Interrater agreement in the assessment of motor manifestations of Huntington's disease

Abstract

With prospects improving for experimental therapeutics aimed at postponing the onset of illness in preclinical carriers of the Huntington's disease (HD) gene, we assessed agreement among experienced clinicians with respect to the motor manifestations of HD, a relevant outcome measure for preventive trials in this population. Seventy-five clinicians experienced in the evaluation of patients with early HD and six non-clinicians were shown a videotape compiled from the film archives of the United States-Venezuela Collaborative HD Research Project. Observers were asked to rate a 2-3-minute segment of the motor examination for each of 17 at-risk subjects. The rating scale ranged from 0 (normal) to 4 (unequivocal extrapyramidal movement disorder characteristic of HD). As measured by a weighted kappa statistic, there was substantial agreement among the 75 clinicians in the judgment of unequivocal motor abnormalities comparing scale ratings of 4 with ratings that were not 4 (weighted kappa = 0.67; standard error (SE) = 0.09). Agreement among the non-clinicians was only fair (weighted kappa = 0.28; SE = 0.10). Even under the artificial conditions of a videotape study, experienced clinicians show substantial agreement about the signs that constitute the motor manifestations of illness in subjects at risk for HD. We expect these findings to translate to a similar level of interobserver agreement in the clinical trial setting involving experienced investigators examining live patients.