of long-term intrathecalmorphine therapy. Pain Physician 2007;10:357–65

11. Chaney MA: Side effects of intrathecal ad epidural opioid.Can J Anaesth 1995;42:891–903

12. Abs R, Verhelst J, Maeyaert J, et al: Endocrine consequencesof long-term intrathecal administration of opioids. J ClinEndocrinol Metab 2000;85:2215–22

13. Anderson VC, Cooke B, Burchiel KJ: Intrathecal hydromor-phone for chronic nonmalignant pain: a restrospectivestudy. Pain Medicine 2001;2:287–97

14. Hassenbusch SJ, Portenoy RK, Cousins M, et al: Polyanal-gesic Consensus Conference 2003: an update on the man-agement of pain by intraspinal drug delivery—report of anexpert panel. J Pain Symptom Manage 2004;27:540–63

15. Wilcox SK: Persistent hiccups after slow-release morphine.Palliat Med 2005;19:568–9

16. Lauterbach EC: Hiccup and apparent myoclonus afterhydrocodone: review of the opiate-related hiccup and my-oclonus literature. Clin Neuropharmacol 1999;22:87–92

17. Bagheri H Cismondo S, Montastruc JL: Drug-inducedhiccup: a review of France pharmacological vigilance data-base [in French]. Therapie 1999;54:35–9

18. Souadjian JV, Cain JC: Intractable hiccup: etiologic factorsin 220 cases. Postgrad Med 1968;43:72–7

19. Kolodzik PW, Eilers MA: Hiccups (singultus): review andapproach to management. Ann Emerg Med 1991;20:565–73

20. Bilotta F, Pietropaoli P, Rosa G: Nefopam for refractoryhiccups in the palliative care population. Am J Hosp PalliatCare 2003;20:149–54

21. Harley C: Extradural corticosteroid infiltration: a follow-upstudy of 50 cases. Ann Phys Med 1967;9:22–8

22. Abram SE, O’Connor TC: Complications associated withepidural steroid injections. Reg Anesth 1996;21:149–62

23. Cohen SP, Lubin E, Stojanovic M: Intravenous lidocaine inthe treatment of hiccup. South Med J 2001;94:1124–5

24. Neeno TA, Rosenow EC 3rd: Intractable hiccups: considernebulized lidocaine. Chest 1996;110:1129–30

25. Cousins MJ, Veering BT: Epidural Neural Blockade. NeuralBlockade in Clinical Anesthesia and Management of Pain.Philadelphia, PA, Lippencott-Raven, 1998

26. Mankikian B, Cantineau JP, Bertrand M, et al: Improvementof diaphragmatic function by a thoracic extradural blockafter upper abdominal surgery. Anesthesiology 1988;68:379–86

27. Pansard JL, Mankikian B, Betrand M et al: Effects of tho-racic extradural block on diaphragmatic electrical activityand contractility after the upper abdominal surgery. Anes-thesiology 1993;78:63–71

28. Slipman CW, Shin CH, Patel RK, et al: Persistent hiccupassociated with thoracic epidural steroid injection. Am JPhys Med Rehabil 2001;80:618–21

29. Mahajan G, Fishman SM: Major opioids in pain manage-ment, in Benzon HT, Raja SN, Molloy RE, et al. (eds):Essentials of Pain Medicine and Regional Anesthesia, ed 2.Philadelphia, PA, Elsevier Churchill Livingstone, 2005, pp94–112

1022 Ruan et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

Authors:Kristi L. Kirschner, MDRebecca Brashler, LCSWTeresa A. Savage, PhD, RN

Affiliations:From the Donnelley Family DisabilityEthics Program, RehabilitationInstitute of Chicago, Chicago, Illinois(KLK, RB, TAS); Departments ofPhysical Medicine and Rehabilitationand Medical Humanities & Bioethics,Northwestern University FeinbergSchool of Medicine, Chicago, Illinois(KLK, RB); and Department ofMaternal Child Nursing, University ofIllinois at Chicago, Chicago, Illinois(TAS).

Correspondence:Kristi L. Kirschner, MD,Rehabilitation Institute of Chicago,345 E. Superior St., Room 1136,Chicago, IL 60611.


DOI: 10.1097/PHM.0b013e31815b77b2

Ashley X

ABSTRACT

Kirschner KL, Brashler R, Savage TA: Ashley X. Am J Phys Med Rehabil2007;86:1023–1029.

The recent case of Ashley X has sparked much recent public debate andcontroversy, and raises critical questions for physiatrists and rehabilitation pro-fessionals. The case came to light when Gunther and Diekema published anarticle in the October 2006 issue of Archives of Pediatric and AdolescentMedicine describing a novel growth attenuation treatment for Ashley X, a 6-yr-oldgirl with developmental disabilities. Her parents also published a blog aboutAshley, with detailed explanations for their rationale in choosing this treatment onbehalf of Ashley which involved beliefs about her future and quality of life.Ashley’s parents refer to the series of interventions as the Ashley Treatment intheir blog, and this phrase has also been adopted by the popular press and otherswho have commented on the case. In this article we present an analysis of theAshley X treatment and use a disability ethics approach to examine the perspec-tives of various stakeholders involved, including Ashley and other girls withextensive disabilities, parents, physicians, and bioethics committees. We con-clude with critical questions for physiatrists and other disability specialists who arein a unique position to examine medical controversies involving people withdisabilities.

Key Words: Disability, Ethics, Evidence-Based Medicine, Pediatrics

Recently the parents of a young girl with disabilities sought controversialmedical interventions to attenuate her growth.1 This case, described by Guntherand Diekema2 in an article published in the October 2006 issue of Archives ofPediatric and Adolescent Medicine, touches on many concerns central to thepractice of rehabilitation medicine. The following commentary on the Ashley Xtreatment uses a disability ethics approach3 to examine the perspectives ofvarious stakeholders involved. Such an approach opens up alternative ways toframe the issues, various treatment options, and considerations for the role ofhealth care providers as moral agents facilitating difficult decisions on behalf ofchildren with significant cognitive and developmental disabilities.

BACKGROUNDAccording to her physicians, Ashley X is a disabled girl with a diagnosis of

static encephalopathy, the etiology of which is unknown.2 She was born at fullterm after an unremarkable pregnancy and uncomplicated birth. She has twohealthy siblings and lives with her parents, both of whom have college educa-

December 2007 Commentary 1023

COMMENTARY

Ethics

tions.2 She is described as being dependent in allactivities of daily living, and completely immobile,hence the parents’ term “Pillow Angel”—that is,she stays where they put her.1 She is nonverbal andreceives all of her nutrition via a feeding tube.Ashley X responds to others in her environment viasmiles and vocalizations. Her developmental stageis described as that of a 3-mo-old infant, and herphysicians have indicated that they do not expectthis to change.

When Ashley X was 6 yrs 7 mos old, andbeginning to show signs of early puberty, her par-ents approached a pediatric endocrinologist for anevaluation. In the 6 mos before the evaluation withthe endocrinologist she had advanced from the50th to the 75th percentile for length. The parents“particularly feared that continued growth eventu-ally would make it untenable for them to care fortheir daughter at home, despite their strong desireto do so . . .. After extensive consultation betweenparents and physician, a plan was devised to atten-uate growth by using high-dose estrogen.”2 (1014)

To avoid complications from high-dose estrogen,Ashley also underwent a hysterectomy and breastbud removal. It is this combination of interven-tions that the parents later refer to as the AshleyTreatment.

The parents’ blog provides additional informa-tion about their rationale and states that Ashleyrequires all the care of a baby.1 The parents felt thatthey were approaching the limits of their ability tolift Ashley at 65 lbs. Other than her parents, thegrandmothers were her only care providers andthey were beginning to have difficulty lifting Ash-ley. They felt that by accelerating the closure of hergrowth plates it would be easier to care for her inthe home and would improve her quality of life byfacilitating mobility as well as community par-ticipation. Ashley’s parents were also concernedabout the potential for sexual abuse if she wereto develop secondary sex characteristics, discom-fort from menstruation, large or fibrocystic breasts(which were common in her family)and difficultywith wheelchair harness fitting. A number of othersecondary benefits and concerns were raised in-cluding the prevention of skin breakdown, infec-tions, and ease in bathing. Quoting from the par-ents’ blog, they also state that “given Ashley’smental age a nine and a half year old body is moreappropriate and more dignified than a fully grownfemale body.”1

DISABILITY ETHICSDisability ethics incorporates a disability stud-

ies approach to traditional clinical medical ethics.Disability studies is an emerging interdisciplinaryfield that is grounded in the lived experience ofdisability and in the social model of disability.4

Simplistically stated, the medical model of disabil-ity characterizes disability as a defect intrinsic tothe individual.5 The primary goal in the medicalmodel is to cure or normalize the person whenpossible, to treat disease and relieve pain. In thesocial model, disability is seen as inherently neu-tral, and a product of the interactions between theperson and the environment.5 Neither model fullyexplains all aspects of disability. More nuancedmodels see disability as a “complex interaction ofbiological, psychological, cultural and sociopoliti-cal factors which cannot be extricated except withimprecision.”6 (p38) In the analysis of the AshleyTreatment that follows, we have a clear illustrationof just how complex these interactions are and thechallenges they pose for healthcare providers.

Ethical Issues for Stakeholders: Ashleyand Other Young Girls with Cognitiveand Physical Disabilities

In December 2006, the United Nations GeneralAssembly adopted the Convention on the Rights ofPersons with Disabilities, stating that “every per-son with disabilities has a right to respect for his orher physical and mental integrity on an equal basiswith others.”7 The American College of Obstetricsand Gynecology (ACOG) came to a similar conclu-sion in their position paper on the “Sterilization ofwomen, including those with mental disabilities.”8

They concluded that hysterectomy should not beused as the primary and initial treatment for deal-ing with menstruation, or for nonvoluntary steril-ization for girls with disabilities. “In disabledwomen with limited functional capability, indica-tions for major surgical procedures remain thesame as in other patients. In all cases, indicationsfor surgery must meet standard criteria, and thebenefits of the procedure must exceed known pro-cedural risks.”8 The concerns raised by caregiversfor Ashley regarding menstrual cramps and hy-giene issues are only speculative. It is critical toassess if there are less risky and reversible treat-ments available that might be equally effective be-fore embracing riskier and permanent interven-tions.

ACOG further notes that using medical treat-ments for social or quality-of-life reasons can beproblematic. Many of the concerns raised in sup-port of the Ashley Treatment—such as Ashley’sfuture care, risk of sexual abuse, wheelchair fitting,and community participation—cannot be resolvedsimply by medical efforts to keep her small, byperforming an hysterectomy or mastectomy. Onthe other hand, physicians are often in a position tomake referrals to other specialists (e.g., indepen-dent living centers, rehabilitation professionals,etc.) when it is clear that the problems identifiedextend beyond medical and biological factors to

1024 Kirschner et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

environmental and social concerns. The ACOG andUN perspectives can both be summarized as a rightto bodily integrity for people with cognitive anddevelopmental disabilities. Involuntary (against theperson’s wishes) or nonvoluntary (for those such asAshley who cannot participate in the decision mak-ing) medical interventions and surgical proceduresin general should only be provided when there areclear health benefits.

Furthermore, people with cognitive and devel-opmental disabilities have been harmed in the pastby controversial and experimental medical treat-ments and need added protections. The history ofmedicine is replete with examples of harm to peoplewith disabilities by well-intentioned doctors, usingpopular but misguided and inadequately tested treat-ments such as prefrontal lobotomies.9,10 New treat-ments for those with disabilities must undergo thesame scrutiny and rigorous research to elucidate theactual benefits and risks as do new medical interven-tions for the general population. Many people withcognitive and developmental disabilities are unable toprovide informed consent and rely on others to makedecisions on their behalf. They are particularly vul-nerable and therefore require additional safeguards.

Ashley and other children with cognitive anddevelopmental disabilities have interests and needsapart from those of their parents. It is difficult toseparate Ashley’s interests from those of her par-ents—clearly her future well-being will largely bedetermined by the ability of her family to continueto provide or arrange for a loving home and 24-hrcare. Children with developmental disabilities whoare dependent upon their parents for all of theirdaily care and who may never be able to commu-nicate their opinions and desires are often not seenas having distinct interests. Yet, it is precisely be-cause of such dependencies and vulnerability thathealth care professionals need to be aware thatsociety has placed some parameters around thetypes of decisions parents can and cannot make.For example, it is no longer acceptable that parentsof newborns with Down syndrome can refuse life-saving surgeries for treatable esophageal atresias orheart anomalies, or for parents to refuse blood trans-fusions or high benefit life-saving interventions fortheir children based upon the parental religious be-liefs (e.g., Jehovah’s Witnesses, or Christian Sci-ence)11,12 There are also limits in place regardingparental decisions for pediatric sterilization includingstate laws such as the one in Washington whereAshley had her surgery.13,14 In rare instances, thestate is empowered to step in and take over the careof a child under the parens patriae power when theparents or guardians are unable or unwilling to act inthe best interest of the child.

Finally, being described as a girl with the men-tal age of an infant may unnecessarily limit expec-

tations for the future and ignores Ashley’s capacityfor social and emotional development. Develop-mental skills and abilities of children with perva-sive needs, in the absence of known progressiveneurological disorders, can change with time, nur-turance and stimulation, just as they can changefor children without disabilities.

Ethical Issues for Stakeholders: Ashley’sParents and Other Caregiving FamilyMembers

The lack of adequate social supports, includingreliable attendant care, quality respite services andaffordable community-based living programs for peo-ple with disabilities remains problematic. Parentswould not need to seek risky medical interventions tominimize difficulties with caregiving if they wereoffered reasonable alternatives. If Ashley’s parentswere convinced that she would always receive lovingand capable care, either in their home or in alterna-tive community-based programs, regardless of hersize and sexual maturation, their concerns could havebeen mitigated. This appalling lack of support lies atthe center of this debate and cannot be ignored orfixed with medical interventions such as surgery anddrugs.

It is also true that parental caregivers havebeen harmed by judgmental health care profession-als who have incomplete appreciation of their livedexperience as caregivers. There has been a ten-dency in medicine to criticize and pathologize fam-ily members who seek solutions to caregiving chal-lenges. Parents are often viewed as selfish if theyinstitutionalize their children, yet overprotective ifthey keep them at home without outside assis-tance. Parents are also often left with inadequateresources and information when making decisionson behalf of their children. When analyzing thetreatment options for Ashley, rehabilitation profes-sionals need to be careful not to contribute to thispattern by blaming her parents for their efforts toseek unconventional solutions in the absence ofquality resources for families of children with dis-abilities.

It is also important to note that in most in-stances parents are in the best position to makedecisions on their child’s behalf. Parents are im-bued with the legal and moral responsibility to carefor their children.15 In making medical decisions,they are expected to use a “best interest” standard,weighing the proposed benefits and burdens ofeach decision.16 Some healthcare professionals andbioethicists argue that considering the child’s “bestinterests” in isolation is unrealistic given that par-ents also have to balance the needs and interests ofother family members.17,18 The lives of physicians,health care professionals and other stakeholdersare not affected by these decisions in the same way


as the lives of Ashley’s family members. Thereforethe parents’ decisions, while not without limits,must carry significant weight.

Ethical Issues for Stakeholders:Physicians and Healthcare Professionals

People with disabilities are high users of med-ical services, and most physicians are not trained todeal with the complex issues that accompany dis-ability. Patients with permanent disabilities havehistorically posed particular challenges for healthcare professionals. People with disabilities oftenrequire health interventions to prevent diseases andcomplications from occurring, but invariably haveenvironmental and social components that compli-cate their degree of function, social participation,and quality of life. Many physicians have neverheard of the social model of medicine, or GeorgeEngel’s biopsychosocial model.19 Nor are they fa-miliar with disability resources and specialists(such as physiatrists, independent living centers,etc.). Finally, numerous studies have demonstratedthat health care professionals are not very accurateat predicting what life with disability is like and,indeed, consistently underestimate the quality oflife as expressed by people who live with disabilitiesand chronic conditions.20–22 Such a dearth of in-formation and experience places both families andphysicians at a great disadvantage when attemptingto make decisions on behalf of a child with exten-sive disabilities.

Physicians have responsibility to frame treat-ment options based on principles of evidence-basedmedicine (EBM). EBM involves “the conscientious,explicit, and judicious use of current best evidencein making decisions about the care of individualpatients. The practice of evidence-based medicinemeans integrating individual clinical expertise withthe best available external clinical evidence fromsystematic research.”23 Gunther and Diekema2 ac-knowledge that there were few available data toguide them in their analysis of risks, benefits andefficacy of the Ashley Treatment for patients withAshley’s condition. Though data on the risks ofgeneral anesthesia and surgery are available, thedata on the risks, benefits, and efficacy of high-doseestrogen treatment are sketchier. In the end, Ash-ley’s physicians extrapolated from studies on theuse of high dose estrogen treatment with constitu-tionally tall adolescent girls in the 1950s to supporttheir belief that the treatment would be safe andpossibly efficacious.2

It is also interesting to consider the assump-tions of benefit made in light of more recent dataon growth curves for children with moderate tosevere cerebral palsy (CP), a diagnosis that is moreakin to Ashley’s condition than the study group ofnondisabled, tall adolescent girls. The researchers

found that these children with CP, though oftensmaller than their nondisabled peers, tended tohave better health and social participation if theyhad better growth.24 Clearly much is not knownabout the effects of growth on health and socialparticipation for children with significant develop-mental disabilities.

Though in general, medical professionals donot offer treatments where the risks exceed theexpected or known benefits, exceptions do occur.An example might be when a disease process, suchas cancer, may result in death without treatment.There may even be times when some risky treat-ments—such as mastectomy—are initiated pro-phylactically, because the risk of developing breastcancer is believed to be very high (such as for awoman who carries the BRCA genes). The decisionto perform surgery prophylactically vs. increasingthe breast cancer screening protocols is not uni-versally accepted, though, and the fact that awoman (i.e., the decision is not offered to children)can weigh the risk and benefits and make her owndecision based on her values is a critical variable inthe decision making.

There are also examples of prophylactic in-terventions that are widely adopted on behalf ofchildren—like immunization practices. Theseare practices in which the benefits are believed tobe very high (though the actual risk of infectivitymay be low) and the risk from the treatment is alsobelieved to be very low. Ashley does not easily fitinto either category. One could argue that she doesnot have a serious disease that warrants risky med-ical interventions, and as a prophylactic interven-tion, the risk of the Ashley Treatment is substantialand not proportionate to the identified problem.Others may believe, as Ashley’s physician’s con-cluded, that “these risks do not seem to be un-reasonable and are not out of line with the risksof other medical interventions these childrenundergo.”2 (1016 –1017) The final risk/benefit anal-ysis is contingent upon how the problem is de-fined, how the risk is assessed and how alterna-tive options are identified.

Ethical Issues for Stakeholders:Bioethics Committees and/orConsultants

Ethics committees and consultation serviceswere established as advisory bodies to provide aforum for discussion of complex issues in clinicalcare. Since Karen Ann Quinlan, the courts haveasked that hospital-based ethics committees beformed and serve as the primary resource to fam-ilies and health care providers for ethical dilem-mas, rather than the courts. “[This would allow]some much needed dialogue regarding these issuesand [force] the point of exploring all of the options


for a particular patient.”25 It is now a requirementof the Joint Commission for Accreditation ofHealthcare Organizations (JCAHO) that all hos-pitals have some ethics mechanism26 Organiza-tions such as the American Society for Bioethicsand Humanities have worked out recommenda-tions for core competencies for those who provideethics consultation.27 Critical to a robust discus-sion of complex ethics issues is ensuring that the“right” composition of stakeholder perspectives isrepresented. Due diligence—the gathering of per-tinent information and relevant perspectives—isessential to this process.

Disability organizations have raised concernsthat the ethics committee process in the Ashley caseprevented the sort of due process protections that thejudicial system would have provided. Not Dead Yet, adisability activist organization opposed to assistedsuicide and euthanasia, issued a statement condemn-ing the ethics committee and viewed the AshleyTreatment as doing an “end-run” around Ashley’sconstitutionally guaranteed right of due process.28

The Disability Rights Education and Defense Fundbelieves Ashley was denied her “basic human rightsthrough draconian interventions to her person” andthe conduct of the physicians and ethics committeeshould be “widely questioned.”29 Three other groupswho have long-served children with disabilities issuedstatements. The American Association on Intellectualand Developmental Disabilities categorizes the AshleyTreatment as “unjustifiable nontherapy” and askswhy an independent legal advocate or expert in dis-ability rights was not part of the review process.30

United Cerebral Palsy and The Arc, in their jointstatement, believed the counsel on behalf of the eth-ics committee to the parents and physicians was“insufficient . . . since they failed to account for Ash-ley’s fundamental rights as a person with intellectualand developmental disabilities . . . consideration ofAshley and her stake in her own civil and humanrights was absent.”31 The Disability Rights Washing-ton (formerly the Washington Protection and Advo-cacy System) critiqued the Ethics Committee processfurther, agreeing that a court order was legally re-quired before the hysterectomy could be performed.14

As part of the settlement agreement with SeattleChildren’s Hospital they recommended that a disabil-ity rights advocate be included in future deliberationsregarding the Ashley Treatment or growth limitinginterventions for children with developmental dis-abilities.14

Stakeholder representation is critical butthis recommendation is not without complexity.Somebody who has a lived experience of disabil-ity may be able to authentically represent a dif-ferent perspective than the professionals whotypically sit on hospital ethics committees. How-ever, not everybody with a disability thinks alike

or holds the same opinion. How should the dis-ability representative be selected? And can a cog-nitively intact adult with a disability representAshley’s interests any better than her parents orothers on the committee? Certainly, every hos-pital ethics committee must think criticallyabout these questions, the issue of representa-tion, and the group process.

Whereas some have critiqued the ethicscommittee proceedings, others within the bio-ethics community have supported the processthat took place in Washington and laudedDiekema and Gunther for publishing this case.Three bioethicists writing in Scientific Ameri-can.Com (Joel E. Frader, MD, Norman Fost, MD,MPH, and Benjamin Wilfond, MD) “came downfirmly on the side of the parents and the decisionof the original ethics board” with a few reserva-tions.32 Norm Frost writes, “whatever disagree-ments critics may have, it is not possible to saythis decision was made casually or quickly, with-out careful consideration for the relevant factsand arguments.”32 They believe that the child’sbest interests remained the focus of the ethicscommittee’s decision and that going to court forpermission on every invasive procedure involv-ing children would be untenable.

Ethical Issues for All StakeholdersA critical question in the Ashley case is

whether it is ever justified to surgically alter thebody of an otherwise healthy and nonconsentingchild? Quality-of-life criteria were the foundationfor the arguments put forward by the parents andphysicians.1,2 Although there are examples of qual-ity-of-life criteria justifying surgical interventions,such as in esthetic surgery, an important criteria iswho defines the “quality-of-life” goal of treatment.People who elect to have esthetic surgery make thedecision for themselves based on their assessmentof quality-of-life benefits and medical risks. Quality-of-life criteria have also been used in decisions totreat children, such as the use of synthetic growthhormone for boys who are constitutionally short(but not growth hormone deficient), or high-doseestrogen for girls who were constitutionally tall(ironically the only limited experience with thistreatment as applied in the Ashley case), as shortboys and tall girls have been considered sociallyand culturally undesirable.33,34 The use of medicaltreatments for such nonmedical, quality-of-life in-dications have been considered controversial, andin general, have fallen out of favor.

Medical historian, Alice Dreger,35 (p261) whohas studied conjoined twins and people with inter-sex conditions offers the following advice to par-ents contemplating such surgeries:


The use of quality-of-life criteria when makingdecisions on behalf of people who cannot speak forthemselves is further problematic because of thelimited ability of third parties to predict what con-stitutes a “good” quality of life for others.36 Thesocial model of disability would question whetherthe identified problem in this case is intrinsic toAshley’s body or embedded in a lack of environ-mental supports, negative social attitudes and theway others see her or frame her possibilities. Evi-dence for this concern is implicit—both in theparents’ blog and public commentaries, one ofwhich states, “given Ashley’s mental age a nine anda half year old body is more appropriate and moredignified than a fully grown female body.”1 In ap-parent support of this position Ashley’s parentsquote George Dvorsky, a member of the board ofdirectors for the Institute for Ethics and EmergingTechnologies, “The estrogen treatment is not whatis grotesque here. Rather, it is the prospect ofhaving a full-grown and fertile woman endowedwith the mind of a baby.”1

What if the fundamental issue driving the AshleyTreatment is this sentiment—the fact that society isuncomfortable with adults who have cognitive dis-abilities? If so, should health care professionals eversanction or be complicit in using medical treatmentsto alter the bodies of people with disabilities to makethem more acceptable, more lovable, or more com-

fortable to others? And if not, should there be prohi-bitions on certain types of medical treatments forchildren and those who cannot consent for them-selves?

SUMMARYAshley’s treatment undoubtedly encompasses

issues related to her biology, her psychologicalwell-being, her cultural/familial environment andsociopolitical factors. Her impairments, the re-stricted options available to her family, the limits ofmedicine and the history of discrimination againstchildren with cognitive/developmental delays allcome together to make this case particularly poi-gnant and compelling. Given our unique trainingand experience, it is imperative that physiatrists, aswell as other specialists in disability, address theseconcerns and, along with other stakeholders, beintimately involved in this ethics debate.

ACKNOWLEDGMENTSThe authors would like to thank Debjani

Mukherjee PhD, Carmen Cicchetti MA, Elliot J.Roth MD, and Raymond H. Curry MD for theircritical review of the manuscript.

REFERENCES1. The “Ashley Treatment”: Toward a better quality of life for

“Pillow Angels.” Available at: http://ashleytreatment.spaces.live.com. Accessed March 6, 2007

2. Gunther DF, Diekema DS: Attenuating growth in childrenwith profound developmental disability. Arch Pediatr Ado-lesc Med 2006: 1013–7

3. Savage TA, Gill CJ, Kirschner KL: (Special section intro-duction on disability ethics. J Clin Ethics 2004;15:256–63

4. Albrecht GL, Seelman KD, Bury M: The Formation of Dis-ability Studies. In Albrecht GL, Seelman, KD, Bury M, eds.Handbook of Disability Studies. Thousand Oaks, California:Sage Publications;2001: 1–8

5. Oliver M: The Politics of Disablement. London: Macmillan;1990

6. Shakespeare T: Disability Rights and Wrongs. London:Routledge; 2006

7. United Nations General Assembly: Convention on theRights of Persons with Disabilities. Available at: http://www.un.org/disabilities/convention/conventionfull.shtml. AccessedMarch 6, 2007

8. Committee on Ethics, American College of Obstetrics andGynecology: Sterilization of Women, including those withMental Disabilities. Available at http://www.acog.org/from_home/publications/ethics/ethics056.pdf. Accessed March 6,2007

9. Kopell BH, Rezai AR: Psychiatric neurosurgery: a historicalperspective. Neurosurg Clin N Am 2003:181–97

10. Krugman S: The Willowbrook hepatitis studies revisited:ethical aspects. Rev Infect Dis 1986;8:157–62

11. American Academy of Pediatrics Committee on Bioethics:Ethics and the care of critically ill infants and children.Pediatrics 1996;98:149–52

12. American Academy of Pediatrics Committee on Bioethics:Religious objections to medical care. Pediatrics 1997;99:279–80

13. Committee on Bioethics: Sterilization of minors with de-velopmental disabilities. Pediatrics 1999;104:337–40

Make sure you ask yourself and your surgeon

more than once: Given the risk, should you

ever choose, on behalf of someone else (even

your own child), a surgery offered purely for

psychological and social reasons? Surgeons

often tell me ‘you can’t change society so you

have to change children’s bodies,’ but clearly in

many cases we do try to change society. We’ve

tried to eliminate sexism that tries to keep girls

down just because their bodies are different

than boys, we’ve tried to eliminate racism that

keeps some kids down just because their bodies

are darker than others. So consider the

possibility that, instead of changing your

child’s body, we should start expecting others

to change their minds.


14. Carlson DR, Dorfman DA: Investigative report regarding the“Ashley Treatment.” Washington Protection & Advocacy Sys-tem, 2007. Available at: http://www.disabilityrightswa.org/news-1/Investigative%20Report%20Regarding%20the%20Ashley%20Treatment.pdf

15. Miller PS: Toward truly informed decisions about appear-ance-normalizing surgeries, in Parens E (ed): SurgicallyShaping Children. Baltimore, Johns Hopkins UniversityPress, 2006, pp 211–26

16. President’s Commission for the Study of Ethical Problemsin Medicine and Biomedical and Behavioral Research: De-ciding to Forego Life-Sustaining Treatment. Washington,DC: U. S. Government Printing Office, 1983, pp 214–29

17. Ross LF: Children, Families, and Health Care Decision-Making. New York, Clarendon Press, 1998

18. Kopelman LM: The best-interests standard as threshold,ideal, and standard of reasonableness. J Med Phil 1997;22:271–89

19. Engel G: The need for a new medical model. Science 1977;196:129–36

20. Albrecht GL, Devlieger PJ: The disability paradox: highquality of life against all odds. Soc Sci Med 1999;48:977–88

21. Blaymore Bier J, Liebling JA, Morales Y, Carlucci M: Par-ents’ and pediatricians’ views of individuals with meningo-myelocele. Clin Pediatr 1996;35:113–7

22. Gerhart KA, Koziol-McLain J, Lowenstein SR, WhiteneckGG: Quality of life following spinal cord injury: knowledgeand attitudes of emergency care providers. Ann Emerg Med1994;23:807–12

23. Sackett DL, Rosenberg WMC, Muir Gray JA, et al: Evidencebased medicine: what it is and what it isn’t. BMJ 1996;312:7023

24. Stevenson RD, Conaway M, Chumlea WD, et al: Growth andhealth in children with moderate-to-severe cerebral palsy.Pediatrics 2006;118:1010–7

25. In re Quinlan, 70 N.J. 10, 355 A. 2d 647, cert denied, 429U.S. 922 (1976)

26. Joint Commission on Accreditation of Healthcare Organizations:

Comprehensive Accreditation Manual for Hospitals: The Offi-cial Handbook. Oakbrook Terrace, IL, Joint Commission onAccreditation of Healthcare Organizations, 2006

27. American Society for Bioethics and Humanities: Core Com-petencies for Health Care Ethics Consultation. Glenview,American Society for Bioethics and Humanities, 1998

28. Not dead yet: statement on “growth attenuation” experi-mentation. Available at: http://www.notdeadyet.org/docs//Growth_AttenuationPR0107.html. Accessed March 10,2007

29. Disability Rights Education & Defense Fund: Modify thesystem, not the person. Available at: http://www.dredr.org/news/ashley.shtml. Accessed March 6, 2007

30. Board of Directors of the American Association on Intellec-tual and Developmental Disabilities: Unjustifiable non-therapy: a response to Gunther & Diekema (2006), and tothe issue of growth attenuation for young people on thebasis of disability. Available at: http://www.aaidd.org/Policies/board_positions/growth.shtml. Accessed March 6, 2007

31. UCP & The Arc joint statement on “Ashley’s Treatment.”Available at: http://www.ucp.org/ucp_generaldoc.cfm/1/9/10020/10020-10020/7108?PrintFriendly�Yes. Accessed March8, 2007

32. Mims C: The pillow angel case—three bioethicists weigh in.Scientific American Com January 5, 2007

33. Gill DG: “Anything you can do, I can do bigger?”: the ethicsand equity of growth hormone for small normal children.Arch Dis Child 2006;91:270–2

34. Barnard ND, Scialli AR, Bobela S: The current use of estro-gens for growth-suppressant therapy in adolescent girls.J Pediatr Adolesc Gynecol 2002;15:23–6

35. Dreger AD: What to expect when you have the child youweren’t expecting, in Parens E (ed): Surgically ShapingChildren. Baltimore, Johns Hopkins University Press, 2006,pp 211–26

36. Kothari S, Kirschner KL: Abandoning the golden rule: theproblem with putting ourselves in our patient’s place. TopStroke Rehabil 2006;13:68–73

CME Self-Assessment Exam

AnswersAmerican Journal of Physical Medicine & RehabilitationVol. 86, No. 12 • December 2007

CME Article Number 2:J. A. Saveika, et al.

1. B

2. A

3. A

4. A

5. D

December 2007 CME Self-Assessment Exam 1029

Exacerbation of Habitual Dislocation ofUlnar Nerve by Concurrent Dislocationof Triceps Muscle: Complementary Roleof Dynamic Ultrasonography toElectrodiagnosisMyung Jae Yoo, MD, Dennis DJ Kim, MD,and Mooyeon Oh-Park, MDFrom the Department of Rehabilitation Medicine, AlbertEinstein College of Medicine, Montefiore Medical Center,Bronx, New York.

0894-9115/07/8612-1030/0American Journal of Physical Medicine & RehabilitationCopyright © 2007 by Lippincott Williams & WilkinsDOI: 10.1097/PHM.0b013e31815b7d07

A 50-yr-old woman with past medical history of leftelbow fracture in childhood presented with 3 mos ofnumbness and tingling of the left ring and little fingers.Physical examination was remarkable for cubitus varusdeformity, atrophy and weakness of the left hand intrinsicmuscles, and a positive Tinel sign over the left retrocon-dylar groove. During elbow flexion, snapping of a cordlikestructure was noted bilaterally over the medial epicondyle.

Electrodiagnosis demonstrated markedly slow con-duction velocity of the left ulnar nerve across the retro-condylar groove. Fibrillation potentials and polyphasic mo-tor unit action potentials were noted in the left first dorsalinterosseous, abductor digiti quinti, flexor digitorum pro-fundus, and flexor carpi ulnaris muscles. These findingswere compatible with entrapment neuropathy of left ulnarnerve at the retrocondylar groove.

To evaluate the “snapping” structure over the medialepicondyle, dynamic ultrasonography was performed using abroadband linear transducer with medium frequency of 10MHz (LogiqE, General Electric, Inc). The left ulnar nerve wasmarkedly enlarged and hypoechoic compared with the con-tralateral side (Fig. 1). At 70 degrees of elbow flexion, dislo-cation of the left ulnar nerve over the medial epicondyle wasdemonstrated; at 90 degrees of elbow flexion, sonographicimaging revealed concomitant dislocation of the triceps mus-cle pushing the ulnar nerve and producing an appreciablesnap (Fig. 2). Ultrasonography of the contralateral elbowrevealed habitual dislocation of the ulnar nerve, without clin-ical or electrodiagnostic findings of nerve dysfunction. On thebasis of the electrodiagnostic and ultrasonographic findings,snapping triceps syndrome with resulting ulnar nerve entrap-ment at the elbow was diagnosed. The patient was referred for

operative management, including transposition of the ulnarnerve and medial triceps, as well as corrective valgus osteot-omy of the distal humerus.

A transient snapping sensation during elbow flexion canbe caused by either dislocation of the ulnar nerve or themedial portion of the triceps muscle.1 In this case, a palpablesnap over the medial epicondyle was initially interpreted as ahabitual dislocation of the ulnar nerve. Dynamic ultrasonog-raphy differentiated concurrent dislocation of the tricepsmuscle as well as the ulnar nerve on the symptomatic sidefrom isolated dislocation of the ulnar nerve on the asymp-tomatic side.2 This case illustrates the complementary role ofultrasonography to electrodiagnosis in evaluating underlyinganatomic structures of entrapment neuropathy such as snap-ping triceps syndrome, posterior interosseous nerve syn-drome, or tarsal tunnel syndrome.REFERENCES1. Spinner RJ, Goldner RD: Snapping of the medial head of the triceps

and recurrent dislocation of the ulnar nerve. Anatomical and dynamicfactors. J Bone Joint Surg Am 1998;80:239–47

2. Jacobson JA, Jebson PJ, Jeffers AW, et al: Ulnar nerve dislocation andsnapping of triceps syndrome: diagnosis with dynamic sonography—report of three cases. Radiology 2001;220:601–5

All correspondence and requests for reprints should be addressed to MooyeonOh-Park, MD, Department of Rehabilitation Medicine, Albert Einstein College ofMedicine, Montefiore Medical Center, 111 E 210th Street, Bronx, NY 10467.

FIGURE 1 Ultrasonography of the left elbow in extended posi-tion. Markedly enlarged and hypoechoic ulnar nerve(n) in retrocondylar groove. Medial epicondyle (m).

FIGURE 2 Ultrasonography of the left elbow in flexed posi-tion. Ulnar nerve (n) and medial head of tricepsmuscle (t) are dislocated over the apex of medialepicondyle (m).

VISUAL VIGNETTE

1030 Yoo et al. Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

LETTERS TO THE EDITOR

Re: Introduction to Nanotechnology: PotentialApplications in Physical Medicine andRehabilitation

To the Editor: We find the article titled “Introductionto nanotechnology: potential applications in physical med-icine and rehabilitation” by Gordon and colleagues (March2007) interesting and timely.1 Scientist now attempt tomanipulate matter at the scale of atoms and molecules,using the term nanotechnology to revolutionize manufac-turing across all industry sectors, including tissue engi-neering, drug delivery, imaging, diagnostics, surface tex-turing, and biointerfaces that could impact the practice ofphysical medicine and rehabilitation in the future. Gov-ernments and conglomerates must establish strict codes ofethics for nanotech products to avoid future mismanage-ment of this novel technology. Lack of appropriate ethicalcodes of conduct in this science may lead to disasters tomankind as well as nature. Perhaps it is time to rememberthat natural organisms harbor excellent nanotech protec-tion shaped by millions of years of natural selection—lotusleaves possess micro- and nanotextures on the surface.Similarly, animals from bees to birds have unique nano-texturing that engenders water repellency and lack of ad-hesion. Prehistoric cave painters apparently used nature’snano materials in their painting processes.

Globally, industry and governments spend nearly $10billion on nanotech research and development annually(Table 1). The National Science Foundation in the UnitedStates predicts that nanoscale technology will capture a $1trillion market by 2011. Thousands of tons of nanomate-rials are already being produced each year, and more than300 claimed nanotechnology products have surfaced onthe market.2 Nano particles like nano paint come when thegeneral public has great concern regarding geneticallymodified organisms and toxic chemical residues in humanand pet foods. Thus, the public in developed and develop-ing nations may undoubtedly demand from their govern-ments more transparency and safer nanotechnology.3

Unfortunately, there are no ethical and professional regu-lations for large- or small-scale nanotech industries, andno impact studies have been mandated to assess exposure

to engineered nanomaterials in air, water, and terrestrialecosystems. Because nanotechnologies are diverse and ex-posures to nonmaterial will vary widely, assessing exposureand potential impacts on health or the environment willrequire a multidisciplinary approach for assessments.

Among the large array of nanomaterials and processes,the quantum dots are of particular importance for theirpossible therapeutic and diagnostic medical applications.These nanoparticles consist of a metalloid core and a shellthat surrounds the core and renders the quantum dots’bioavailability. The further addition of a specific coatingcan give the quantum dots a desired bioactivity, which isessential for promising medical applications such as tar-geted drug delivery and in vivo biomedical imaging. Hu-man exposure to quantum dots may result from environ-mental, workplace, and therapeutic exposure. There maybe a risk of bioaccumulation of these materials withinorgans and tissues (e.g., in lungs) with still unexplored orunderexplored health risks. Introduction of quantum dotsinto the environment may occur via waste streams fromindustry, research, and clinical settings. The persistence ofthese materials in the environment may be quite long, andthe environmental exposure will depend on the partition ofquantum dots between water, air, and various soil types.4

Particular impacts of nanomaterials are foreseen on theecology of soils, which often already contain and movearound small particles. Nanoparticles could easily be ab-sorbed by earthworms, possibly allowing them to move upthe food chain and reach humans.5

Global understanding of nanotechnology-specific risksis essential if large and small industries are to operate ona level playing field and if developing economies are not tobe denied essential information on designing safe nano-technologies. The ethical issues fall into the areas of eq-uity, privacy, security, environment, and metaphysicalquestions concerning human–machine interactions. Awide debate is needed on the implications and democraticcontrol of nanotechnologies worldwide, to restrict nano-technology before it becomes a lawless science. The debateshould not be restricted to meetings of expert scientists ortargeted on human/environmental health and safety is-sues. Ethical implications of nanotechnology must be ad-dressed by people from all walks of life.

Govindasamy AgoramoorthyDepartment of Pharmacy

Tajen UniversityYanpu, Taiwan

Chiranjib ChakrabortyDepartment of Biotechnology

College of Engineering & TechnologyGreater Noida, India

TABLE 1 Government R&D spending innanotechnology ($U.S., in millions)

Region 2001 2005

European Union �225 �1050Japan �465 �950United States �465 1,081Others �1535 �4081

Source: National Nanotechnology Initiative (United States).

December 2007 Letters to the Editor 1031

REFERENCES1. Gordon AT, Lutz GE, Boninger ML, Cooper RA: Introduction to

nanotechnology: potential applications in physical medicine and reha-bilitation. Am J Phys Med Rehabil 2007;86:225–41

2. The Nanotechnology Consumer Products Inventory. Washington, DC,Woodrow Wilson International Center for Scholars, 2006

3. Maynard AJ, Aitken RJ, Butz T, et al: Safe handling of nanotechnology.Nature 2006;444:267–9

4. Hardman RA: A toxicological review of quantum dots: toxicity dependson physico-chemical and environmental factors. Environ Health Per-spect 2006;114:165–72

5. Brumfiel G: Nanotechnology: a little knowledge. Nature 2003;424:246

DOI: 10.1097/01.phm.0000297451.49425.9d

Response Letter

The Authors Respond: We thank Drs. Agoramoorthyand Chakraborty for their letter in which they express con-cern for safety and regulation of nanotechnology and the needfor shared international information. In our article (March2007), we raise safety issues by citing known toxicity ofnanoparticles to animals in reference citations 122–126.1 Weagree with Drs. Agoramoorthy and Chakraborty about theimportance of environmental impact studies, and we areencouraged by the initial steps taken to address these issues(references 127 and 128 in our article).

The issue of nanotechnology regulation has receiveda great deal of international attention among materialsscientists, legal scholars, and government agencies. Thereader is referred to several review articles that carefullydiscuss the sharing of resources and expertise, companyprotectionism, trade conflicts, and models for transna-tional regulation based on international agreements onenvironmental pollutants, arms proliferation, and globalethics.2–5 Substantive progress will be limited, however, byan inadequate framework for discussion and implementa-tion of policy, combined with a scientific uncertainty re-garding toxicity and bioaccumulation of nanomaterials. Ofnote, the first law regarding nanotechnology regulationhas already been instituted in the United States. In December,2006, the city of Berkeley amended its hazardous materialslaw to include nanostructures, requiring researchers andmanufacturers to report what materials they are workingwith and how they are handling their disposal. The “BerkeleyNanotechnology Regulation” represents an initial step anddemonstrates that oversight is achievable.

Assaf T. Gordon, MDDepartment of Physical Medicine and Rehabilitation

Rusk Institute, New York UniversityNew York, New York

REFERENCES1. Gordon AT, Lutz GE, Boninger ML, Cooper RA: Introduction to

nanotechnology: potential applications in physical medicine and reha-bilitation. Am J Phys Med Rehabil 2007;86:225–41

2. Marchant GE, Sylvester DJ: Transnational models for regulation ofnanotechnology. J Law Med Ethics 2006;34:714–25

3. Renn O, Roco MC: Nanotechnology and the need for risk governance.J Nanoparticle Res 2006;8:153–91

4. Mehta MD: From biotechnology to nanotechnology: what can we learnfrom earlier technologies? Bull Sci Tech Soc 2004;24:34–9

5. Reynolds GH: Nanotechnology and regulatory policy: three futures.Harv J Law Technol 2003;17:179–205

DOI: 10.1097/01.phm.0000297450.49425.d4

Re: Sacroiliac Joint Pain: Anatomy,Biomechanics, Diagnosis, and Treatment

To the Editor: Foley and Buschbacher1 are to be com-mended on their informative as well as thorough review ofthe sacroiliac joint (SIJ) as a pain generator (December 2006).As they readily acknowledge, “the literature on SIJ pain iscomplex and controversial.” Three decades ago, when weevaluated the SIJ as a source of lumbosacral and pelvic pain,it was also underappreciated, as it remains so even today.2

Although Dreyfuss et al.3 state in their review “that nophysical examination test [has] demonstrated diagnosticvalue and that SIJ pain is resistant to identification byhistorical and physical examination,” our experience inthis regard would suggest otherwise. Although we wouldagree there is no one clinical or laboratory test that ispathomonic of SIJ dysfunction, an appropriate history andphysical examination buttressed by dynamic pelvic x-raysand an immediate reduction in pain in response to ananesthetic injection of the joint can be diagnostic. In thesupine position, an asymmetric restriction of hip externalrotation on the symptomatic side associated with increasedpain over the SIJ as it is passively rotated into abduction isalso a frequent clinical finding (Fig. 1). A standing pelvicx-ray followed by alternate-leg weight bearing may also dem-onstrate excessive instability of the pubic symphysis, oftenexceeding 3 mm (Fig. 2). Usually, this slip is greater on theside of the symptomatic SIJ. Symphyseal instability mirrorssimilar rotary luxation at the symptomatic SIJ as this artic-ulation responds to shifts in the center gravity horizontally aswell as vertically during normal ambulation (Fig. 3).

An immediate reduction in SIJ pain after a combinedinjection of a steroid and Xylocaine is also diagnostic.

FIGURE 1 Asymmetric restriction of hip external rotation onthe symptomatic side is associated with increasedpain over the sacroiliac joint as it is passivelyrotated into abduction.

1032 Letters Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

Except in the presence of extreme obesity, this procedurecan easily be accomplished without x-ray guidance. En-trance to the articulation itself can be assured by notingrestricted side-by-side movement of the syringe as limitedby the congruent joint surfaces As a caveat, it is importantto note that SIJ may not be the primary pain generator.Instead, it may be a referred site of discomfort, both via itsdirect innervation and secondary to more remote mechan-ical problems—among others, a degenerative disc atL5–S1 and/or spinal stenosis, as well as degenerative dis-ease of the hip.

Myron M. LaBan, MD, MMSc, FACP, FAAPMRWilliam Beaumont Hospital

Royal Oak, Michigan

REFERENCES1. Foley BS, Buschbacher RM: Sacroiliac joint pain: anatomy, biomechanics,

diagnosis and treatment. Am J Phys Med Rehabil 2006;85:997–1006

2. LaBan MM, Meerschaert JR, Taylor RS: Symphysis and sacroiliac jointpain associated with pubic symphysis instability. Arch Phys Med Re-habil 1978;56:470–2

3. Dreyfuss P, Michaelsen M, Pauza K, McLarty J, Bogduk N: The value of

medical history and physical examination in diagnosing sacroiliacjoint pain. Spine 1996;21:2594–602

DOI: 10.1097/01.phm.0000297443.85757.8a

Response Letter

The Authors Respond: The authors appreciate Dr.LaBan’s thoughtful comments. We agree that although nosingle physical exam test is particularly predictive, a panelof three or more tests can be extremely helpful.

The tests described by Dr. LaBan may prove useful tothose examining patients with sacroiliac joint pain. Hopefully,further research will be performed to test those hypotheses.They make sense to the authors, but they have not yet beenvalidated in the literature.

Dr. LaBan’s caveat regarding other regional pain gen-erators that masquerade as SIJ pain is a good one thatshould be heeded by all practicing clinicians.

Brian S. Foley, MDRalph M. Buschbacher, MD

Department of PM&RIndiana University School of Medicine

Indianapolis, Indiana

DOI: 10.1097/01.phm.0000297442.78134.fe

Re: The Myofascial Trigger Point Region:Correlation Between the Degree of Irritabilityand the Prevalence of Endplate Noise

To the Editor: The article “The myofascial triggerpoint region: correlation between the degree of irrita-bility and the prevalence of endplate noise” by Kuan etal.1 (March 2007) is a valuable addition to the currentbody of knowledge surrounding myofascial pain. Theauthors suggest a possible mechanism of action pro-posed originally by Simons et al.2 in which hypoxia andischemia are related to local release of inflammatorysubstances, which may sensitize muscle nociceptors. Wewould like to add that our recent study3 in 2005 alsosupports this contention.

We used a novel microdialysis technique to samplemuscle with and without myofascial trigger points forlevels of protons, inflammatory mediators, cat-echolamines, neuropeptides, and cytokines. The resultsshow distinct local biochemical differences between sub-jects with a spontaneously painful myofascial trigger point(active MTrP) in the upper trapezius muscle comparedwith those with a nonpainful trigger point (latent MTrP)and those without a trigger point. Analysis for pH demon-strated that active MTrPs possessed relatively higher acidiclevels (lower pH) than the other muscle groups, whichcould be associated with conditions attributed to hypoxia andischemia. Similarly, levels of bradykinin, serotonin, norepineph-rine, substance P, CGRP, TNF-�, and IL-1� were all at relativelyhigher concentrations in active MTrPs.

FIGURE 3 Diagram showing rotary luxation at the symptom-atic sacroiliac joint during normal ambulation.

FIGURE 2 Standing pelvic x-ray showing excessive instabil-ity of the pubic symphysis.


We feel that the combination of these two studiesprovides good support to the proposed mechanism of myo-fascial trigger point pain.

Jay P. Shah, MDRehabilitation Medicine Department

National Institutes of HealthBethesda, Maryland

Sagar Parikh, MS IVUniversity of Medicine and Dentistry of New Jersey

New Brunswick, New Jersey

Jerome Danoff, PT, PhDDepartment of Exercise Science

George Washington University Medical CenterWashington, DC

Lynn H. Gerber, MDCollege of Health and Human Services

George Mason UniversityFairfax, Virginia

REFERENCES1. Kuan T-S, Hsieh Y-L, Chen S-M, Chen J-T, Yen W-C, Hong C-Z: The

myofascial trigger point region: correlation between the degree ofirritability and the prevalence of endplate noise. Am J Phys MedRehabil 2007;86:183–9

2. Simons DG, Travell JG, Simons LS: Travell & Simons’s MyofascialPain and Dysfunction: The Trigger Point Manual, vol 1, ed 2. Balti-more, Williams & Wilkins, 1999

3. Shah JP, Phillips TM, Danoff JV, Gerber LH: An in vivo microanalyticaltechnique for measuring the local biochemical milieu of human skel-etal muscle. J Appl Physiol 2005;99:1977–84

DOI: 10.1097/01.phm.0000297445.62886.08

Response Letter

The Authors Respond: We highly appreciate the com-ments by Shah et al. We also regret that we have not cited theirpaper1 as one of the references in our article.2 Both their work1

and ours2 have definitely contributed substantially to increas-ing the credibility of myofascial trigger point (MTrPs). Theformer has further confirmed the energy crisis hypothesis,3

and the latter has proven the correlation between MTrP andendplate noise.4,5 The great pioneer in MTrP, David G.Simons, developed these two important theories more than10 yrs ago.

Chang-Zern Hong, MDDepartment of Physical Therapy

HungKuang UniversitySha-Lu, Taichung, Taiwan

Ta-Shen Kuan, MDDepartment of Physical Medicine and Rehabilitation

National Cheng Kung UniversityTainan, Taiwan

REFERENCES1. Shah JP, Phillips TM, Danoff JV, Gerber LH: An in vivo microanalytical

technique for measuring the local biochemical milieu of human skel-etal muscle. J Appl Physiol 2005;99:1977–84

2. Kuan T-S, Hsieh Y-L, Chen S-M, Chen J-T, Yen W-C, Hong C-Z: Themyofascial trigger point region: correlation between the degree ofirritability and the prevalence of endplate noise. Am J Phys MedRehabil 2007;86:183–9

3. Simons DG: Myofascial pain syndrome due to trigger points, in GoodgoldJ (ed): Rehabilitation Medicine. St. Louis, Mosby, 1988, pp 686–723

4. Simons DG, Hong C-Z, Simons LS: Prevalence of spontaneous elec-trical activity at trigger spots and at control sites in rabbit skeletalmuscle. J Musculoskelet Pain 1995;3:35–48

5. Simons DG: Clinical and etiological update of myofascial pain fromtrigger points. J Musculoskelet Pain 1996;4:93–121

DOI: 10.1097/01.phm.0000297444.62886.41

Re: Characteristics and Treatment ofHeadache After Traumatic Brain Injury

To the Editor: The authors of “Characteristics andTreatment of Headache After Traumatic Brain Injury” (Lewet al.,1 July 2006) are to be complimented on their com-prehensive review of “posttraumatic headache (PTH).”They readily acknowledge the present confusion inherentin both the “accepted terminology” and the exact patho-genesis of “cerviogenic headache” pathology when bothintracranial and cervical spine trauma have occurred si-multaneously.

Although controversy in this regard persists, in ourown 40-yr clinical experience, posttraumatic occipital neu-ralgia (PTON) is all too often the principle causal agent incases of persistent PTH. The discomfort of PTON is locatedprimarily in the occipital area, often radiating to the vertexas a direct consequence of injury to the greater and/orlesser occipital nerves at their C1–C2 spinal roots of origin.

When digital pressure is applied to these nerves overthe occipital notch, the patient may experience eitherhyper- and/or hypoesthesia in the occipital nerve’s distri-bution. Thereafter, an occipital nerve block can be bothdiagnostic and therapeutic in confirming the origin of theheadache by alleviating the symptoms of the PTON.

A 25-gauge needle in this procedure is inserted 3.5 cminferolaterally to the occipital protuberance, and a combi-nation of both bupivacaine hypochloride (Marcaine) and acorticosteroid (1 ml of each) is injected. In a review of tenPTH patients, Lew report a good response to a greateroccipital nerve block in eight patients and, in an additionaltwo patients, a partial response.2

Additional symptoms often accompanying PTON in-clude, among others, a “light and sound hypersensitiv-ity,” as well as vertigo, nausea, and facial hyperesthesia.At the upper spinal cord levels, the fibers of the descend-ing tract of CN V, accompanied by the tracts of CN VII,IX, and X, interdigitate with the ganglion cells of CI andCII. Subsequently, these tracts cross the midline to-gether and progress retrogradely as the ascending tractof CN V. At the C1–C2 spinal levels, this tract is vulner-able to trauma, with the potential of inciting brainstemsymptoms.

As the authors themselves conclude, “headaches aremultifactorial and involve a combination of central andperipheral mechanisms.” We would agree that a definitivediagnosis as related to intracranial and/or extracranial pa-thology remains paramount to efficient and appropriatetreatment.

Myron M. LaBan, MD, MMScRoyal Oak, Michigan


REFERENCES1. Lew HL, Lin P-H, Fuh J-L, Wang SJ, Clark DJ, Walker WC: Charac-

teristics and treatment of headache after traumatic brain injury. Afocused review. Am J Phys Med Rehabil 2006;85:619–27

2. Hecht JS: Occipital nerve blocks in posterior concussive headaches: arespective review and report of ten patients. J Head Trauma Rehabil2004;19:58–71

DOI: 10.1097/01.phm.0000297448.64672.d1

Response Letter

The Authors Respond: The authors appreciate the in-sightful comments by Dr. Myron M. LaBan regarding ourrecent article on headache after traumatic brain injury (TBI).1

Incidentally, after publication of this manuscript, we havereceived feedback from private practitioners of different spe-cialties, as well as academicians, who have echoed our per-ception that most headaches involve a combination of centraland peripheral mechanisms. They also share our observationsregarding the heterogeneity in terminology and the paucity ofwell-controlled studies in this field. The consensus is thatclinicians should perform a comprehensive and systematicevaluation before initiating appropriate treatment. What wepropose in Figure 2 is one possible algorithm that can beexpanded on, according to the local resources and level ofexpertise of the individual clinician. As described by Dr. La-Ban, occipital nerve block is one example of a dual diagnosticand therapeutic approach that is relatively time-efficient toperform and that could prove to be useful to patients. Like allother methods we have alluded to, this procedure deservesformal evaluation in the specific context of headache afterTBI. Retrospective case review and report2 can certainly pro-vide impetus for further studies. Hopefully, with increasinginterest and awareness in the field, more prospective, ran-domized, controlled trials can be conducted to enhance thestandard of care for headache management in patients withTBI.

Henry L. Lew, MD, PhDStanford University School of Medicine

PM&R Service, VA Palo Alto Health Care SystemPalo Alto, California

David J, Clark MD, PhDStanford University School of Medicine

Anesthesiology Service, VA Palo Alto Health Care SystemPalo Alto, California

William C. Walker, MDVirginia Commonwealth University

PM&R Service, McGuire VAMCRichmond, Virginia

REFERENCES1. Lew HL, Lin P-H, Fuh J-L, Wang SJ, Clark DJ, Walker WC: Charac-

teristics and treatment of headache after traumatic brain injury. Afocused review. Am J Phys Med Rehabil 2006;85:619–27

2. Hecht JS: Occipital nerve blocks in posterior concussive headaches: a

retrospective review and report of ten patients. J Head Trauma Rehabil2004;19:58–71

DOI: 10.1097/01.phm.0000297446.70510.d4

Re: Shoulder Impingement Syndrome:Relationships Between Clinical, Functional,and Radiologic Findings

To the Editor: We read with interest the article byArdic and colleagues.1 The authors detected the relation-ship between clinical, functional, and radiologic diagnostic(ultrasonography, magnetic resonance imaging [MRI]) mea-sures in patients with subacromial impingement syndrome(SIS) and compared the diagnostic performances of ultra-sonography and MRI. In this study, MRI has been used as agold standard. However, MRI mainly evaluates the anato-mopathologic conditions in the rotator cuff (RC) and sur-rounding structures. It has also been shown that a significantnumber of asymptomatic shoulders may reveal abnormalinternal signals and even complete ruptures of RC tendons,by MRI.2,3 These tears were increasingly frequent with ad-vancing age and were compatible with normal subjects. MRIevaluation of elder asymptomatic patients (60 yrs of age andabove) showed an incidence of 28% full-thickness tears.3

Frost et al.4 report that age was related to MRI based onsupraspinatus pathology more so than clinical signs of im-pingement.

In addition, Neer5 and Hawkins–Kennedy6 impinge-ment tests are not enough for diagnosis of SIS. These testsmay be positive in other shoulder disorders. SIT should beperformed as a reference test for differential diagnosis ofSIS.5,7–9 Neer5,7 identifies SIT in his reports as a most usefulmethod of separating impingement lesions from other causesof shoulder pain. SIT does not investigate the anatomopatho-logic condition in RC, but it depresses the pain generators inthe subacromial area where the impingement occurs. Obvi-ous relief of pain and almost total improvement in passiveand/or active shoulder range of motion after SIT suggest thatthe test is positive and discriminates SIS from other shoulderpathologies.

As a result, MRI is not the gold standard for diagnosisof SIS. SIS is a clinical diagnosis that is done with SIT andimpingement tests.

Ilknur Aktas, MDDepartment of Physical Medicine and Rehabilitation

Saygi HospitalMarmara University Institute of Neurological Sciences

Istanbul, Turkey

Kenan Akgun, MDDepartment of Physical Medicine and Rehabilitation

Cerrahpasa Medical FacultyUniversity of Istanbul

Istanbul, Turkey

REFERENCES1. Ardic F, Kahraman Y, Kacar M, Kahraman MC, Findikoglu G,

Yorgancioglu ZR: Shoulder impingement syndrome: relationships be-tween clinical, functional, and radiologic findings. Am J Phys MedRehabil 2006;85:53–60

2. Chandnani V, Ho C, Gerharter J, et al: MR findings in asymptomatic


shoulders: a blind analysis using symptomatic shoulders as controls.Clin Imaging 1992;16:25–30

3. Sher JS, Uribe JW, Posada A: Abnormal findings on magnetic reso-nance images of asymptomatic shoulders. J Bone Joint Surg Am1995;77:10–5

4. Frost P, Andersen JH, Lundorf E: Is supraspinatus pathology as definedby magnetic resonance imaging associated with clinical sign of shoul-der impingement? J Shoulder Elbow Surg 1999;8:565–8

5. Neer CS: Impingement lesions. Clin Orthop 1983;173:70–7

6. Hawkins RJ, Abrams JS: Impingement syndrome in the absence ofrotator cuff tear. Orthop Clin North Am 1987;18:373–82

7. Neer CS: Cuff Tears, Biceps Lesions and Impingement. ShoulderReconstruction. Philadelphia, Saunders, 1990, pp 41–142

8. Calis M, Akgun K, Birtane M, Karacan I, Calis H, Tuzun F: Diagnosticvalues of clinical diagnostic tests in subacromial impingement syn-drome. Ann Rheum Dis 2000;59:44–7

9. Birtane M, Calis M, Akgun K: The diagnostic value of magnetic reso-nance imaging in subacromial impingement syndrome. Yonsei Med2001;42:418–24

DOI: 10.1097/01.phm.0000297449.72296.3d

Response Letter

The Authors Respond: Many thanks for your interestto our article.1 First of all, our studies were done by anupper-extremity rehabilitation team consisting of expertsin this area using standardized evaluation and therapysystems/algorithms.

We have to use comprehensive diagnostic and thera-peutic tools for patients’ referral to shoulder rehab, as wementioned before.1 Special shoulder tests commonly per-formed by upper-extremity professionals can be divided forinstability, impingement, labral lesions, scapular stability,muscle tendon pathology, brachial plexus, and thoracicoutlet syndrome.2,3 Common shoulder impingement testsare the Neer–Walsh, Hawkins–Kennedy, and posterior in-ternal impingement tests.3 We used the first and the sec-ond impingement tests because the third is found primar-ily in overhead athletes and workers.3

The positive Neer–Walsh impingement test suggestsrotator cuff (RC) pathology like the positive Hawkins–Kennedy impingement test.2 If one of these tests elicitspain, subacromial impingement test (SIT) with 10 ml of0.1% lidocaine to the subacromial space will eliminate thepain of repeat testing.2 Because of medicolegal issuesand the approval procedure of this study by the IRB, wedidn’t choose the subacromial impingement test to as-sist in the diagnosis. It is not clear that these injectionshave proven useful in directing care toward successfuloutcomes, possibly because of the confounding placebofactor. Actually, Henkus et al.4 report that injections insubacromial bursa were inaccurate, with false-positiveand -negative results in their randomized controlled MRIstudy with level II evidence. We didn’t prefer SIT, because itis invasive; it requires patient’s approval and is only 66%accurate.1,4 Our opinion for the diagnosis of subacromialimpingement syndrome (SIS) was noninvasive clinical andradiologic work-ups.

We selected patients that had no history of trauma andhad shoulder pain of �3 mos, were unresponsive to analgesicmedication after 3 wks, and had a diagnosis of suspected SISwaiting for physical therapy and shoulder rehab.1 It is ridic-ulous to make clinical and radiologic assessments for asymp-tomatic shoulders. We carefully evaluate the patients withchronic shoulder pain demographically, clinically, and radio-logically for the treatment options. Diagnosis and treatmentshould be based on combination of these data.

As we state in our article, “MRI has been used as thegold standard because our patients did not want arthro-scopic evaluation.” In addition, MRI has been accepted asthe “best imaging tool” in many shoulder problems, in-cluding RC, labrum, capsule, biceps tendon/anchor, osse-ous structures, arthritis, and neural impingement.5,6 It iswell known that MRI is the only technique that can accu-rately image the early changes of this condition, in partic-ular bursal thickening and effusion (subacromial bursitis),edema, and inflammatory changes of the rotator cuff andits tendons. SIS is a clinical diagnosis, and MRI provides uswith critical information regarding the size and location ofa tear, the specific tendons involved, the degree of muscu-lar atrophy and tendon retraction, and the quality of thetorn edges.5,6

Neer’s7 impingement concept depicts the syndrome asa lifelong process with three stages. Stage I impingementconsists of reversible edema and hemorrhage typically inactive patients �25 yrs. Stage II implies fibrosis and ten-donitis. Stage III represents degeneration and rupture of-ten associated with osseous changes most commonly inpatients �40 yrs.5 Therefore, early diagnosis and treat-ment are critical to prevent progression and improveshoulder function. Frequently, however, clinical signs andsymptoms are nonspecific, and the diagnosis is often de-layed until a full-thickness defect in the RC has developed.6

Symptomatic elderly patients may have stage III SISmainly treated by shoulder rehab. However, asymptomaticindividuals are not our interest area.

Finally, there were very limited numbers of random-ized controlled trials for SIS using clinical and MRIvariables. MR/MR arthrography are still the best imag-ing tools in many shoulder problems. MRI has proved tobe highly sensitive and accurate for the diagnosis ofstage II–III SIS, with excellent correlation of arthro-scopic and/or miniopen surgical findings. Specified clin-ics in shoulder rehab should use comprehensive diag-nostic and therapeutic work-ups.

F. Figen Ayhan, MDZ. Rezan Yorgancioglu, MD

Ankara Education and Research HospitalDepartment of PM&R

Division of Hand and Upper Extremity RehabAnkara, Turkey


REFERENCES1. Ardic F, Kahraman Y, Kacar M, et al: Shoulder impingement

syndrome: relationships between clinical, functional, and radiologicfindings. Am J Phys Med Rehabil 2006;85:53–60

2. Braddom RL: Physical Medicine & Rehabilitation. China, SaundersElsevier, 2007

3. Magee DJ: Orthopedic Physical Assessment. Philadelphia, Saunders,2002

4. Henkus HE, Cobben LP, Coerkamp EG, et al: The accuracy of sub-

acromial injections: a prospective randomized MRI study. Arthroscopy2006;22:277–82

5. Stoller D, Tirman P, Bredella M, et al: Diagnostic Imaging Orthopae-dics. Utah, Amirsys, 2004

6. Greenspan A: Orthopedic Imaging: a practical approach. Philadelphia,Lippincott Williams & Wilkins, 2004

7. Neer CS II: Impingement lesions. Clin Orthop 1983;173:70–7

DOI: 10.1097/01.phm.0000297455.15615.5e


Erratum

In the article by Ersoz et al., published in the September 2007 issue of the American Journal of PhysicalMedicine & Rehabilitation, Tables 1 and 4 should appear as follows:

REFERENCE1. Ersoz M, Ulusoy H, Oktar MA, Akyuz M: Urinary tract infection and bacteriuria in stroke patients: Frequencies, pathogen

microorganisms, and risk factors. Am J Phys Med Rehabil 2007;86:734–41

TABLE 1. Comparison of the stroke subgroups with respect to demographic and clinicparameters for the frequencies of symptomatic urinary tract infection

Factor Frequency (%) P

Age, yrs 0.067�65 11/56 (19.6)�65 19/54 (35.2)

Gender 0.147Male 10/49 (20.4)Female 20/61 (32.8)

Level of education 0.055Ungraduated 18/46 (39.1)Primary school 9/44 (20.5)Secondary/high school or university

Type of lesion 0.858Hemorrhagic 8/28 (28.6)Ischemic 22/82 (26.8)

Side of lesion 0.354Right 6/29 (20.7)Left 24/81 (29.6)

Bladder emptying method 0.041Spontaneous 23/96 (24.0)Indwelling catheter

Residual urine, �50 ml 0.039Absent 15/77 (19.5)Present 8/19 (41.2)

Functional ambulation class 0.147�2 20/61 (32.8)�2 10/49 (20.4)

Brunnstrom stage class of upper extremity 0.0121––2 23/59 (39.0)3––4 3/26 (11.5)5––6 4/25 (16.0)

Brunnstrom stage class of lower extremity 0.0551––2 16/39 (41.0)3––4 9/44 (20.5)5––6 5/27 (18.5)

TABLE 4. Microorganisms isolated in urine (�105 CFU/ml) and their frequencies in all strokepatients, in patients who are voiding spontaneously, and in patients with indwelling catheters

Spontaneous(%) (n � 32)

Indwelling Catheter(%) (n � 11)

All Patients(%) (n � 43)

Escherichia coli 19 (59.4) 6 (54.5) 25 (58.1)Klebsiella pneumoniae 7 (21.9) 0 (0) 7 (16.3)Pseudomonas aeruginosa 0 (0) 4 (36.4) 4 (9.3)Staphylococcus aureus 3 (9.4) 1 (9.1) 4 (9.3)Proteus mirabilis 3 (9.4) 0 (0) 3 (7.0)

1038 Erratum Am. J. Phys. Med. Rehabil. ● Vol. 86, No. 12

American Journal of Physical Medicine and Rehabilitation · American Journal of Physical Medicine...

Documents

Transcript of American Journal of Physical Medicine and Rehabilitation · American Journal of Physical Medicine...