Symptoms of 100 Patients With Electromyographically Verified Carpal Tunnel Syndrome.

Stevens JC, Smith BE, Weaver AL, et al (Departments of Neurology and Neurologic Surgery and Division of Regional and International Medicine, Mayo Clinic, Scottsdale, Ariz; Section of Biostatistics, Mayo Clinic and Mayo Foundation, Rochester, Minn), Muscle Nerve. 1999;22:1448-1456.

The primary purpose of this study was to determine the distribution of symptoms in patients-, with symptomatic, electromyographically verified carpal tunnel syndrome (CTS). The secondary purpose was to assess the usefulness of a hand symptom questionnaire (HSQ) and a hand symptom diagram (HSD) in the diagnosis of CTS.

Subjects were recruited from patients who were referred to an electromyography (EMG) laboratory for evaluation of CTS or upper-extremity paresthesia. Under standardized temperature conditions, nerve conduction studies of the median and ulnar nerve were performed and compound muscle action potential amplitudes were determined for all patients. Carpal tunnel syndrome was diagnosed if the median palmar absolute latency was prolonged and the median-ulnar latency difference was greater than 0.4 milliseconds or, in the cases in which the median palmar latency was less than 2.3 milliseconds, if the median ulnar difference exceeded 0.4 milliseconds. Subjects were included in the study if EMG findings supported the diagnosis of CTS, if the subject had symptoms of hand paresthesia, and if a medical record and interview revealed no other medical condition that could be causing the upper-extremity paresthesias. Evidence of ulnar nerve involvement as determined through nerve conduction studies was a reason for exclusion.

The HSQ and the HSD were completed by the 100 subjects who met the inclusion criteria. Individuals with upper-extremity paresthesia who did not have CTS also filled out the questionnaire and diagram. Two investigators graded all diagrams and questionnaires and placed each subject into a "distal median," "ulnar," or "other" group. Questionnaires were then given in random order to an internist, neurologist, and neurosurgeon. Using the questionnaires, each physician rated each hand as "not CTS," "undecided," "possible CTS," or "definite CTS." The process was repeated again using the hand diagrams.

Frequencies, percentages, means, standard deviations, and ranges of variability were used to summarize data. For the purpose of analysis, physician ratings were collapsed from the initial 4 categories to 2 categories: "CTS" ("possible CTS" or "definite CTS") and "no CTS" ("undecided" or "not CTS"), and positive predictive value and negative predictive values were estimated.

The results indicated that the most common symptoms (the descriptions of which appeared on the HSQ) were the following: paresthesia when reading, relief of sensory symptoms by shaking the hands, being awakened from sleep by paresthesia or pain more than once per week, having numb hands on awakening in the morning, and paresthesia when driving. Symptoms were confined to the fingers innervated by the median nerve in 40% (HSQ) and 45% (HSD) of the subjects. The middle finger was chosen as the most affected digit more than twice as often as the index finger or thumb. The little finger, though frequently affected, was identified as the most severely affected finger in only 2.5% of the hands examined. Symptoms were mostly palmar in 69.2% of the subjects (HSQ), although 10.7% indicated that their symptoms were mostly dorsal. Symptoms proximal to the fingers were found most frequently in the palm (56.0%) followed by the wrist (30.2%), forearm (21.4%), elbow (10.1%), arm (5.7%), and shoulder (1.9%).

Analysis revealed that the average sensitivity of the physicians using the HSD to predict CTS was 71.3%. This sensitivity value was 74.0% for the HSQ and 88.0% when the HSQ and HSD were combined. The average positive predictive values for the HSD, HSQ, and the combined HSD and HSQ were 83.4%, 82.2%, and 81.0%, respectively. The average negative predictive values for the HSD, HSQ, and the combined HSD and HSQ were 50.0%, 52.9%, and 65.8%, respectively.

The authors concluded that a wide range of sensory symptoms can be seen in CTS. These symptoms are most commonly present in the median and ulnar digits. Symptoms usually involve 3 or 4 digits, frequently involve the dorsum of the hand, and can radiate proximally into the palm and wrist. The authors believed that combining the questions on the HSD and HSQ that are related to aggravating and relieving factors and the most severely affected finger provides a quick and sensitive method of screening for CTS. Although the combined data of the HSD and HSQ were sensitive tot the diagnosis of CTS, the researchers reported that the accuracy was only 74.5% to 80.6%. The authors suggested that confirmation of CTS by EMG is desirable.

Nancy R Talbott, PT University of Cincinnati Cincinnati, Ohio