For the side-to-side confrontational testing of the abductor digiti minimi, which is innervated by the ulnar nerve, the patient holds the hands approximately eight to ten inches (twenty to twenty-five centimeters) from the face, with the palms facing the face and the fingertips at about eye level. Both long fingers are held vertical to the horizontal plane, and the remaining fingers are abducted to create an angle of approximately 90 degrees between the two little fingers when the tips of the little fingers are brought together. The patient then pushes the hands together while simultaneously using the little fingers to prevent this action (Fig. 2-A). A negative finding on the test—that is, equal strength on each side—is indicated by a symmetrical giving-way of the ring and little fingers as they come together (Fig. 3). If only one finger gives way, the test is positive (Figs. 4-A and 4-B). Abduction of the thumb is tested in a similar manner (Fig. 2-B), with giving-way on one side considered a positive result.
Two hundred and seventeen consecutive outpatients who were referred to an electrodiagnostic laboratory at a Veterans Affairs Medical Center were evaluated. The first 100 patients to be referred were included in the series regardless of whether they had symptoms in a lower or upper extremity. This was done in order to estimate the rate of positive test results in individuals who do not have symptoms in an upper extremity. Subsequently, only those patients who had symptoms in an upper extremity were included. The criteria for exclusion of a patient included the lack of an upper extremity, a digit, or a part of a digit; a joint deformity or arthritis; the lack of full extension of the metacarpophalangeal joints or of the proximal or distal interphalangeal joints; the lack of at least 30 degrees of symmetrical active abduction of both little fingers; the loss of sensation in the little finger; extensive weakness of the hand; and an inability to understand the instructions.
Giving-way on one side was recorded as a positive test. The test was not performed in a blinded manner. I taught all of the patients how to participate in the examination and I conducted the test. The findings of the manual strength tests were recorded before the performance of other evaluations. These included electrodiagnostic studies of the median and ulnar motor nerves as well as of the median, ulnar, and radial sensory nerves at the wrist; a needle examination of the cervical paraspinal muscles; and strength-testing of a sample of the muscles of the upper extremity, so that all of the major nerve roots as well as peripheral nerves and the trunks and cords of the brachial plexus were included in the testing. The contralateral side was evaluated, and additional studies were performed if they were indicated clinically.
Fifty of the 217 patients were referred because of symptoms in a lower extremity. As those patients did not have a positive finding for weakness on either standard muscle testing or confrontational testing, they were not included in the remaining evaluations. There were 157 men (94 per cent) and ten women (6 per cent) who were referred because of a problem related to the upper extremity. The mean age (and standard deviation) was 52.6 ± 14.5 years (range, twenty-two to eighty-four years). Nine patients were excluded because they were unable to perform the tests for a variety of reasons. Strength-testing of the muscles innervated by the ulnar nerve (the abductors of the index and little fingers) was performed on all 158 patients. Testing of the muscles innervated by the median nerve (the abductors of the thumb) was performed on 131 patients.
The findings on the electrodiagnostic studies were normal in eighty-four patients. Positive findings included median neuropathy at the wrist (sixty-eight patients), ulnar neuropathy at the elbow (forty-six patients), caudad cervical radiculopathy (twenty-nine patients), mid-cervical radiculopathy (seven patients), brachial plexopathy and peripheral neuropathy (three patients each), and ulnar neuropathy at the Guyon canal (two patients). Some patients had more than one diagnosis. The number of patients in each group was too small to determine whether the patients who had more than one diagnosis had a greater likelihood of having a positive result on strength-testing than those who had a single diagnosis.
Analysis of the Muscles Innervated by the Ulnar Nerve
Seventy-four of the patients had electrodiagnostic evidence consistent with weakness of the muscles innervated by the ulnar nerve (caudad cervical radiculopathy, low brachial plexopathy, or ulnar neuropathy). Sixty-one of these patients had a positive finding on the confrontational test, and only ten had a positive finding on the standard manual test. The eighty-four patients who had normal findings on the electrodiagnostic studies had normal findings on the manual and confrontational strength tests. The sensitivity of the standard manual test for weakness of the muscles innervated by the ulnar nerve as compared with the positive results on the electrodiagnostic studies was 14 per cent, and the specificity was 100 per cent. The predictive value of a positive result was 100 per cent (ten divided by ten), and the predictive value of a negative result was 57 per cent (eighty-four divided by 148).
The sensitivity of the confrontational test of the muscles innervated by the ulnar nerve was 82 per cent, and the specificity was 100 per cent. The predictive value of a positive result was 100 per cent (sixty-one divided by sixty-one), and the predictive value of a negative result was 87 per cent (eighty-four divided by ninety-seven). The sensitivity and specificity of the confrontational test were not associated with age. Thirteen patients had a false-negative result. Three of those patients had bilateral ulnar neuropathy; five, ulnar neuropathy at the elbow; and five, cervical radiculopathy.
Analysis of the Muscles Innervated by the Median Nerve
Sixty-nine patients had electrodiagnostic evidence of weakness of the muscles innervated by the median nerve. The findings included caudad cervical radiculopathy, low brachial plexopathy, and median neuropathy. Eleven patients had bilateral median neuropathy at the wrist. Four of the sixty-nine patients had a positive result on the standard manual test of muscle strength, and eleven had a positive result on the confrontational test. Sixty-two patients had no electrodiagnostic evidence of weakness of the muscles innervated by the median nerve, and the findings of the standard and confrontational tests were also negative.
The sensitivity of the standard manual test for evaluating the strength of abduction of the thumb as compared with electrodiagnostic evidence of a neural lesion was 6 per cent, and the specificity was 100 per cent. The predictive value of a positive result was 100 per cent (four divided by four), and the predictive value of a negative result was 49 per cent (sixty-two divided by 127). The sensitivity of the confrontational test for weakness of the median nerve was 16 per cent, and the specificity was 100 per cent. The predictive value of a positive result was 100 per cent (eleven divided by eleven), and the predictive value of a negative result was 52 per cent (sixty-two divided by 120).
Electrodiagnostic testing is commonly used to diagnose carpal tunnel syndrome, ulnar neuropathy, and cervical radiculopathy, which are among the most frequent causes of weakness of the intrinsic muscles of the hand6-8. The results are often used to assess the reliability of other tests2,9,10, as electrodiagnostic studies document the functional status of the nerves. In fact, some authors have reported that the association between the findings of an electrodiagnostic study and operative evidence of disc herniation is equal to or better than that of the findings on a myelogram, a computerized tomography scan, or a magnetic resonance imaging study6. Electrodiagnostic studies have a low rate of false-positive results. In addition, they can be used to identify peripheral neuropathy and to differentiate neurological from musculoskeletal pain5. In the present study, the results of electrodiagnostic studies were used as the standard against which the test results were compared.
The limitations of the present study include the absence of universally accepted standard criteria for diagnosing nerve injury and the generalization of results from a group of patients that was predominantly male and somewhat older than the general population.
The data indicated that standard manual strength-testing of the muscles innervated by the median nerve and those innervated by the ulnar nerve was not diagnostically sensitive, possibly because mild weakness could not be detected. Furthermore, confrontational testing was not useful for assessing the strength of the abductors of the thumb, possibly because the abductor pollicis longus compensates for weakness of the abductor pollicis brevis. However, the sensitivity, specificity, and predictive value (for both positive and negative results) of the confrontational test for assessing weakness of the muscles innervated by the ulnar nerve were excellent.
The proper application of the confrontational test and an understanding of its limitations are important, as the test is a non-specific screening tool for assessing an imbalance in muscle strength due to a variety of disorders. The test requires the cooperation of the patient, and it may be difficult to perform when a patient has a problem with communication due to a factor such as a hearing loss, a language barrier, or sometimes a limited intellect. When it is used appropriately (as has been described), the test appears to be a valuable screening tool for weakness with a neurological etiology. I have found the test to be useful in clinical practice.