Dexamethasone Added to Lidocaine Prolongs AxillaryBrachial Plexus BlockadeAli Movafegh, MD*, Mehran Razazian, MD*, Fatemeh Hajimaohamadi, MD*, andAlipasha Meysamie, MD†

*Department of Anesthesiology and Critical Care, Dr. Ali Shariati Hospital; and †Department of Community Medicine,Tehran University of Medical Sciences, Tehran, Iran

Different additives have been used to prolong regionalblockade. We designed a prospective, randomized,double-blind study to evaluate the effect of dexameth-asone added to lidocaine on the onset and duration ofaxillary brachial plexus block. Sixty patients scheduledfor elective hand and forearm surgery under axillarybrachial plexus block were randomly allocated to re-ceive either 34 mL lidocaine 1.5% with 2 mL of isotonicsaline chloride (control group, n � 30) or 34 mL lido-caine 1.5% with 2 mL of dexamethasone (8 mg) (dexa-methasone group, n � 30). Neither epinephrine nor bi-carbonate was added to the treatment mixture. We useda nerve stimulator and multiple stimulations techniquein all of the patients. After performance of the block,sensory and motor blockade of radial, median, muscu-locutaneous, and ulnar nerves were recorded at 5, 15,and 30 min. The onset time of the sensory and motor

blockade was defined as the time between last injectionand the total abolition of the pinprick response andcomplete paralysis. The duration of sensory and motorblocks were considered as the time interval between theadministration of the local anesthetic and the first post-operative pain and complete recovery of motor func-tions. Sixteen patients were excluded because of unsuc-cessful blockade. The duration of surgery and the onsettimes of sensory and motor block were similar in thetwo groups. The duration of sensory (242 � 76 versus 98� 33 min) and motor (310 � 81 versus 130 � 31 min)blockade were significantly longer in the dexametha-sone than in the control group (P � 0.01). We concludethat the addition of dexamethasone to lidocaine 1.5%solution in axillary brachial plexus block prolongs theduration of sensory and motor blockade.

(Anesth Analg 2006;102:263–7)

I ncreasing the duration of local anesthetic action isoften desirable because it prolongs surgical anes-thesia and analgesia. Different additives have been

used to prolong regional blockade. Vasoconstrictorscan be used to vasoconstrict vessels, thereby reducingvascular absorption of the local anesthetic. Clonidinehas also been studied in combination with local anes-thetics (lidocaine, mepivacaine, bupivacaine) in axil-lary brachial plexus block (1). Some studies have dem-onstrated the analgesic effect of local spinal andsystemic corticosteroids in combination with bupiva-caine (2,3). Dexamethasone microspheres have beenfound to prolong the block duration in animal andhuman studies (4–6), and adding methylprednisoloneto local anesthetic increases the duration of axillary

brachial block (7). However, these findings must beverified in a randomized prospective clinical trial.

The present placebo-controlled clinical trial evalu-ates the effect of dexamethasone added to lidocaine onthe onset and duration of axillary brachial plexusblock.

MethodsAfter institutional approval and obtaining informedpatient consent, 60 ASA physical status I-II patientsaged 20–50 yr scheduled for elective short to moderate(�90 min) hand and forearm surgery under axillarybrachial plexus block were included in the study. Pa-tients with a history of peptic ulcer disease, diabetesmellitus, hepatic or renal failure, pregnant women,and those receiving any premedications (includingopioids, benzodiazepines, and clonidine) were ex-cluded from the study.

Patients were allocated into 2 groups in a controlled,randomized double-blind design using a computer-generated randomization list to receive either 34 mLlidocaine 1.5% with 2 mL of isotonic saline chloride

Accepted for publication August 24, 2005.The authors have no financial or proprietary interest in the

instrumentation used.Address correspondence and reprint requests to Ali Movafegh

MD, Department of Anesthesiology and Critical Care, No: 57, 25th

Ave, Jahan Ara Street, Tehran 1438794811, Iran. Address e-mail tomovafegh@sina.tums.ac.ir or ali@movafegh.com.

DOI: 10.1213/01.ane.0000189055.06729.0a

©2006 by the International Anesthesia Research Society0003-2999/06 Anesth Analg 2006;102:263–7 263

(control group, n � 30) or 34 mL lidocaine 1.5% with2 mL of dexamethasone (8 mg) (dexamethasonegroup, n � 30). Neither epinephrine nor bicarbonatewere added to mixtures. All local anesthetic solutionsand adjuvant drugs were prepared by an anesthesiol-ogist not involved in the performance of brachialplexus block, patient care, or data collection.

On arrival to the operating room, standard moni-toring was established (pulse oximetry, electrocardi-ography, and noninvasive arterial blood pressuremonitoring) and oxygen was delivered via a Venturifacemask at a rate of 3 L/min. After insertion of a20-gauge IV catheter in a peripheral vein in the con-tralateral arm and administration of 1 �g/kg IV fent-anyl, axillary block was performed with the patient inthe supine position and the upper arm abducted 90°and the elbow flexed at 110°. A nerve stimulator (Poly-medic®) with a 24-gauge 7 cm Sprotte needle was usedfor precise localization of each nerve. The stimulationfrequency was set at 3 Hz, the duration of stimulationat 0.1 ms, and the intensity of the stimulating currentwas initially set to deliver 3 mA and was then grad-ually decreased. The position of the needle was con-sidered to be acceptable when an output current �0.7mA still elicited a slight distal motor response in eachof the nerve distributions (thumb opposition for me-dian, thumb abduction for radial, thumb adduction orulnar deviation of the hand for ulnar, and flexion offorearm on the arm for musculocutaneous nerves). Weused multiple stimulations technique in all of the pa-tients. Increments of anesthetic mixture (8 mL/nervein total) were injected through a stationary needleafter identifying the 4 nerves in each patient in thefollowing order: median, radial, ulnar, and musculo-cutaneous. The remaining 4 mL was injected subcuta-neously as the needle was withdrawn to block theintercostobrachial nerve. In case of blockade failure inany of the nerve distributions (i.e., if the patient didnot achieve satisfactory levels of anesthesia) or in anyof the nerve distributions that complete sensory (zeroin verbal analog scale) or motor (zero in Levvott ratingscale) block did not reach, the patients were excludedfrom the study and replaced on the randomization list,even when the block was adequate for surgery.

Sensory and motor blockade of radial, median, mus-culocutaneous and ulnar nerves were recorded after 5,15, and 30 min and every 10 min after the end of thesurgery. Sensory blockade of each nerve was assessedby pinprick and compared with the same stimulationon the contralateral hand. Sensory blockade of eachnerve was rated by the patient on a verbal analog scalefrom 100% (normal sensations) to 0% (no sensation).Motor block was evaluated by thumb abduction (ra-dial nerve), thumb adduction (ulnar nerve), flexion ofthe elbow in supination and pronation of the forearm(musculocutaneous), and thumb opposition (median

nerve). Measurements were performed using a modi-fication of the Lovett rating scale from 6 (normal mus-cular force) to 0 (complete paralysis) (1). The onsettime of the sensory and motor blockade was definedas the time between the end of last injection and thetotal abolition of the pinprick response and completeparalysis in all of the nerve distributions. The durationof sensory block was considered as the time intervalbetween the administration of the local anesthetic andthe first postoperative pain, and the duration of motorblock was defined as the time interval between thelocal anesthetic administration and complete recoveryof motor functions. The patients and the anesthesiol-ogist who evaluated the sensory and motor blockadeswere blinded as to the mixture used.

In the absence of historical data, the necessary sam-ple size was estimated based on a pilot study of 10patients. Dexamethasone added to lidocaine pro-longed the duration of the axillary brachial plexussensory block from 100 � 45 min to 165 � 73 mincompared with placebo. It was estimated that a min-imum of 27 patients in each group would be requiredto have an 80% power of detecting a 35-min differenceat a significance level of 0.05. Statistical analysis wasperformed with SPSS for Windows (SPSS Inc., Chi-cago, IL), version 11.5. For statistical analysis of de-mographic data and for comparison of groups, �2,Kruskal-Wallis test, Mann-Whitney U-test, and inde-pendent Student’s t-test analyses were performed.

ResultsSixteen patients were excluded from study because ofunsuccessful blockade (Table 1). The mean patientage, weight, and height and the duration of surgerywere similar in the two groups (Table 1). There was nosignificant difference in the onset time of the sensory(14 � 5 min in dexamethasone group versus 11 � 4min in control group) and motor (26 � 7 min indexamethasone group versus 22 � 8 min in controlgroup) block between groups. The duration of sensory(242 � 76 min in dexamethasone group versus 98 � 33min in control group) and motor (310 � 817 min indexamethasone group versus 130 � 31 min in control

Table 1. Patient Characteristics

Controlgroup

Dexamethasonegroup

Age (yr) 30 � 10 32 � 10Weight (kg) 70 � 4 68 � 3Height (cm) 165 � 7 168 � 4Sex (F/M) 9/11 7/13Duration of surgery (min) 60 � 23 62 � 28Unsuccessful blocks (n) 10 6

Values are expressed as mean � sd. There are no significant differencesbetween groups.

264 REGIONAL ANESTHESIA MOVAFEGH ET AL. ANESTH ANALGPROLONGING AXILLARY BRACHIAL PLEXUS BLOCKADE 2006;102:263–7

group) blockade were significantly longer in the dexa-methasone than in the control group (P � 0.01). Thesensory and motor blockade of axillary block of eachnerve distribution at the 5-, 15-, and 30-min points ispresented in Table 2; there was no statistical differencebetween groups (Table 2).

DiscussionThe present study indicates that the addition of 8 mgdexamethasone to 34 mL lidocaine 1.5% for axillarybrachial plexus block results in a significant increasein duration of sensory and motor blocks but that theonset time of sensory and motor blockade is similar.

In our study, duration of axillary brachial plexusblock with lidocaine (98 � 33 minutes and 130 � 31minutes for sensory and motor blockade) is moder-ately shorter than other works (8).

Ten patients in the control group and 6 patients inthe dexamethasone group (approximately 21% of thepatients in this study) failed to achieve satisfactorylevels of anesthesia and required induction of generalanesthesia. The 21% incidence of failure may be fre-quent (9), but this is comparable to an incidence thathas been reported in some previous studies on variouslocal anesthetics (8,10–12). Also, the incidence of blockfailure did not differ significantly between groups.

Previous works demonstrated that the addition ofcorticosteroid microspheres to local anesthetic pro-longed duration of blockade of the peripheral nerves(4–6). In one study, a prolonged percutaneous block-ade of sciatic nerve in rat using bupivacaine-dexamethasone microspheres was demonstrated (4).In another study, incorporation of dexamethasone intobupivacaine microspheres significantly prolonged in-tercostal nerve block in sheep (5). It has been reportedthat the intercostal injection of dexamethasone-containing bupivacaine microcapsules produces a

prolonged duration of anesthesia and analgesia (6).These authors believe that there is a causative relation-ship between the suppression of inflammation and theremarkably longer duration of effect (5). In our study,however, the prolongation of blocks by adding dexa-methasone could have been caused by a completelydifferent mechanism.

Other preliminary data suggest that methylpred-nisolone can increase the duration of sensory andmotor block. In this study (7), patients were dividedinto 2 groups to receive solutions containing 20 mLmepivacaine, 20 mL bupivacaine, 0.2 mL epinephrine,and, in one group, 40 mg methylprednisolone wasadded to this solution. The authors found that theduration of sensory analgesia (23 hours versus 16hours; P � 0.01) and motor block (19 hours versus 13hours; P � 0.001) were significantly longer in thesteroid group. The authors believed that the applica-bility of these findings to clinical practice should beverified in a randomized prospective clinical trial.

Although corticosteroids have been used success-fully for postoperative pain relief in oral, general, andorthopedic surgery (13,14), other studies have not cor-roborated these reports (15,16).

The mechanism of the analgesia induced by corti-costeroids is not fully understood. This effect is sus-pected to be mediated by their antiinflammatory orimmune-suppressive effects (17,18). The use of corti-costeroids as an adjuvant to local anesthetic for pe-ripheral nerve blocks rarely has been described, andits mechanism of action is not clearly understood.Corticosteroids cause skin vasoconstriction on topicalapplication. The vasoconstriction effects of topical ste-roids are mediated by occupancy of classical glucocor-ticoid receptors rather than by nonspecific pharma-cological mechanisms (19,20). According to thetraditional theory of steroid action, steroids bind to

Table 2. The Sensory and Motor Blockade of Axillary Block of Each Nerve Distribution at 5-, 15-, 30-min Points

Control group Dexamethasone group

5 min 15 min 30 min 5 min 15 min 30 min

Sensory blocka

MC 15% 75% 100% 10% 75% 100%RAD 20% 80% 100% 10% 75% 100%MED 25% 85% 100% 15% 85% 100%ULN 20% 80% 100% 15% 75% 100%

Motor blockb

MC 3.7 � 0.46 2.4 � 1.5 1.3 � 0.11 3.3 � 0.80 2.5 � 0.74 1.2 � 0.50RAD 4 � 0.97 2.3 � 1.04 1.3 � 0.55 3.8 � 0.95 2.2 � 0.76 1 � 0.22MED 3.8 � 0.66 2.28 � 0.92 1.2 � 0.55 3.6 � 0.82 2.5 � 0.52 1.1 � 0.3ULN 4 � 0.91 2.3 � 1.05 1.4 � 0.58 3.8 � 0.81 2.4 � 0.67 1.3 � 0.55

There are no significant differences between groups.MC � musculocutenous nerve; RAD � radial nerve; MED � median nerve; ULN � ulnar nerve.a Sensory blockade was assessed by pinprick and compared with the same stimulation on the contralateral hand. Values show the percent of no sensation in

each nerve distribution.b Motor blockade was measured using a modification of Lovett rating scale from 6 (normal muscular force) to 0 (complete paralysis). Values are expressed

as mean � sd.

ANESTH ANALG REGIONAL ANESTHESIA MOVAFEGH ET AL. 2652006;102:263–7 PROLONGING AXILLARY BRACHIAL PLEXUS BLOCKADE

intracellular receptors and modulate nuclear tran-scription. In our study, dexamethasone produced arelatively rapid effect which cannot be explained bythe above mechanism (21). Therefore, vasoconstric-tion, the presumed mechanism of action for epineph-rine’s adjunctive effect on local anesthetics, is proba-bly not responsible for block prolongation bydexamethasone. Corticosteroids may have a local ef-fect on the nerve; the dexamethasone effect may berelated to this action (22).

One possibility is that prolongation of local anes-thetic block occurs because of systemic effects of dexa-methasone. Some authors believe that analgesic prop-erties of corticosteroids are the result of their systemiceffect (13,14). Unfortunately, a control group receivingparenteral administration of the same dose of dexa-methasone was not considered when this study wasdesigned. It was not the aim of this study, however, toelucidate the mechanism of action of dexamethasone.Because of our positive results, the question ofwhether these results were attributable to a local orsystemic effect warrants further investigation.

The safety of dexamethasone use in a nerve sheathmay raise some concerns. In animal experiments, re-peated intrathecal injections of small-dose betametha-sone (23) and triamecilonon acetate (24) did not inducespinal neurotoxicity. In one study, after approximately2000 intrathecal injections of dexamethasone (8 mg) in200 patients for treatment of posttraumatic visual dis-turbance, no neurological disorders were found at1-month follow up (25). Nerve injury is a rare compli-cation of dexamethasone injection, and it usually oc-curs in the context of needle trauma (26,27).

The use of dexamethasone as an adjuvant to localanesthetics for peripheral nerve block has not beendescribed; we used a dose of 8 mg because adminis-tration of this dose seems to be safe in adults. Adverseeffects with a single dose of dexamethasone are prob-ably extremely rare and minor in nature, and previousstudies have demonstrated that short-term (�24hours) use of dexamethasone was safe (16,28).

Adding a steroid to local anesthetic solution maynot be indicated for all patients. For example, diabeticpatients may experience hyperglycemia and patientswith a continuing infectious process may be detrimen-tally affected by the antiinflammatory effects of ste-roids. The use of dexamethasone to increase the du-ration of action of local anesthetics is not an indicationof this drug. This study led us to hypothesize that itmay be useful in situations in which epinephrine mustbe used with caution (e.g., hypertension, ischemicheart disease).

In our center, lidocaine is routinely used for re-gional block procedures, which is why we chose lido-caine for this study. Considering cardiovascular tox-icity, lidocaine is safer than bupivacaine. However,

repetition of this study could be done with longer-acting drugs, such as bupivacaine or ropivacaine, toassess the effect of dexamethasone on duration ofsensory and motor blockade.

In conclusion, the addition of dexamethasone tolidocaine 1.5% solution in axillary brachial plexusblock prolongs the duration of sensory and motorblockade. Further studies are needed to evaluate theoptimal dose of dexamethasone to be used for pro-longed brachial plexus block as well as the mechanismof this effect.

References1. Kapral S, Goolann G, Walt B, et al. Tramadol added to mepiva-

caine prolongs the duration of an axillary brachial plexus block-ade. Anesth Analg 1999;88:853–6.

2. Glasser RS, Knego RS,Delashaw JB, Fessler RG. The periopera-tive use of corticosteroids and bupivacaine in the managementof lumbar disc disease. J Neurosurg 1993;78:383–7.

3. Mirzai H, Tekin I, Alincak H. Perioperative use of corticosteroidand bupivacaine combination in lumbar disc surgery: a ran-domized controlled trial. Spine 2002;27:343–6.

4. Castillo J, Curley J, Hotz J, et al. Glucocorticoids prolong ratsciatic nerve blockade in vivo from bupivacaine microspheres.Anesthesiology 1996;85:1157–66.

5. Droger C, Benziger D, Gao F, Berde CB. Prolonged intercostalsnerve blockade in sheep using controlled-release of bupivacaineand dexamethasone from polymer microspheres. Anesthesiol-ogy 1998;89:969–74.

6. Kopacz DJ, Lacouture PG, Wu D, et al. The dose response andeffects of dexamethasone on bupivacaine microcapsules for in-tercostals blockade (T9 to T11) in healthy volunteers. AnesthAnalg 2003;96:576–82.

7. Stan T, Goodman E, Cardida B, Curtis RH. Adding methylpred-nisolone to local anesthetic increases the duration of axillaryblock. Reg Anesth Pain Med 2004;29:380–1.

8. Adnan T, Elif AA, Ayshe K, et al. Clonidine as an adjuvant forlidocaine in axillary brachial plexus block in patients withchronic renal failure. Acta Anesthesiol Scand 2005;49:563–68.

9. Iohom J, Machmachi A, Diarra D. The effects of clonidine addedto mepivacaine for paronychia surgery under axillary brachialplexus block. Anesth Analg 2005;100:1179–83.

10. Dunlop DJ, Graham CM, Watt JM. The practical use of axillarybrachial plexus block for hand surgery. J Hand Surg 1995;20:677–8.

11. Vainionpaa VA, Haavisto ET, Huha TM, et al. A clinical andpharmacokinetic comparison of ropivacaine and bupivacaineinaxillary plexus block. Anesth Analg 1995;81:534–8.

12. McGlade DP, Kalpokas MV, Mooney PH, et al. A comparison of0.5% ropivacaine and 0.5% bupivacaine for axillary brachialplexus anesthesia. Anaesth Intensive Care 1998;26:515–20.

13. Aasboe V, Raeder JC, Groegaard B. Betamethasone reducespostoperative pain and nausea after ambulatory surgery.Anesth Analg 1998;87:913–7.

14. Baxendale BR, Vater M, Lavery KM. Dexamethasone reducespain and swelling following extraction of third molar teeth.Anaesthesia 1993;48:961–4.

15. Liu K, Hsu CC, Chia YY. Effect of dexamethasone on postop-erative pain and emesis. Br J Anaesth 1998;80:85–6.

16. Tan P, Liu K, Peng CH, et al. The effect of dexamethasone onpostoperative pain and emesis after intrathecal neostigmine.Anesth Analg 2001;92:228–32.

17. McCormack K. The spinal actions of nonsteroidal anti-inflammatory drugs and the dissociation between their anti-inflammatory and analgesic effects. Drugs 1994;47:28–45.

18. Ahlgren SC, Wang JF, Levine JD. C-fiber mechanical stimulus-response functions are different in inflammatory versus neuro-pathic hyperalgesia in the rat. Neuroscience 1997;76:285–90.

266 REGIONAL ANESTHESIA MOVAFEGH ET AL. ANESTH ANALGPROLONGING AXILLARY BRACHIAL PLEXUS BLOCKADE 2006;102:263–7

19. Marks R, Barlow JW, Funder JW. Steroid-inducedvasoconstriction: glucocorticoid antagonist studies. J Clin EndoMeta 1982;54:1075–7.

20. Seidenari S, Di Nardo A, Mantovani L, Giannetti A. Parallelintraindividual evaluation of the vasoconstrictory action andthe anti-allergic activity of topical corticosteroids. Exp Dermatol1997;6:75–80.

21. Taguchi H, Shingu K, Okuda H, Matsumoto H. Analgesia forpelvic and perineal cancer pain by intrathecal steroid injection.Acta Anaesthesiol Scand 2002;46:190–3.

22. Devor MD, Gorvin-Lippmann R, Raber P. Corticosteroids sup-press ectopic neural discharge originating in experimental neu-romas. Pain 1985;22:127–37.

23. Latham JM, Fraser RD, Moore RJ, et al. The pathologic effects ofintrathecal betamethasone. Spine 1997;22:1558–62.

24. Abram SE, Marsala M, Yaksh TL. Analgesic and neurotoxiceffects of intrathecal corticosteroids in rats. Anesthesiology1994;81:1198–205.

25. Sugita K, Kobayashi S, Yokoo A, et al. Intrathecal steroid ther-apy for post-traumatic visual disturbance. Neurochirurgia(Stuttg) 1983;26:112–7.

26. Frederick HA, Carter PR. Injection injuries to the median andulnar nerves at the wrist. J Hand Surg Am 1992;17:645–7.

27. Stahl S, Kaufman T. Ulnar nerve injury at the elbow aftersteroid injection for medial epicondylitis. J Hand Surg Br1997;22:69–70.

28. Splinter WM, Rhine EJ. Low-dose ondansetrone with dexameth-asone more effectively decreases vomiting after strabismus sur-gery in children than does high-dose ondansetron. Anesthesi-ology 1998;88:72–5.

ANESTH ANALG REGIONAL ANESTHESIA MOVAFEGH ET AL. 2672006;102:263–7 PROLONGING AXILLARY BRACHIAL PLEXUS BLOCKADE