Interview data versus questionnaire data in the diagnosis of carpal tunnel syndrome in epidemiological studies

Abstract

Aims. In most epidemiological studies on carpal tunnel syndrome (CTS), the case definition is based on questionnaire data with or without neurophysiological testing. The aim of this study was to test if the use of questionnaire data results in mis-classification of cases. Methods. In two studies, involving 940 and 311 participants, respectively, the people indicating CTS symptoms (tingling) in a questionnaire were clinically interviewed. In a subpopulation (n = 404), all went through an interview regardless of their questionnaire answers. Results. Only 35-45% of the participants reporting tingling once a week or more in the questionnaire actually had symptoms consistent with CTS when interviewed. The remaining 55-65% had no or infrequent symptoms or symptoms because of other disorders. Few potential CTS cases were missed. The positive predictive values were 0.48 (95% CI = 0.30-0.66) and 0.52 (95% CI = 0.38-0.67) on the right and left hands, respectively. The sensitivities, specificities and negative predictive values ranged from 0.87 to 1.00. Conclusions. Questionnaire information overestimates the prevalence of CTS symptoms. However, asking about tingling in a questionnaire is a simple and sensitive first step to detect potential CTS cases, but symptoms should be confirmed by interview.