RELIABILITY BETWEEN 2 OBSERVERS USING A PROTOCOL FOR DIAGNOSING ESSENTIAL TREMOR

Protocols with demonstrated reliability have been established for the diagnosis of numerous movement disorders, whereas in the essential tre mor (ET) literature, there is no discussion about the reliability of d iagnostic protocols. Lack of knowledge of the reliability of diagnosti c protocols in ET limits the use of these protocols because reliabilit y is an essential requirement for scientific quality in data managemen t. The ob objective of this study was to determine the reliability of a protocol for diagnosing ET. The protocol consists of a Tremor Interv iew, a videotaped Tremor Examination, and a diagnostic algorithm. Eigh ty-three subjects with ET, identified in a community-based health stud y in Washington Heights-Inwood, New York, were matched with 83 control subjects from the same community. These subjects and their relatives are being recruited to participate in the Washington Heights-Inwood Ge netic Study of ET. Two hundred twenty-six subjects have been evaluated to date (35 ET cases, 40 controls, 151 relatives). All 226 underwent an 84-item Tremor interview and 26-item videotaped Tremor Examination. Diagnoses (normal, possible ET, probable ET, definite ET) were indepe ndently assigned by two blinded neurologists specializing in movement disorders. The kappa statistic, k, was used to determine diagnostic ag reement between these two neurologists. The concordance rate between t wo raters using diagnostic categories definite ET, probable ET, possib le ET, and normal was 80%; k(w) = 0.84 (near perfect to perfect agreem ent). The concordance rate between two raters using two diagnostic cat egories (definite ET and normal) was 100%; k = 1.00 (perfect agreement ), There was high correlation between the two raters' total tremor sco res (r = 0.89, p < 0.00001). This diagnostic protocol is highly reliab le. Research in ET would greatly benefit from diagnostic protocols wit h demonstrated reliability.