ANN ARBOR --- A nerve test widely used to diagnose carpal tunnel syndrome should not be relied upon to give a "yes-no" answer to the question of whether a person has the painful hand condition, a new University of Michigan study says.
In fact, the results show, people without any typical carpal tunnel syndrome symptoms can register abnormal results on nerve conduction tests depending on how the test is performed---and not all of those who complain of symptoms like wrist pain and tingling fingers will show nerve damage during the test. What's more, the test can produce very different results depending on who performs it.
For the estimated 850,000 Americans who visit their doctors each year complaining of carpal tunnel-like symptoms, and the 260,000 who have corrective wrist surgery annually, the study shows the importance of being examined by an experienced doctor using a range of tests.
Published in a recent issue of the journal Muscle and Nerve, the U-M study is the first of its size and kind. It was performed by a team of experts from several areas of the university, who conclude that doctors should use a continuous scale to evaluate test results, rather than the current practice of choosing a cutoff point to distinguish between normal and abnormal results.
"When job placement or surgery hangs in the balance, the reliability of a test like this becomes even more important," says co-author and U-M health scientist Deborah Salerno, Ph.D. "The results of our study were mixed, and, at times, the difference in reliability was striking. For reliable results, you want to have an experienced examiner do the test, and use appropriate criteria to see if the results are normal."
The study looked at 158 workers, mostly women, whose work includes a large amount of time spent at a computer keyboard. Not all the subjects complained of symptoms typical of carpal tunnel syndrome. Each underwent two rounds of nerve conduction tests, spaced three weeks apart, that electrically stimulated two nerves in the wrist and examined the difference in response time.
Two U-M physicians performed the tests. In the first round, each physician tested both wrists of each subject. In the second round, three weeks later, tests were performed by one of the physicians in the dominant wrist only.
The examiners tested two nerves for abnormalities: the median nerve, which runs down the middle of the arm and wrist and controls the palm, the thumb, and several fingers; and the ulnar nerve, which runs through the outer arm and wrist to control the back of the hand and the small and ring fingers.
The median nerve is more affected by carpal tunnel syndrome, in which overuse of the hand and wrist causes tissue swelling. This, in turn, narrows the gap between the carpal, or wrist, bones and ligaments through which the nerve runs, causing pressure on the nerve. Women between ages 30 and 60 are most commonly affected.
Although a variety of tests can detect conditions like carpal tunnel syndrome, nerve conduction studies have been used as the gold standard. The test uses a small jolt of electricity to stimulate each nerve, then measures the time it takes for the stimulus to reach the muscle. A second stimulation of the same nerve at a different spot, or of the other major nerve, allows the lag time, known as the latency, to be compared. This lets doctors spot the existence and the extent of nerve damage in different locations.
In the U-M study, the results recorded by each examiner for each patient were compared, as were each examiner's results from both their exams of the workers. In both cases, results showed that certain tests were better than others. Both between examiners and within the same examiner, median nerve tests were more reliable than ulnar nerve tests.
A key question was how differences between examiners would affect a diagnosis of carpal tunnel syndrome. That is, would the first examiner rate a test result as "normal" when the second rated it as "abnormal"?
To explore this, the researchers used two cutoff points to define "normal." The first defined a difference of 0.5 milliseconds (msec) or more between the median and ulnar tests in the same wrist as abnormal. The second used a difference of 0.8 msec or more. Results showed that between examiners, both cutpoints had good to excellent reliability.
However, using the two definitions of normal resulted in marked differences in results. Each examiner had higher reliability with the higher cutoff point. "Presumably, the higher cutpoint reflected more serious nerve problems. This would make the abnormalities easier to notice, and results more reliable," says Salerno.
In light of this, the researchers concluded that a system with more ordered categories (such as normal, equivocal, mild, or definite abnormality) would better describe results, rather than using absolute cutoff points to define "normal."
Interestingly, between the two examiners, 36 (22 percent) of the workers in the first round had abnormal results. In contrast, over half (53 percent) of the workers in the study reported pain, burning, tingling, or numbness in the hands or wrists, which are typical symptoms of certain nerve disorders.
"Nerve conduction studies have been considered the gold standard for carpal tunnel syndrome, but abnormal results alone, without symptoms, do not define the disorder," says Salerno. "Both symptoms and physical findings are critical for accurate test interpretations."
Given that a diagnosis of carpal tunnel syndrome can lead to work-area and duty adjustments, prescription of medication, or surgery, accurate and reliable testing---using tests beyond nerve conduction studies---should be performed, the researchers conclude.
Co-authors of the study and patient examiners were Robert Werner, M.D., associate professor of physical medicine and rehabilitation (PM&R); and James W. Albers, M.D., professor of neurology, both of whom also hold appointments in the Department of Environmental Health Sciences at the U-M School of Public Health. Other co-authors included Mark P. Becker, Ph.D., professor of biostatistics in the School of Public Health; and Thomas J. Armstrong, Ph.D., professor of industrial and operations engineering, and of biomedical engineering, in the College of Engineering.
Cite This Page: