PEDIATRIC/CRANIOFACIAL

Documentation of the Incidents Associated withMandibular Distraction: Introduction of a NewStratification System

Pradip R. Shetye, D.D.S.,M.D.S.

Stephen M. Warren, M.D.Daniel Brown, M.D.

Judah S. Garfinkle, D.M.D.Barry H. Grayson, D.D.S.

Joseph G. McCarthy, M.D.

New York, N.Y.

Background: This article aims to assess the spectrum of unfavorable events orincidents encountered during mandibular distraction and to evaluate the dif-ference in the incident rates among the following treatment groups: (1) nativebone distraction using an external device, (2) native bone distraction using aninternal device, and (3) grafted bone distraction using an external device.Methods: This retrospective study examined the records of 141 patients treatedby mandibular distraction over a 16-year period. Of the total 141 patients, 56underwent unilateral mandibular distraction and 85 underwent bilateral man-dibular distraction, contributing to a total of 226 sided distraction procedures.The number of procedures performed on native bone using external devices was149, versus 41 internal devices. There were 36 distractions performed on graftedbone with external devices. Incidents were broadly classified into three groupsbased on a severity index. A minor incident was one that resolved satisfactorilywith minimal or no invasive intervention. A moderate incident was one thatresolved satisfactorily with moderate clinical intervention. A major incident wasone that did not resolve or could not be resolved with surgical intervention, andcompromised treatment outcome.Results: The major incident rate was 5.31 percent (total of 226 distractionprocedures). A higher rate of major incidents was observed when distractinggrafted bone. The overall minor incident rate was 26.99 percent and the mod-erate incident rate was 20.35 percent.Conclusion: Mandibular distraction can be considered a safe and predictableprocedure for lengthening/augmenting the mandible in patients with lower jawdeficiencies. (Plast. Reconstr. Surg. 123: 627, 2009.)

Distraction osteogenesis has become a widelyused technique to reconstruct the hypo-plastic mandible since the report of Mc-

Carthy and colleagues in 1992.1 The technique hassubsequently been applied to the maxilla and cra-nium to correct various craniofacial anomalies.Distraction osteogenesis has been established asthe preferred technique in many clinical situa-tions because of the relative safety of the proce-dure, reduced hospital stay, lower incidence ofblood transfusion, elimination of the harvest ofbone graft, and a concomitant expansion of theassociated soft tissue.2 Over the years, the biologyof the distraction zone has been studied exten-

sively in the laboratory and the technique has beenrefined.3 These advances have been characterizedby an improved surgical procedure and modifica-tions of the distraction devices, including trans-formation from more bulky externally placed de-vices to small internally buried devices.

As is true of all procedures, distraction osteo-genesis is not without its incidents. Surgical inci-dents are not necessarily the result of an error insurgical technique or management.4 Several au-thors have reported the incident rates associatedwith mandibular distraction. These reports varyfrom a rate as low as 2.5 percent to as high as 35percent.2,5,6 The large variation in the reporting of

From the Institute of Reconstructive Plastic Surgery, NewYork University Medical Center.Received for publication August 3, 2008; accepted August12, 2008.Copyright ©2009 by the American Society of Plastic Surgeons

DOI: 10.1097/PRS.0b013e3181956664

Disclosure: None of the authors has a financialinterest to disclose in relation to the content of thisarticle.

www.PRSJournal.com 627

incident rate is primarily attributable to the oftenambiguous and sometimes arbitrary definition ofsurgical complications/incidents as traditionallyreported by surgeons. During any surgical proce-dure and the associated perioperative period, thesurgeon may encounter various events that rangein severity from cardiac arrest to a localized sutureinfection. The management (and outcome) ofsuch clinical problems can differ widely. Some ofthe unintended events may not lead to adverseoutcomes but do merit documentation to be ableto assess the overall potential risk of the surgicalprocedure and the associated increase in healthcare expenses. In addition, one report5 was basedon a survey of many authors, an inaccuratemethod of documenting perioperative incidents.Several authors have endeavored to classify surgi-cal incidents based on different grading systems.7–9

Many of the reported classification system areelaborate and complex and ultimately difficult toapply in the day-to-day clinical setting.

In this article, the authors aim to assess thespectrum of incidents encountered during man-dibular distraction and to evaluate the differencein the incident rates between distraction osteo-genesis in the following treatment subgroups: (1)native mandibular bone distraction using an ex-ternal device, (2) native bone distraction using aninternal device, and (3) grafted bone distractionusing an external device. This article also ad-dresses the management of these problems andprecautions that are necessary to prevent them.

PATIENTS AND METHODSThis retrospective study examined records of

patients treated for mandibular distraction at theNew York University Institute of ReconstructivePlastic Surgery over a 16-year period (1989 to2005). All patients with unilateral and bilateralmandibular distraction were included, but pa-tients treated for mandibular midline and trans-port distraction were excluded from this study.One hundred forty-one patients formed the sam-ple for this study; there were 81 men and 60women. The mean age was 6.3 years, with a rangeof 0.1 years to 41.9 years. The diagnoses of allpatients were recorded (Table 1 ). Of the total 141patients, 56 underwent unilateral mandibular dis-traction and 85 underwent bilateral mandibulardistraction, for a total of 226 sided (56 � 85 � 2)distraction procedures. The number of proce-dures that were performed on native bone usingexternal devices was 149 versus 41 with internaldevices. There were 36 distractions performed ongrafted bone (external devices only).

After surgical osteotomy and fixation of thedevice, there was a latency period of 6 days (0 to7 days). The device was activated at a rate of 1mm/day: 0.5 mm in the morning and 0.5 mm inthe evening. The patient’s parents were instructedto activate the device and the activation was su-pervised periodically. With the multiguided de-vice, the force vector angulation was changed toachieve the desired direction when necessary. Themean activation period was 30.2 days, with a meandistraction length of 25.45 mm (as calibrated onthe device). Activation of the device was termi-nated when the clinical objective was achieved.2The mean consolidation period of 60.4 days(range, 45 to 75 days) was followed by device re-moval. The progress of distraction was evaluatedwith periodic radiographs. All incidents encoun-tered during the distraction process were docu-mented in the patient’s record.

Stratification of Surgical Incidents: The SeverityScale

We designed a stratification system, based onthe severity of the event, to classify incidents ob-served during any surgical intervention and theassociated perioperative period. The three broadcategories of incidents are minor, moderate, andmajor. The key to understanding and using thisstratification scheme is based on the basic termi-nology. Incidents are defined as an unintendedevent, however trivial or commonplace, that couldhave harmed or did harm a patient.10 An adverseoutcome is defined as an unintended and unwantedevent or state occurring during or following med-ical care, that is so harmful to a patient’s healththat (adjustment of) treatment is required or thatpermanent damage results.11 The perioperativeperiod is defined as that interval period beginningat the commencement of the surgical procedureand ending up to 60 days postoperatively (ex-tended in this study to the conclusion of the con-

Table 1. Diagnoses of the 141 Patients Included inthe Study

Diagnoses No. of Patients

Craniofacial microsomia 77Treacher Collins syndrome 22Robin sequence 16Nager syndrome 4Developmental micrognathia 13TMJ ankylosis 4Femoral hypoplasia 3Cerebrocostomandibular syndrome 1Tumor 1Total 141TMJ, temporomandibular joint.

Plastic and Reconstructive Surgery • February 2009

628

solidation period or removal of the distractiondevice).We propose the following stratificationmethod (Table 2) to report all incidents encoun-tered during and following any surgical proce-dure.

Minor IncidentA minor incident is defined as a perioperative

event that does not lead to an adverse outcomeand can be resolved with or without noninvasivetherapy. An example of such an event could be theresulting difficulty in opening the mouth duringthe mandibular distraction procedure or a tem-porary mechanical problem faced in the activationof the device. Both problems will not lead to anadverse outcome and can be resolved without in-vasive intervention.

Moderate IncidentA moderate incident is defined as a perioper-

ative event that could result in an adverse outcomebut can be resolved with invasive therapy or returnto the operating room. An example of such anincident would be a mechanical failure of a dis-traction device during activation. To avoid com-promising the clinical outcome, the patientshould be returned to the operating room to re-place the device.

Major IncidentA major incident is defined as a perioperative

event that results in an adverse outcome and didnot resolve or could not be resolved with invasivetherapy. An example of such a incident would bepremature consolidation of the distraction bonysegments, fibrous nonunion, or nerve injury (e.g.,neurotmesis). All of these incidents have a signif-icant impact on the patients.

The severity scale can be applied easily to anysurgical procedure. The potential qualificationsfor admission to each category are listed in Table3.

RESULTSThe rate of minor incidents, moderate inci-

dents, and major incidents observed in all the

categories of mandibular distraction is presentedin Table 4. Minor incidents were observed in 26.99percent of all procedures, and moderate incidentswere reported in 20.35 percent of distraction pro-cedures. The overall rate of major incidents was

Table 2. Perioperative* Incident Stratification

Event Definitions Examples

Minor incident Event that does not lead to adverse outcomesand can be resolved with or withoutnoninvasive therapy

Pin-site infection, trismus, hypertrophicscarring

Moderate incident Event that could result in an adverseoutcome but can be resolved with invasivetherapy

Premature ossification, device breakage,improper vector

Major incident Event that results in an adverse outcome anddid not or could not be resolved withinvasive therapy

Tooth injury, fibrous nonunion, devicefailure

*The perioperative period extends from the date of surgical procedure until removal of device.

Table 3. Subcategories of Minor, Moderate, andMajor Incidents

I. Minor incidentA. Hardware related

1. Difficulty in device activation2. Device backup3. Improper vector4. Inadequate device length

B. Hard- and soft-tissue related1. Pain at bony generate site2. Hypertrophic scar3. Cyst caused by pins4. Psychological problem5. Neurapraxia6. Infection7. Trismus during treatment or device removal8. Parotid gland injury

II. Moderate incidentA. Hardware related

1. Inadequate device length2. Improper vector3. Device backup4. Unstable device5. Unstable device following trauma6. Device deformation or breakage

B. Hard- and soft-tissue related1. Premature ossification2. Scar requiring revision

III. Major incidentA. Hardware related

1. Unstable device2. Device deformation breakage

B. Hard- and soft-tissue related1. Tooth follicle perforation and dislodgement2. Premature consolidation3. Fibrous union4. TMJ ankylosis and degenerative changes5. Fracture6. Nerve injury (e.g., neurotmesis)

C. Medical1. Systemic infection/sepsis2. Embolism3. Cardiac event4. Anoxia/airway problem5. Death

TMJ, temporomandibular joint.

Volume 123, Number 2 • A New Stratified Reporting System

629

5.31 percent. Table 5 lists the types of events doc-umented for the three subgroups of mandibulardistraction. Figures 1 through 3 depict the per-centage of minor, moderate, and major incidents.

DISCUSSIONIt has been 18 years since the distraction tech-

nique was first applied to reconstruct a humanmandible. Over the intervening years, there hasbeen significant development and refinement ofthe distraction process, making it one of the mostwidely used and effective reconstructive tech-niques in craniofacial surgery. It has proven

uniquely useful in reconstructing the hypoplasticmandible, especially in growing children. Becauseof our long-term and ongoing clinical experience,we decided to present this review of our periop-erative incidents as a part of a clinical outcomestudy.

A comprehensive literature review on the re-porting of surgical incidents demonstrated thelack of a suitable classification system that couldencompass all of the events encountered duringthe associated perioperative period. Most reportsignore minor events/incidents and note only se-vere complications. This has led to a generalizedunderreporting of incidents that could greatly af-fect both the patient and clinician experience withthe procedure and ultimate clinical outcome.12

We believe that it is vitally important to report allminor and moderate incidents encountered dur-ing the perioperative period for all surgical pro-cedures. The reporting of minor and moderateincidents helps the clinician to be aware of allpossible events and to focus on improving the

Table 4. Number and Percentage of Incidents bySeverity Index

Type of Events (n � 226) No. (%)

No incidents 107 (47.3)Minor incidents 61 (26.99)Moderate incidents 46 (20.35)Major incidents 12 (5.31)

Table 5. Percentage of Minor, Moderate, and Major Incidents in Three Mandibular Distraction Groups

Native BoneExternal Devices

(n � 95)* (%)

Native BoneInternal Devices(n � 25)† (%)

Bone GraftExternal Devices

(n � 21)‡ (%)

Minor incidentDifficulty in device activation 2.68 4.88 8.33Device backup 0.00 14.63 0.00Improper vector 0.67 0.00 0.00Inadequate device length 0.67 4.88 0.00Pain at bony regenerate site 4.70 9.76 11.11Hypertrophic scar 0.67 0.00 0.00Cyst caused by pins 0.67 0.00 0.00Psychological problem 0.00 2.44 0.00Neurapraxia (inferior alveolar nerve) 0.00 2.44 0.00Infection 8.05 7.32 8.33Trismus during treatment or device removal 0.00 9.76 0.00Parotid gland injury 1.34 2.44 0.00

Moderate incident§Improper vector 2.01 2.44 0.00Device backup 0.67 4.88 0.00Loose device 6.71 2.44 11.11Loose device following trauma 4.70 0.00 11.11Device deformation breakage 0.67 2.44 5.56Premature ossification 2.68 2.44 0.00Scar requiring revision 2.01 0.00 2.78

Major incident��Unstable device 0.67 0.00 5.56Tooth follicle or damaged teeth 0.67 0.00 0.00Premature consolidation 0.67 0.00 2.78Fibrous union 0.00 2.44 0.00TMJ ankylosis and degenerative changes 1.34 0.00 5.56Fracture 0.67 0.00 0.00

TMJ, temporomandibular joint.*Distraction, n � 149.†Distraction, n � 41.‡Distraction, n � 36.§Inadequate device length, a moderate incident, did not occur with any of the procedures.��Device deformation breakage, nerve injury (e.g., neurotmesis), medical infection, embolism, cardiac event, anoxia/airway problems, anddeath did not occur with any of the procedures.

Plastic and Reconstructive Surgery • February 2009

630

Fig. 1. Percentage of minor incidents observed in three mandibular distraction groups.

Fig. 2. Percentage of moderate incidents observed in three mandibular distraction groups.

Volume 123, Number 2 • A New Stratified Reporting System

631

technique to minimize the occurrence of suchevents. The absence of a classification system thataddresses and reports minor and moderate inci-dents prompted us to develop a new stratifiedreporting system, the severity scale. We have usedthe new stratification system to report the inci-dents rates associated with mandibular distrac-tion.

Minor IncidentsThe incidences of minor incidents observed in

all three patient subgroups are listed in Table 5and Figure 1. In patients treated with internaldistraction devices, higher incidences of distrac-tion device backup (14.63 percent) and trismus(9.76 percent) were reported. The device backupor reverse slippage during activation was the mostcommon problem associated with the first gener-ation of buried or internal distraction devices.This problem could be attributed to the sleekerdevice design that does not withstand the recip-rocal rebound forces of soft tissue and the musclepull as the device is being activated. These findingshave also been well documented when using pal-atal expanders, such as those used routinely inorthodontic therapy to correct a narrow maxillarydentoalveolar arch. Device backup could lead to a

failure of the osteotomized segments to moveapart with subsequent premature consolidation. Itis critical to monitor the activation of the deviceand to confirm activation by taking serial radio-graphs (not a simple undertaking in the youngpatient) to ensure that the bony segments arebeing distracted. This problem can be compen-sated for by adding additional turns when activat-ing the device. Recognition of this internal dis-traction device problem, however, has resulted inthe addition of a ratchet mechanism in the secondgeneration of buried devices to prevent backingup.

A high incidence of trismus was also evidentwith the internal distraction device. We speculatethat these patients had trismus secondary to thedistraction device interfering with temporoman-dibular joint function. From previous work, weknow that the condyle and disk undergo reversiblechanges when loaded during distraction.13 Be-cause the internal distraction devices can deliverdistraction vectors as steep as 70 degrees relativeto the maxillary occlusal plane angle, the internaldevice may drive the condyle into the temporo-mandibular joint with greater force than the ex-ternal distraction devices, causing more trismus.14

This hypothesis, however, remains to be proven.

Fig. 3. Percentage of major incidents observed in three mandibular distraction groups.

Plastic and Reconstructive Surgery • February 2009

632

In the group of patients undergoing grafted bonedistraction, pain at the bony generate site (11.11percent) and difficultly in activation (8.33 per-cent) were common problems encountered. Therate of infections was similar in all three groups ofpatients. The most common site of infection wasthe pin site.

Inadequate device length occurred in nearly 5percent of patients treated with an internal dis-traction device. Since we treated growing chil-dren, we included an overcorrection in our treat-ment plan. Even when we achieved our desiredovercorrection, nearly all patients will require ei-ther a second distraction osteogenesis or a man-dibular orthognathic procedure at skeletal matu-rity to achieve centric occlusion. Therefore, if ourtreatment plan estimated 10 mm of overcorrec-tion, but the device length limited us to 7 mm, wewould classify this as “inadequate devicelength”—a minor incident. It is important to notethat we have also listed inadequate device lengthunder the moderate incident category, eventhough this event never occurred. How could in-adequate device length be categorized as a mod-erate incident? If, for example, we estimated that10 mm of overcorrection would achieve satisfac-tory retroglossal space for tracheostomy decannu-lation following distraction but we achieved only7 mm and the tracheostomy could not be re-moved, we would classify this as a moderate inci-dent because of inadequate device length and per-form a second procedure to obtained the desiredadvancement and remove the tracheostomy. Thisdid not occur in the 226 distraction proceduresreported here.

Moderate IncidentsThe incidence of moderate events reported in

all three of the patient groups was 20.35 percent(Table 5). Improper vector occurred in 2 percentof patients treated with external devices, and im-proper vector selection occurred in 2.44 percentof patients treated with internal semiburied de-vices. We hypothesize that improper vector trajec-tory occurred in patients treated with an externaldistraction device because the muscles of masti-cation prevented the mandible from respondingto the change in vector direction. For example, ifthe patient had a large preexisting anterior openbite and little masseter function, as the distractedmandible moved anteriorly, the tight strap mus-cles and absence of masseter activity prevented themandible from following the external distractiondevice vector change as we attempted to close the

anterior open-bite. In contrast, the vector of man-dibular trajectory is fixed once the internal semi-buried devices is placed and has been describedmathematically by Dec et al. and Vendittelli etal.14,15 Therefore, inaccuracies in internal devicevector may be the result of inaccurate vector se-lection.

A higher incidence of external device insta-bility was observed in the grafted bone group(11.11 percent) compared with the native bonedistraction group (6.71 percent). The distractionof grafted bone presents challenges, as the bonevaries in volume and density when compared withthe native bone of the mandible. Inadequate re-tention of the pins in the grafted bony segmentswas the main problem. Because of small bonevolume and inherent instability of the pins whenplaced in grafted bone, alternate methods of an-chorage must be explored. Dislodgment of exter-nal devices because of trauma was also reported tobe higher in grafted bone (11.11 percent) whencompared with native bone (4.70 percent). Pro-jection of the external device makes it prone totrauma, especially in children. Most traumas oc-curred as a result of rough play. A loose device atthe end of the consolidation period does not poseany threat, but device loosening during the acti-vation phase necessitates returning the patient tothe operating room for reapplication of the de-vice.

Major IncidentThe major incident rate was 5.31 percent. In

grafted bone, loosening of distraction devices andtemporomandibular joint ankylosis each had thehighest major incident rate (5.56 percent), fol-lowed by premature consolidation (2.78 percent).Temporomandibular joint ankylosis is attribut-able primarily to a lack of adequate cartilaginousinterface between the grafted bone and the re-constructed glenoid fossa. The compressive forcesresulting from the distraction technique lead to abone-to-bone contact in the area of the recon-structed temporomandibular joint space, result-ing in bony fusion. In patients who had nativebone distraction by internal devices, fibrous unionwas the major incident (2.44 percent). This may beattributed to device instability or internal devicebackup.

CONCLUSIONSMandibular distraction osteogenesis can be

considered a safe and predictable procedure forlengthening the mandible in patients with cranio-

Volume 123, Number 2 • A New Stratified Reporting System

633

facial anomalies. Grafted bone has a higher rate ofmajor incidents than distraction of native bone.Internal buried devices have a higher rate of mi-nor events but a lower rate of major incidentscompared with external devices.

The newly stratified reporting system, the se-verity scale, is a simple and efficient method ofdocumenting all of the incidents observed duringmandibular distraction. Because of stratification,the incidents having a negative impact on clinicaloutcome can be easily identified. In addition toreporting major incidents, documenting minorand moderate incidents helps the clinician to beaware of all possible events that can occur duringdistraction, and the clinician can then focus onimproving the technique to minimize the occur-rence of such events. This system of classificationcan be modified easily and adapted to any surgicalprocedure; it should be particularly helpful in thedevelopment of surgical outcome studies.

Joseph G. McCarthy, M.D.Institute of Reconstructive Plastic Surgery

New York University Medical Center560 First Avenue

New York, N.Y. 10016josephmccarthy@med.nyu.edu

REFERENCES1. McCarthy JG, Schreiber J, Karp N, Thorne CH, Grayson BH.

Lengthening the human mandible by gradual distraction.Plast Reconstr Surg. 1992;89:1–10.

2. McCarthy JG, Katzen JT, Hopper R, Grayson BH. The firstdecade of mandibular distraction: Lessons we have learned.Plast Reconstr Surg. 2002;110:1704–1713.

3. Karp NS, McCarthy JG, Schreiber JS, Sissons HA, Thorne CH.Membranous bone lengthening: A serial histological study.Ann Plast Surg. 1992;29:2–7.

4. Adar R, Bass A, Walden R. Iatrogenic complications in sur-gery: Five years’ experience in general and vascular surgeryin a University Hospital. Ann Surg. 1982;196:725–729.

5. Mofid MM, Manson PN, Robertson BC, et al. Craniofacialdistraction osteogenesis: A review of 3278 cases. Plast ReconstrSurg. 2001;108:1103–1107.

6. van Strijen PJ, Breuning KH, Becking AG, Perdijk FG, Tuin-zing DB. Complications in bilateral mandibular distractionosteogenesis using internal devices. Oral Surg Oral Med OralPathol Oral Radiol Endod. 2003;96:392–397.

7. Veen EJ, Steenbruggen J, Roukema JA. Classifying surgicalcomplications: A critical appraisal. Arch Surg. 2005;140:1078–1083.

8. Dindo D, Demartines N, Clavien PA. Classification of surgicalcomplications: A new proposal with evaluation in a cohort of6336 patients and results of a survey. Ann Surg. 2004;240:205–213.

9. Pomposelli JJ, Gupta SK, Zacharoulis DC, Landa R, Miller A,Nanda R. Surgical complication outcome (SCOUT) score: Anew method to evaluate quality of care in vascular surgery.J Vasc Surg. 1997;25:1007–1005.

10. Bhasale AL, Miller GC, Reid SE, Britt HC. Analysing potentialharm in Australian general practice: an incident-monitoringstudy. Med J Aust. 1998;169:73–76.

11. Kievit JJJ, Sanders FBM. Adverse outcome registration andquality improvement (in Dutch). Med Contact 1999;54:1363–1365.

12. Smith AF, Goodwin D, Mort M, Pope C. Adverse events inanaesthetic practice: qualitative study of definition, discus-sion and reporting. Br J Anaesth. 2006;96:715–721.

13. McCormick SU, Grayson BH, McCarthy JG, Staffenberg D.Effect of mandibular distraction on the temporomandibularjoint: Part 2. Clinical study. J Craniofac Surg. 1995;6:364–367.

14. Dec W, Peltomaki T, Warren SM, Garfinkle JS, Grayson BH,McCarthy JG. The importance of vector selection in preop-erative planning of unilateral mandibular distraction. PlastReconstr Surg. 2008;121:2084–2084.

15. Vendittelli BL, Dec W, Warren SM, et al. Predicting theresponse to bilateral mandibular distraction Plast ReconstrSurg. (in press).

Plastic and Reconstructive Surgery • February 2009

634