Does a screening questionnaire for essential tremor agree with the physician's examination?

In research studies of essential tremor (ET), monetary and geographic factors often necessitate diagnosis by interview rather than by examination. Few attempts have been made to determine the validity of a screening instrument for ET. A total of 242 subjects (33 definite or probable ET, 54 possible ET, and 155 normal) were part of a community-based family study of ET in northern Manhattan. Subjects underwent a tremor interview and videotaped tremor examination. The interview included 12 screening questions for ET. Two neurologists rated the severity of tremor on videotape and assigned diagnoses. The subjects' responses to the 12 screening questions were compared with the neurologic examination and the neurologists' diagnoses. There was an association between the number of affirmative responses on the 12-item screen and the severity of tremor during the videotaped tremor examination (r = 0.66, p < 0.01). Including those with probable and definite ET and defining a positive screen result as an affirmative response to at least i of 12 screening questions, the sensitivity of the screening questionnaire was 73%, the specificity was 96%, and the positive predictive value was 80%. The responses to the questionnaire closely correlated with the severity of tremor on neurologic examination. The questionnaire had moderate validity for subjects with probable or definite ET; subjects with mild ET often screened negative. Hence, in population- based studies and family studies of ET, it is important to examine all subjects for the presence of mild asymptomatic ET.