Lepr Rev (2009) 80, 34 50

Sensitivity and specicity of nerve palpation, monolament testing and voluntary muscle testing in detecting peripheral nerve abnormality, using nerve conduction studies as gold standard; A study in 357 patients
FATEMA ABBAS KHAMBATI, VANAJA PRABHAKAR SHETTY, SUNIL DATTATRAYA GHATE & GOSPI DOLLY CAPADIA The Foundation for Medical Research, 84-A, R.G. Thadani Marg, Worli, Mumbai 400 018, India Accepted for publication 13 February 2009
Summary Objective To determine sensitivity and specicity of clinical tools viz. nerve palpation (NP), monolament (MF), and voluntary muscle testing (VMT), for assessing peripheral nerve function impairment (NFI) in leprosy, using nerve conduction studies (NCS) as gold standard. Study population and Methods 357 untreated multibacillary (MB) leprosy patients were assessed using above tests. The nerves assessed were left and right ulnar, median, radial cutaneous, sural, common peroneal and posterior tibial. The concordance between the clinical and NCS tests was done for each nerve. The sensitivity and specicity of clinical tests for detecting nerve impairment was determined, using NCS as gold standard. Analysis was performed using SPSS version 10.0. Results The sensitivity of NP ranged between 71% to 88% for all nerves, except the median (43%) and sural (59%) nerves. Specicity was . 60% for all, but low for ulnar (34%) and common peroneal (40%) nerves. The specicity of MF testing was . 80% and of VMT assessment was . 90% for all nerves. The sensitivity of MF testing ranged between 35 44%, while of VMT assessment was very low i.e. 4 5%, the maximum was for the ulnar nerve (25%). Detection sensitivity of MF testing and VMT assessment improved two fold when combined with NP and was closely comparable to NCS test ndings. Conclusions Both MF testing and VMT assessment showed good specicity, but moderate to low sensitivity. NP was less specic but more sensitive than MF testing and VMT assessment. Combining NP with MF testing and VMT assessment gives a two fold improvement in the sensitivity for assessing nerve damage and could therefore serve as the most useful clinical tools for diagnosis of leprosy and detecting nerve damage at eld level.
Correspondence to: Vanaja P. Shetty, The Foundation for Medical Research, 84-A, R.G. Thadani Marg, Worli, Mumbai 400 018, India (Tel: 91 22 24934989/24932876 (also fax); e-mail: fmr@fmrindia.org, fmr@vsnl.net)

34

0305-7518/09/064053+17 $1.00

q Lepra

Correlation between clinical and electrophysiological tests

35

Introduction Early detection and treatment of nerve function impairment (NFI) is of paramount importance in leprosy. When detected and treated early, primary impairments may be reversible. Problems associated with leprosy neuropathy include loss of sensory and autonomic nerve functions and muscle strength. In practice, touch/pressure and muscle strength are the modalities tested in leprosy patients.1,2 However, the issue of importance lies in the sensitivity, reliability and reproducibility of any standard assessment method. Owen and Stratford (1995)3 have reviewed several commonly used methods for sensory testing and concluded that the ball-point test, cotton wool and pin prick tests recommended by WHO (1970)4 were all cheap and easy to use, but were not sensitive enough to be of practical value. Assessment of pressure using a barograph5 temperature using hot and cold test tubes6 as well as thermotester7 were being used at some centres. Similarly, the sweat function test8 and arteriographic pattern9 were also practiced for assessing autonomic impairment but proved cumbersome to be used at the eld level. The Semmes Weinstein monolaments (MF), used for assessing sensory nerve function are advocated on the grounds that the results are reliable, since the force required to bend the accurately manufactured monolaments is relatively constant and repeatable when used by skilled examiners, and since they are graded (1 5), they provide a quasi-quantitative estimate of sensory loss.10,11 Similarly the voluntary muscle testing (VMT) described by Brandsma (2000),12 is an important technique for the assessment and evaluation of motor nerve dysfunction. Application of monolaments outside the eld of leprosy viz. diabetes has also been recorded. In diabetes it is used as a screening tool to identify patients at risk for foot complications in primary case setting.13 However the MF test is sometimes criticized on the grounds that the ready-made monolaments are relatively expensive and not always easy to obtain. Also the laments may be lost or need to be replaced after they have become bent and a test using multiple monolaments is time consuming for widespread use.14 Similarly, by VMT assessment, minor degrees of weakness cannot be detected and the test will be positive when there is considerable weakness of the muscles present.12 However, both MF testing and VMT assessment are established tests, for assessing sensory and motor functions of peripheral nerves respectively and studies have shown their good inter-tester reliability and reproducibility.1,15 17 On the other hand, nerve enlargement at certain sites is one of the cardinal signs of leprosy and often precedes sensorimotor decits.18,19 Studies have reported palpable enlargement of nerves as the most common physical nding in leprosy.20,21 However there are also reports that highlight the moderate level of reliability in assessing thickened nerves when undertaken by paramedical workers (PMWs) at eld level.22,23 The utility of electrophysiological methods particularly nerve conduction studies (NCS) in the detection and monitoring of nerve abnormalities in leprosy and other neuropathies have been well established.24 26 Though not specic to leprosy neuropathy, NCS is by far more reliable,27 reproducible28,29 and has proved a sensitive measure of nerve damage, since it denes small late components originating from demyelinated, remyelimated or regenerated bres.30,31 The large diameter sensory bres have lower thresholds and conduct faster than motor bres by about 5 to 10% and fastest sensory conduction velocity is particularly observed amongst mixed nerves.32 Studies correlating

36

F. A. Khambati et al.

the nerve conduction test ndings with the clinical tests are few and far between.33 35 Studies by van Brakel, et al. (2005),34 as well as Kaplan and Gelber (1985)2 have reported a good concordance between MF testing and sensory nerve conduction (SNC) studies, supporting the validity of monolaments as standard screening test of sensory function and the usefulness of clinical testing modalities for assessing nerve function impairment (NFI). Since NCS is a more sensitive technique for assessing nerve damage, it is considered as gold standard for this study. A cohort of untreated multibacillary (MB) leprosy patients were assessed at baseline, for nerve function impairment using three clinical methods viz. nerve palpation (NP), monolaments testing (MF) and voluntary muscle testing (VMT). All the tests were performed on both right and left sensory nerves including the ulnar, median, radial cutaneous, sural and motor nerves including the ulnar, median, common peroneal and posterior tibial. The results were compared with that of sensory (SNC) and motor (MNC) nerve conduction tests performed on the same nerves. Our aim was to determine the sensitivity and specicity of each of the clinical tests using NCS as gold standard.

Patients and Methods

STUDY POPULATION

Four hundred newly registered multibacillary (MB) leprosy patients were recruited at the Foundation for Medical Research between February 2001 and April 2005 for a larger prospective cohort study where the effect of corticosteroids on the pathogen and pathological process was investigated (submitted for publication to Leprosy Review, 2009). Of these 357 patients were subjected to all the four neurological tests described below (in methods section) and included in this study.

STUDY SUBJECTS

Patients of either sex aged 12 60 years. The denition of multibacillary leprosy was having . 5 skin lesions or with a positive slit skin smear or with more than one enlarged peripheral nerve.36

Methods
NERVE PALPATION (NP)

All the major nerves listed above and their branches were palpated bilaterally at the sites of predilection (see Figure 1) to record the enlargement and graded as 0 (no enlargement), 1 (slightly enlarged), 2 (moderately enlarged) or 3 (very enlarged).37,38 Note: During palpation, the corresponding nerve on the other side was always examined. A nerve was not pronounced as enlarged without comparing the one on the other side. Gross enlargement is obvious. A general rule that was followed was, if in doubt, consider the nerve not enlarged.

Correlation between clinical and electrophysiological tests

37

Figure 1. Sites of potentially palpable nerve enlargement in leprosy.

NERVE FUNCTION ASSESSMENTS

(a) Touch sensibility testing using monolaments (MF) Touch sensibility was tested with a standard set of ve coloured Semmes Weinstein monolaments (MF) as described by Bell Krotoski (1990).11 When applied with a force sufcient to bend the lament, these were respectively equal to application forces of 50 mg score 5, 200 mg score 4, 2 g score 3, 4 g score 2, and 300 g score 1. A score of 5 was given when the thinnest lament was felt and zero, even if the thickest lament was not felt. Normal reference values were up to 200 mg for hands and 2 g for the foot.39,40 The nerves tested were bilateral ulnar (three sites, normal score 11 15), median (three sites, normal score 11 15), radial cutaneous (one site, normal score 4 5) and sural (one site, normal score 4 5). Any nerve scoring below the normal score was considered impaired. Test sites: Ulnar nerve at hypothenar eminence, 5th metacarpal head and distal phalanx of the little nger.

38

F. A. Khambati et al.

Median nerve at 2nd metacarpal head, distal phalanx of the thumb and distal phalanx of the index nger. Radial cutaneous nerve over dorsum of the thumb at the site of motor point. Sural dorsal lateral aspect of the foot.

(b) Voluntary muscle testing (VMT) This was done using the modied Medical Research Council (MRC) scale.12 The muscles tested per nerve were; for ulnar nerve abductor digiti minimi, for median nerve abductor pollicis brevis, for common peroneal nerve tibialis anterior, peroneus longus and brevis, for posterior tibial nerve small intrinsic muscles of feet. Grading criteria were as follows: 5 4 3 2 1 0 full range of movement at the joint with normal resistance full range of movement but with less than normal resistance full range of movement but no resistance partial range of movement with no resistance perceptible contraction of muscles not resulting in joint movements complete paralysis

Criteria for motor impairment was any muscle scoring less than grade 4. Note: All the clinical tests were performed/ascertained and documented by a clinician (SG) with more than 10 years experience in palpating nerves, performing sensory testing using monolaments and voluntary muscle testing and was assisted by a team member (FK). Inter tester variation To validate the results as well as measure reliability, in 40 patients the assessments were performed independently by two team members and the results were compared with that of the clinician (SG) who served as a gold standard. In order to measure reliability, the weighed Kappa statistic (Kw) as reliability coefcient dened by Altman41 was used. Kw may be interpreted as the proportion of agreement between testers corrected for the effect of chance. The computation yields a value for (Kw) in the range 0 to 1, where 0 reects agreement no better than that arising purely by chance and 1 reects perfect agreement. According to the classication by Altman, a (Kw) of 060 or greater indicates good agreement and a (Kw) in excess of 080 indicates very good agreement. Normal values: 25 normal subjects in the age group of 23 52 years were tested by the three clinical tests (NP, MF testing, VMT assessment). By NP all the nerves graded as (0 not enlarged). By MF testing all the nerves scored either 5 or 4 with total score of 11 15. Similarly by VMT assessment, all the muscles scored 5 with total score of 25.

(c) Sensory nerve conduction measurements (SNC) Sensory action potential (SAP) parameters were measured (Neurocare 2000 EMG machine, Biotech Ltd, Mumbai), bilaterally in four major sensory nerves, at a xed stimulation recording distance of 14 cm. The ulnar and median nerves were stimulated at little and index ngers respectively with pick up at wrist, while radial cutaneous nerve action potential was recorded at the rst web space and stimulated at the radial side of the forearm. The sural nerve was picked up behind the lateral malleolus and stimulated at mid calf.

Correlation between clinical and electrophysiological tests

39

(d) Motor nerve conduction measurements (MNC) Compound muscle action potential (CMAP) parameters were measured bilaterally in four motor nerves, following stimulation at standard distal and proximal sites. Abductor digiti minimi muscle served as pick up for ulnar nerve, below and above the sulcus, stimulated at wrist, the distance between the sites xed at 10 cm. Abductor pollicis brevis muscle was the pick up for median nerve, stimulated at wrist and elbow. Extensor digitorum brevis muscle was the pick up for the common peroneal nerve, stimulated at the ankle and behind the capitulum bulae. Abductor hallucis brevis muscle was the pick up for posterior tibial nerve, stimulated behind the medial malleolus and at the popliteal fossa. The measured value for latency, amplitude, and conduction velocity were recorded and stored in a separate Access database. Skin temperatures were measured electronically at wrist and ankle and the measured latencies and velocities normalised for a temperature of 338 C at the time of analysis using standard formulae.42 Filter settings: 2 Hz 10 KHz. Type and size of the electrode: Silver silver chloride surface electrodes of 1 cm. diameter were used for pick up. Stimulating electrode was bipolar with a xed distance of 25 cms. between the two poles. Note: The nerve conduction (sensory and motor) test was undertaken by a physiotherapist (GC) with 30 years experience in electrophysiology. Normal values: Normal values for sensory and motor conduction studies were obtained by studying the same 25 normal subjects that were clinically assessed in the age group of 23 to 52 years with no apparent disease of the peripheral nerves. The normal values i.e. Mean ^ standard deviation (cut-off) for each individual nerve is presented in Table 1 below.

Concordance between clinical and electrophysiological tests Findings of the three clinical testing methods employed for assessing nerve function impairment; NP, MF testing and VMT assessment were compared with SNC and MNC test ndings to determine the sensitivity and specicity of these testing methods singly and in combination. (Note: These tests are routinely done at our outpatient clinic). Combination: The combination of the tests were done as; MF testing VMT assessment, MF testing NP, VMT assessment NP, MF testing VMT assessment NP which meant when either by single or by all two or three tests a nerve showed impairment it was set to impaired and only if by all the tests it showed normal, it was set to unimpaired.
Table 1. Normal values for nerve conduction studies Sensory nerves CV (m/sec) Amp (mV) Motor nerves CV (m/sec) Amp (mV) Ulnar 6246 ^ 1016 (52) 71 ^ 28 (4) Ulnar 596 ^ 114 (48) 158 ^ 51 (11) Med 6650 ^ 159 (50) 195 ^ 9 (11) Med 602 ^ 127 (47) 168 ^ 64 (10) R.cut 6593 ^ 1483 (51) 253 ^ 138 (12) C.per 462 ^ 59 (40) 68 ^ 28 (4) Sural 6347 ^ 1384 (49) 216 ^ 102 (11) P. tib 485 ^ 78 (40) 17 ^ 72 (10)

Lower limits of normal (cut off) calculated as mean 1 SD.

40

F. A. Khambati et al.

DEFINITIONS

Concordance The direct agreement between the results of two tests in terms of impaired and not impaired. Sensitivity This is the measure of probability of correctly diagnosing a diseased condition. Specicity This is the measure of probability of correctly identifying a non-diseased condition. The NCS test ndings were taken as reference values for determining the sensitivity and specicity of clinical test ndings (NP, MF testing, VMT assessment). Type 1 reaction (T1R) Patients presenting with erythema and oedema of skin lesions. There may be accompanying neuritis and oedema of the hands, feet and face.43 Type 2 reaction (T2R) Patients presenting with crops of tender subcutaneous skin lesions. There may be accompanying neuritis, iritis, arthritis, orchitis, dactylitis, lymphadenopathy, oedema and fever.43 Neuritis (N) Patients presenting with acute inammation of one or more peripheral nerve trunk detectable by swelling and/or functional impairment with spontaneous nerve pain and/or nerve tenderness on palpation. Nerve pain or tenderness may or may not be accompanied by paraesthesia.43 Silent neuropathy (SN) Patients with clinically detectable sensory and/or motor impairment but without spontaneous complaints of nerve pain, paraesthesia and/or nerve tenderness on palpation and with no evidence of reaction (either T1R or T2R).43 Ethical clearance: No nancial incentives were given to the patients. Travel expenses were refunded to the patients on occasion. The study followed the International Ethical Guidelines for Biomedical Research Involving Human Subjects (CIOMS/WHO, 1993). The study had the clearance from the Institutional (The Foundation for Medical Research) Ethics Committee. Written consent was obtained from individual study subjects before inclusion in the study using a standard consent form. Data analysis: Analysis was performed using SPSS version 10.0. The signicance of association was tested using Chi-square test. The 95% condence interval was calculated.

Results Of the 357 untreated MB patients included, 280 (78%) were males and 77 (22%) females. Age ranged between 12 60 years; (mean 314) years. On classifying as per RidleyJopling scale, 39% patients were borderline tuberculoid (BT), 20% mid borderline (BB), 24% borderline lepromatous (BL), 12% sub-polar lepromatous (LLs) and 5% were pure neural (PN). A total of 141/357 (39%) patients presented with either Type 1 or Type 2 reaction and/or neuritis or silent neuropathy of , 6 months duration and were treated with Prednisolone 40 mg daily tapered to 5 mg daily over 12 weeks as per WHO recommendation.36 40 mg, 30 mg, 20 mg, 15 mg, 10 mg, 5 mg for 2 weeks each (daily dose). All the patients were subjected to clinical (NP, MF testing, VMT assessment) and NCS tests. The reliability statistics of nerve palpation, monolament testing and voluntary muscle testing of ulnar, median, radial cutaneous, sural, common peroneal and posterior tibial nerves have been summarised in Table 2.

Correlation between clinical and electrophysiological tests

41

Table 2. Agreement between tester and gold standard and the resulting weighed Kappa values for NP, MF testing and VMT assessment (assessed in 40 patients) NP Kw (95% CI) MF Kw (95% CI) VMT Kw (95% CI)

Nerves Sensory nerves Ulnar Median R. cutaneous Sural Motor nerves Ulnar Median C. peroneal P. tibial

062 (045 076) 057 (040 072) 070 (053 082) 055 (039 070) 062 (046 077) 057 (040 073) 073 (059 085) 058 (042 074)

073 (055 086) 080 (063 090) 069 (054 082) 060 (044 074) NA NA NA NA

NA NA NA NA 093 (078 099) 082 (064 093) 094 (079 099) 097 (082 099)

Key: Kw values . 08 very good agreement, $ 0608 good agreement. Note: Each of the outcome variables (NP, MF testing, VMT assessment) were categorized into either Impaired/not impaired for each nerve, for determining agreement between the tester and gold standard.
NERVE PALPATION (NP) FINDINGS

This was assessed in 357 patients, of which 302 (85%) showed 1 to 3 thickening of one or more nerves. Of 4284 nerves examined, 1363 (32%) were graded as 1 (denitely enlarged), 799 (19%) as 2 (moderately enlarged) and 151 (4%) as 3 (very enlarged). The ulnar nerve at the sulcus was the most frequently enlarged (74%), followed by common peroneal (65%), radial cutaneous (55%), sural (50%), posterior tibial (50%) and median (45%) nerves (Table 3).

Table 3. No. (%) of patients and nerves showing enlargement by NP, impairment by clinical (MF testing, VMT assessment) and abnormal nerve conduction (SNC and MNC) tests No. (%) NP Patients Nerves Sensory Ulnar Median R.cut Sural Total Motor nerves Ulnar Median C. per P. tibial Total 302/357 (85) 529/714 (74) 319/714(45) 390/714 (55) 356/714 (50) 1594/2856 (56) 529/714 (74) 319/714 (45) 462/714 (65) 355/776 (50) 1665/2856 (58) MF 147/357 (41) 162/714 (23) 125/714 (17) 163/714 (23) 199/714 (28) 649/2856 (23) NA NA NA NA NA VMT 109/357 (31) NA NA NA NA NA 126/714 (18) 22/714 (03) 11/714 (02) 11/714 (02) 170/2856 (06) SNC 314/357 (88) 361/714 (51) 285/714 (40) 333/714 (47) 517/714 (72) 1496/2856 (52) NA NA NA NA NA MNC 266/357 (75) NA NA NA NA NA 400/714 (56) 236/714 (33) 191/714 (27) 225/714 (32) 1052/2856 (37)

Key: NP nerve palpation, MF monolaments testing, VMT voluntary muscle testing, SNC & MNC sensory and motor nerve conduction, NA not applicable.

42

F. A. Khambati et al.

MONOFILAMENT (MF) TESTING

Assessments using MF testing was assessed in 357 patients of which 147 (41%) showed impaired sensory nerve function. Proportion of nerves showing impairment ranged between 18 and 28%. Sural nerve showed maximum impairment (28%) followed by ulnar (23%), median (17%) and radial cutaneous (23%). Overall, 23% sensory nerves showed impairment by MF testing (Table 3).

VOLUNTARY MUSCLE TESTING (VMT)

Assessment using VMT assessment was done in 357 patients, of which 109 (31%) showed impaired motor nerve function. The ulnar nerve showed maximum impairment (18%) followed by median (3%), common peroneal (1%) and posterior tibial (1%). Overall, 5% motor nerves showed impairment by VMT assessment (Table 3).

SENSORY NERVE CONDUCTION (SNC) STUDY FINDINGS

SNC parameters were tested in 357 patients, of which 314 (88%) showed impairment in the nerves tested. Sural nerve showed maximum abnormality (72%) followed by radial cutaneous (47%), median (40%) and ulnar (36%). Overall, 48% sensory nerves showed abnormal function (Table 3).

MOTOR NERVE CONDUCTION (MNC) STUDY FINDINGS

MNC parameters were tested in 357 patients, of which 266 (75%) showed impairment in the nerves tested. Ulnar nerve showed maximum abnormality (56%) followed by median (33%), posterior tibial (32%) and common peroneal (27%). Overall, 37% motor nerves showed abnormal function (Table 3).

CONCORDANCE BETWEEN THE CLINICAL (NP, MF TESTING, AND VMT ASSESSMENT) AND NCS (SNC AND MNC) TESTS

This was obtained in 357 patients (714 nerves each). Between NP and NCS tests, concordance was seen in 61% nerves. Maximum concordance was observed for ulnar-motor (44%) and sural (44%) nerves. Between MF testing and SNC tests, concordance was seen in 65% nerves and maximum was for sural (27%) nerve. Similarly between VMT assessment and MNC, concordance was seen in 60% nerves and maximum was for ulnar nerve (14%) (Table 4).

SENSITIVITY AND SPECIFICITY OF CLINICAL TESTS (NERVE WISE)

On comparison of NP with SNC & MNC, the specicity was low (30 40%) for ulnar and common peroneal nerves as compared to other sensory and motor nerves (60 80%). On the

Correlation between clinical and electrophysiological tests

43

Table 4. Concordance between clinical (NP)/(MF testing)/(VMT assessment) and NCS (SNC & MNC) tests ndings assessed in 357 patients (714 nerves each) By clinical NP (sensory) By NCS Nerves Ulnar Imp Not imp Median Imp Not imp R.cut Imp Not imp Sural Imp Not imp C.per Imp Not imp P.tibial Imp Not imp enl Not enl enl NP (motor) Not enl MF Imp Not imp VMT Imp Not imp

225 (32) 302 (42) 141 (20) 88 (12) 236 (33) 156 (22) 313 (44) 41 (6)

31 (4) 156 (22) 139 (20) 346 (48) 98 (14) 224 (31) 207 (29) 153 (21)

313 (44) 214 (30) 101 (14) 128 (18) 152 (21) 307 (43) 161 (23) 191 (27)

84 (12) 103 (15) 84 (12) 349 (49) 41 (6) 214 (30) 63 (9) 299 (42)

110 (15) 58 (8) 97 (14) 31 (4) 131 (18) 35 (5) 191 (27) 14 (2)

142 (20) 404 (57) 180 (25) 406 (57) 202 (29) 346 (48) 332 (46) 177 (25)

100 (14) 24 (3) 13 (2) 24 (3) 9 (1) 2 (03) 8 (1) 3 (03)

300 (42) 290 (41) 300 (42) 290 (41) 181 (25) 522 (73) 213 (30) 490 (69)

Key: Imp impaired, enl enlarged. Note: Percentages in parenthesis shows the agreement between clinical (NP, MF, VMT) and (NCS gold standard) test.

other hand, the sensitivity was higher than specicity for radial cutaneous, common peroneal, posterior tibial, and was maximum for ulnar nerve (. 80%) (Table 5). On comparison of MF testing and SNC, for all the nerves, specicity was higher (. 80%) than sensitivity. The sensitivity ranged between 35 and 44% and was maximum in case of ulnar nerve (Table 5). On comparison of VMT assessment and MNC, the highest sensitivity was seen in case of ulnar nerve, while for all other nerves the sensitivity was low (4 5%). The specicity was high for all the nerves (. 90%) (Table 5).

Discussion The study examined: (a) the frequency of abnormality of the major sensory and motor nerves in 357 untreated MB leprosy patients, as detected by NP, MF testing, VMT assessment, SNC and MNC tests, (b) the degree of concordance between the clinical and NCS test ndings and (c) the sensitivity and specicity of each of the clinical tests using NCS as gold standard. Among the clinical tests, NP was found to be the most frequently abnormal with 85% of patients showing one or more enlarged peripheral nerves, however it had lower specicity. The sensitivity of NP was signicantly higher (P , 0001) in patients presenting with reaction and/or neuritis (Gr.A 73%) as compared to patients with no reaction (Gr.B 63%) (Table 6).

44

F. A. Khambati et al.

Table 5. Nerve wise sensitivity and specicity of clinical tests taking NCS as reference test (gold standard) assessed in 357 patients (714 nerves each) Sensitivity % (95% CI) Specicity % (95% CI) Chi square (Yates corrected)

Clinical Testa Sensory nerves Ulnar NP MF Median NP MF R. cut NP MF Sural NP MF Motor nerves Ulnar NP VMT Median NP VMT C. per NP VMT P. tibial NP VMT

Sig.

88 44 50 35 71 39 60 36 78 25 43 5 78 5 71 4

(8392) (3850) (4456) (2941) (6575) (3445) (5664) (3240) (7482) (2129) (3750) (3 9) (7284) (2 9) (6578) (2 8)

34 (2939) 88 (8490) 80 (7683) 93 (9095) 59 (5464) 90 (8793) 79 (7284) 93 (8896) 32 (2738) 92 (8894) 73 (6977) 99 (97100) 40 (3745) 99 (98100) 61 (5765) 99 (98100)

3981 8760 6931 9000 6174 8547 8467 5806 1113 3244 1761 806 2327 6525

P P P P P P P P P P P P P P P P

, , , , , , , , , , , , , , , ,

0001 0001 0001 0001 0001 0001 0001 0001 0001 0001 0001 0005 0001 0001 0001 0001

The specicity was low particularly in case of ulnar and common peroneal nerves (30 and 40% respectively) while for all other sensory and motor nerves it ranged between 60 and 80%. Frequent enlargement of ulnar and peroneal nerves is known in leprosy and assessing enlargement especially of these nerves is considered an important aspect of routine clinical examination.22,38 Sensitivity was highest in case of ulnar nerve (. 80%), a nding very similar to that recorded by Brown et al. (1996).33 Between NP and SNC tests, concordance was seen in 63% of nerves (Table 4). Similarly between NP and MNC tests, concordance was seen in 60% of nerves (Table 4). In the discordant group, a substantial proportion (20%) of nerves particularly ulnar (42%) and common peroneal (43%) showed enlargement ranging from 2 3 , but normal conduction by SNC/MNC (Table 4). Two possible explanations are; (a) conduction of a palpably enlarged nerve may remain within the normal range, as long as sufcient number of normally functioning nerve bres remains. To this effect segmental involvement and enlargement of one or two fascicles of a large nerve trunk, particularly at the tuberculoid and borderline tuberculoid end of the spectrum is fairly common and has been documented.44 (b) The second possibility is that NP may be giving false positive results. However in our study all the tests were ascertained and nally documented by an experienced clinician. A signicant proportion of sural (29%), median (20%), radial cutaneous (14%) and ulnar motor (12%) nerves were normal by palpation but abnormal by NCS (Table 4). There could be two explanations for this discordance; (1) Practical difculties in palpating these nerves due to obesity, oedema and size of the contralateral nerve which are known to inuence the decision on nerve thickness,21,22 or (2) Nerve thickening may be less marked in these patients. At the lepromatous end of the spectrum the nerve enlargement is often more diffused and less marked. In our study, the (Kw) values of inter examiner repeatability varied between 054 073 indicating fair to good agreement yet unsatisfactory (Table 2). The agreement on palpation of the peroneal nerve (073) was better than that on palpation of the ulnar (062)

Correlation between clinical and electrophysiological tests

45

nerve but for both the nerves, it was still fair to good agreement. This is similar to the ndings of previous study conducted in India by Kolappan et al. (1995).22 In contrast Shumin Chen et al. (2006)23 found that palpation of ulnar nerve was better as compared to peroneal nerve. Study by Kolappan et al. (1995)22 also showed that the Kappa statistic (Kw) values of inter examiner repeatability between the medical ofcer and PMW for palpating enlarged nerves varied between 045 and 069 respectively. In view of this, nerve palpation as a standalone test, may not be very reliable. However its advantages are it only requires a pair of experienced hands and is one of the cardinal signs of the disease having diagnostic value. In the eld leprosy screening is usually done by paramedical workers. Standardising PMWs and selecting those with high levels of inter-tester agreement would therefore be important to minimise the measurement errors at eld level and would be useful for research projects as well. Sensitivity of both MF testing and VMT assessment were uniformly low but they showed very good specicity. Like NP the sensitivity of MF testing and VMT assessment was signicantly higher (P , 0001) in patients who presented with reaction and/or neuritis (Gr. A; MF testing 46%, VMT assessment 20%) as compared to patients with no reaction (Gr. B; MF testing 32%, VMT assessment 9%) (Table 6). As has been documented earlier,45 peripheral nerve damage can be demonstrated in almost all the patients with leprosy particularly multibacillary. However reaction episode increases the severity of the problem in terms of extent of nerve damage. A good concordance between MF testing and SNC tests was seen in 65% of nerves and between VMT assessment and MNC concordance was seen in 60% of nerves (Table 4). Common discordance was in terms of normal MF testing and VMT assessment against an abnormal NCS nding, implying that NCS is probably detecting pre-clinical neuropathy. This is in agreement with an earlier documented
Table 6. Nerve wise sensitivity and specicity of clinical tests taking NCS as reference test (gold standard) tested in patients presenting with reaction (Gr.A, n 141) and without reaction (Gr.B, n 216) Sensitivity (%) Clinical Tests Sensory nerves Ulnar NP MF Median NP MF R. cut NP MF Sural NP MF Motor nerves Ulnar NP VMT Median NP VMT C. per NP VMT P. tibial NP VMT Gr.A Gr.B Chi
2

Specicit (%) Sig. P , 0001 P , 005 P 242 P , 0005 P 260 P , 000 P , 0005 P , 0025 P , 001 P , 0005 P , 005 P , 0001 P 029 P 001 P 029 P , 005 Gr.A Gr.B Chi2 Sig. P , 0025 P , 005 P 004 P 111 P 020 P 032 P 090 P 069 P 123 P 161 P 0098 P 264 P 0016 P 035 P , 001 P , 0025

93 51 55 45 75 50 69 42 84 31 49 30 83 05 74 03

84 39 46 28 67 32 55 32 74 18 38 0 73 0 71 0

1138 370 855 1115 910 424 665 941 31 1947

28 84 79 91 57 90 75 90 28 89 72 89 40 99 54 99

38 90 80 94 59 91 81 94 34 93 74 93 41 99 65 99

438 307

379

587 397

Key: Gr. A patients with reaction and/or neuritis, Gr. B patients with no reaction.

46

F. A. Khambati et al.

nding that . 30% of bres in a nerve are destroyed before any clinical sensory decit can be elicited.46 This implies that, even in the face of normal clinical testing, latent nerve damage exists and the disease is well established in the nerve before the diagnosis has been made. On the other hand, the high specicities of MF testing and VMT assessment show that loss of sensation or strength is very likely to be accompanied by nerve conduction abnormalities. It is interesting to note that, a small proportion of nerves impaired by MF testing (5%) and VMT assessment (1%) showed normal SNC and MNC respectively (Table 4). A study by van Brakel et al. (2005)34 also had similar observations. Therefore, MF testing/VMT assessment probably detects functionally important impairment not necessarily reected in nerve conduction, which is studied over a relatively short segment of the peripheral nerve. Some fast conducting bres may have remained unaffected and produced a normal electrophysiological test.2 Demonstration of healthy nerve bers within anaesthetic areas in tissue sections47 further suggest that probably some bres survive long enough to conduct the response, or that the response may be due to discharge from regenerating nerve bres.48 Maximum concordance between VMT assessment and MNC in terms of abnormal by both was seen in case of ulnar nerve (14%) (Table 4). Remaining motor nerves showed abnormal conduction by MNC but normal muscle strength by VMT assessment indicating that results of nerve conduction tests may not necessarily correlate well with motor function abnormality. As in case of sensory nerves, the impairment observed by MNC may also be preclinical and may translate into motor weakness late in the disease.33,49 In the case of peroneal and tibial nerves, the VMT assessment of the foot primarily tests strength of the tibialis anterior, extensor of toes, peroneus longus, and small muscles respectively, while MNC recordings were taken from the extensor digitorum brevis and abductor hallucis brevis respectively. Thus even though VMT assessment as a test is very reliable, it has very poor sensitivity for assessing motor impairment and there is a need for a more sensitive test. In our study the reliability of MF testing varied between 060 and 073 while for VMT assessment varied between 08 and 097 indicating good to very good agreement respectively between the testers (Table 2). A study by Roberts et al. (2007)15 has concluded that, the reliability of MF testing and VMT assessment was very good and good with Kappa statistics (Kw) value of . 075 and . 071 for MF testing and VMT assessment respectively, obtained between different testers and gold standard. Since monolament grades are calculated as the sum of points down at each of three or four sites within the same nerve distribution, the resulting totals or grades range from 0 to 15 or 16. In contrast the VMT assessment scores are based on assessment of a single movement and range from 5 to 0. Therefore some variation in monolament scores is expected, resulting in a reduced level of agreement within one point. By NCS, sural and ulnar across the sulcus were found to be most frequently impaired sensory and motor nerves respectively (Table 3). An earlier study carried out at this centre50 and a recent large cohort study by van Brakel et al. (2005)34 also record sural as the most frequently affected sensory nerve in leprosy. Studies by Sabin & Swift (1984),51 and Kumar et al. (2004)52 note that paralysis is more common in the ulnar distribution than other nerves in leprosy. Cool temperature facilitating bacterial multiplication, repetitive trauma and bony impingement are implicated as some of the underlying causes. Overall by MF testing and VMT assessment, 53% of patients showed impaired nerve function (Table 8), whereas 92% showed abnormal nerve function by NCS. In terms of number of nerves also, proportion of sensory and motor nerves showing abnormality by SNC (48%) was two fold higher than MF testing (23%) and MNC (37%) was seven fold higher

Correlation between clinical and electrophysiological tests

47

Table 7. Sensitivity and specicity of clinical tests (nerve wise) (used singly and in combination) taking NCS as reference test (gold standard) Clinical tests MF VMT NP MF VMT MF NP VMT NP MF VMT NP Sensitivity No. (%) 529/1392 (38) 147/1060 (14) 1637/2438 (67) 850/2121 (40) 994/1374 (72) 738/1047 (70) 1787/2492 (72) Specicity No. (%) 1338/1478 (91) 1750/1823 (96) 1838/3268 (56) 2909/3296 (88) 883/1554 (57) 948/1799 (53) 1769/3341 (53)

Key: MF monolaments testing, VMT voluntary muscle testing, NP nerve palpation.

than VMT assessment (6%) (Table 3) proving NCS is by far more sensitive than MF testing and VMT assessment.45 Combining nerve palpation (NP) with either monolaments (MF) or voluntary muscle testing (VMT) increased the detection sensitivity of clinical tests by two fold (72% nerve wise and 93% patient wise) and was comparable to that detected by nerve conduction studies (Tables 7 and 8). However on combining the three tests, the specicity reduced to 53% nerve wise and 45% patient wise which is low in comparison to MF testing and VMT assessment used singly (. 90% specicity). The role of nerve palpation as a screening test in leprosy diagnosis is undeniable. In that setting, the high sensitivity of NP is of paramount importance. However tests with higher specicity (e.g. MF testing and VMT assessment) are needed to avoid unnecessary treatment. The enlarged nerves are shown to be a risk factor for subsequent nerve function impairment.53 It is therefore advisable that patients with enlarged nerves and/or impairment as assessed by MF testing and VMT assessment should be followed up and examined regularly and appropriately treated for prevention and reversal of nerve damage and disability. Under eld conditions, combining NP with MF testing and VMT assessment could prove to be a practical as well as a sensitive tool for diagnosing leprosy disease and detecting nerve damage in leprosy patients. Also those working as diagnostic assistants in leprosy clinics can be trained to carry out these tests efciently. Electrophysiology could be recommended only
Table 8. Sensitivity and specicity of clinical tests (patient wise n 357) (used singly and in combination) taking NCS as reference test (gold standard) Clinical tests MF VMT NP MF VMT MF NP VMT NP MF VMT NP Sensitivity No. (%) 152/320 (48) 97/269 (36) 296/328 (90) 173/328 (53) 303/328 (92) 300/328 (91) 307/328 (93) Specicity No. (%) 27/29 (93) 27/29 (93) 13/29 (45) 27/29 (93) 13/29 (45) 13/29 (45) 13/29 (45)

Key: MF monolaments testing, VMT voluntary muscle testing, NP nerve palpation.

48

F. A. Khambati et al.

under special circumstances for further conrmation or for more detailed analysis or as a research tool.

Conclusions . Both MF testing and VMT assessment are very specic, but less sensitive for assessing nerve damage. . NP is less specic but more sensitive, as compared to MF testing and VMT assessment. . NP in combination with MF testing and VMT assessment, when undertaken by trained workers improves the sensitivity of assessing nerve damage two fold. . Nerve function assessment (NFA) is important for the diagnosis of leprosy per se, but has an even more crucial role in the prevention and early treatment of nerve function impairment (NFI). The nerve function assessment using clinical tests have been practiced and supported by the ILEP in the research eld,54 and our study ndings provides further support to the ILEP guidelines for assessment of NFI. These tests could serve as the most useful clinical tools for detecting nerve impairment at eld level.

Acknowledgements The authors are grateful to all the patients who participated in this study. We are thankful to Dr. Ganapati and Dr. Pai from Bombay Leprosy Project (Chunnabhatti, Sion) for referral of patients. Our thanks are due to Mr. Uday Thakkar and management of Kusthrog Nivaran Samiti, Shantivan Panvel for referrel of patients and providing an outpatient clinic facility at Panvel. Thanks to Dr. Shubhada Pandya for her valuable guidance in the nerve conduction studies and manuscript preparation and Mr. K. Gandewar, Biostatistician, Department of Preventive Social Medicine, Sion Municipal Hospital, Mumbai for statistical help. Funding: This study was funded by the American Leprosy Mission, ILEP Grant code number 7.01.03.46.

References
1 2 3 4 5 6 7 8 9

Lewis S. Reproducibility of sensory testing and voluntary muscle testing in evaluating the treatment of acute neuritis in leprosy patients. Lepr Rev, 1983; 54: 23 30. Kaplan M, Gelber RH. Evaluation of testing modalities for peripheral neuropathy in lepromatous Hansens disease. Phy Therapy, 1985; 65: 16621665. Owen BM, Stratford CJ. Assessment of the methods available for testing sensation in leprosy patients in a rural setting. Lepr Rev, 1995; 66: 55 62. WHO Expert committee on leprosy. Fourth report, WHO technical series report, 1970; 459: 29. Patil KM, Sudhakar Babu T, Oomen PK, Srinivasan H. Foot pressure measurement in leprosy and foot wear design. Ind J Lepr, 1986; 58: 357366. Palande DD, Bowden REM. Early detection of damage to nerves in leprosy. Lepr Rev, 1992; 63: 6072. Srinivasan H, Stumpe B. Value of thermal sensibility testing in leprosy diagnosis in the eldeld trial of a pocket device. Lepr Rev, 1989; 60: 317 326. Moberg E. Objective methods of determining functional value of sensibility in the hand. J Bone Jt Surg, 1958; 40: 458466. Debi BP, Mohanty HC, Tripathy N, Tompe DB, Sarangi BK. Arteriographic pattern of plantar ulcers in lepromatous leprosy study of 20 cases. Lepr India, 1980; 52: 429 432.

Correlation between clinical and electrophysiological tests

10

49

11

12 13

14

15

16

17

18

19 20

21

22

23

24 25

26 27

28

29

30

31

32

33

34

35

36 37

38 39

40

41 42

Bell-Krotoski JA, Tomancik E. The repeatability of testing with SemmesWeinstein monolaments. J Hand Surg, 1987; 12: 155161. Bell-Krotoski JA. Pocket laments and specications for the SemmesWeinstein monolaments. J Hand Ther, 1990; 3: 2631. Brandsma JW. Monitoring motor nerves function in leprosy patients. Lepr Rev, 2000; 71: 258 267. Sangyeoup L, Hyeunho K, Sanghan C et al. Clinical usefulness of the Two- site test for detecting diabetic peripheral neuropathy. J Korean Med Sci, 2003; 18: 103107. van Brakel WH. Peripheral neuropathy in leprosy. The continuing challenge. Thesis, University of Utrecht, The Netherlands, 1994. Roberts AE, Nicholls PG, Maddali P, Van Brakel WH. Ensuring inter-tester reliability of voluntary muscle testing and monolament sensory testing in the INFIR Cohort study. Lepr Rev, 2007; 78: 122130. Anderson AM, Croft RP. Reliability of Semmes Weinstein monolament and ballpoint sensory testing and voluntary muscle testing in Bangladesh. Lepr Rev, 1999; 70: 305 313. Brandsma JW, van Brakel WH, Anderson AM et al. Intertester reliability of manual muscle strength testing in leprosy patients. Lepr Rev, 1998; 69: 257 266. Tzourio C, Henry P, Boucher P et al. Lepre lepromateure arguments cliniques et electrophysiologiques en faveur dune multiwrite axonale. Acta Leprol, 1989; 7: 5156. Srinivasan H. Nerve damage, surgery and rehabilitation in leprosy. Trop Med Parasitol, 1990; 41: 347349. Pannikar VK, Arunthathi S, Chacko CJG, Fritschi EP. A clinico-pathological study of primary neuritic leprosy. Lepr India, 1983; 55: 212 221. McDougall AC. The clinical examination of peripheral nerves in leprosy. Correspondence. Ind J Lepr, 1996; 68: 378 379. Kolappan C, Selvaraj R, Khudoos A et al. Repeatability of nerve thickness assessment in the clinical examination for leprosy. Lepr Rev, 1995; 66: 224228. Shumen C, Qihua W, Tongsheng C, Ming Z. Inter-observer reliability in assessment of sensation of skin lesion and enlargement of peripheral nerves in leprosy patients. Lepr Rev, 2006; 77: 371 376. McLeod JG, Hargrave JC, Walsh JC et al. Nerve conduction studies in leprosy. Int J Lepr, 1975; 43: 2131. Singh T, Kaur S, Kumar B. A study of motor and sensory nerve conduction in leprosy. Indian J Med Res, 1977; 65: 632 639. Campion D. Electrodiagnostic testing in Hand Surg. J Hand Surg, 1996; 21: 947956. Chaudhry V, Cornblath DR, Mellits ED et al. Inter and intra examiner reliability of nerve conduction measurements in normal subjects. Ann Neurol, 1991; 30: 841843. Pinheiro D, Manzano G, Nobrega J. Reproducibility in nerve conduction studies and F-wave analysis. Clinical Neurophysiology, 2008; 119: 20702073. Nasseri K, Strijers RLM, Dekhuijzen LS et al. Reproducibility of different methods for diagnosing and monitoring diabetic neuropathy. Electromyogr Clin Neurophysiol, 1998; 38: 295 299. Buchtlal F, Rosenfalck A, Behse F. Sensory potentials of normal and diseased nerves. In: Dyck PJ, Thomas PK, Lambert EH (eds). Peripheral neuropathy WB Saunders Co, Philadelphia, 1975; Vol. 1: pp. 442464. Gilliatt RW, Sears TA. Sensory nerve conduction studies in the early recognition of nerve disorders. Muscle Nerve, 1978; 1: 360 367. Dawson GD. The relative excitability and conduction velocity of sensory and motor nerve bers in man. J Physiol (Lond), 1956; 131: 436451. Brown TR, Kovindha A, Wathanadilokkol U et al. Leprosy neuropathy: Concordance of clinical and electrophysiological tests. Ind J Lepr, 1996; 68: 115. Van Brakel WH, Nicholls PG, Das L et al. The INFIR Cohort study: assessment of sensory and motor neuropathy in leprosy at baseline. Lepr Rev, 2005; 76: 277295. Van Brakel WH, Nicholls PG, Wilder-Smith E et al. Early diagnosis of neuropathy in leprosy- comparing diagnostic tests in a large prospective study (The INFIR cohort study). PLOS, 2008; 2: e 212. WHO expert committee on leprosy. WHO technical report series, 1978; No. 874. Croft RP, Richardus JH, Nicholls PG, Smith WCS. Nerve function impairment in leprosy: design, methodology and intake status of a prospective cohort study of 2664 new leprosy cases in Bangladesh: The Bangladesh Acute Nerve Damage Study. Lepr Rev, 1999; 70: 140 159. Donaghy M. Enlarged peripheral nerves. JNNP, 2003; 3: 4045. Anderson AM, van Brakel WH. Age specic normal thresholds for sensibility testing with monolaments in a Nepali population. Int J Lepr, 1998; 66: 69 A. van Brakel WH. Peripheral nerve function assessment. In: Schwarz RJ, Brandsma JW (eds). Surgical reconstruction and rehabilitation in leprosy and other neuropathies EKTA publishers, Nepal Kathmandu, 2004. Altman D. Practical statistics for medical research Chapman and Hall, London and New York, 1991. DeLisa JA, Hang JL, Baran EM et al. Temperature effects on nerve conduction velocities and latencies. Manual of nerve conduction velocity and clinical neurophysiology Raven, New York, 1994, pp. 121.

50
43

F. A. Khambati et al.
Van Brakel WH, Nicholls PG, Das L, Barkataki P. The INFIR Cohort study: Investigating prediction, detection and pathogenesis of neuropathy and reactions in leprosy. Methods and baseline results of a cohort of multibacillary leprosy patients in North India. Lepr Rev, 2005; 76: 1434. Antia NH, Mehta L, Shetty VP, Irani FF. Clinical, electro-physiological, quantitative histologic and ultrastructural studies of the index branch of the radial nerve in leprosy. 1. Preliminary report. Int J Lepr, 1975; 43: 106 113. Khambati FA, Shetty VP, Ghate SD et al. The effect of corticosteroid usage on the bacterial killing, clearance and nerve damage in leprosy. A prospective cohort study: Part 1 study design and baseline ndings of 400 multibacillary patients. Lep Rev, 2008; 79: 121. Antia NH, Shetty VP, Mehta LN. Study of evolution of nerve damage in leprosy. Lepr India, 1980; 52: 48. Jayaraj AP, Chaudhary DS. Epitheloid and sub-epithelial innervations in lepromatous leprosy. Lepr India, 1960; 32: 167169. Marques W, Norma T, Foss MD et al. Near nerve potential in lepromatous leprosy. Muscle Nerve, 2003; 28: 460 463. Dyck PJ. Quantitating severity of neuropathy. In: Dyck PJ, Thomas PK, Grifn J, Low PA, Poduslo JF (eds). Peripheral neuropathy Saunders, Philadelphia, 1993, pp. 686 697. Donde SV, Shah A, Antia NH. Nerve Conduction in leprosy: in vivo and in vitro study. Lepr India, 1983; 55: 12 21. Sabin TD, Swift TR. Leprosy. In: Dyck PJ, Thomas PK, Lambert EH, Bunge R (eds). Peripheral neuropathy 2nd edn., Saunders, Philadelphia, 1984, pp. 19551987. Kumar A, Girdhar A, Girdhar BK. Nerve thickening in leprosy patients and risk of paralytic deformities: a eld based study in Agra, India. Lepr Rev, 2004; 75: 135142. Nicholls PG, Croft RP, Richardus JH et al. Delay in presentation, an indicator for nerve function status at registration and for treatment outcome the experience of the Bangladesh Acute Nerve Damage Study cohort. Lepr Rev, 2003; 74: 349 356. van Brakel WH, Saunderson P, Shetty VP et al. International workshop on neuropathology in leprosy consensus report. Lepr Rev, 2007; 78: 416 433.

44 45

46 47 48 49 50 51 52 53

54