A method for checking interobserver reliability in observational sleep studies

Disturbances of the sleep/wake cycle occur frequently in nursing home residents. Because of the many technical difficulties in studying sleep and rhythms in such patients, systematic behavioral observations offer an alternative approach. In this study we describe a method for determining interrater reliability of such observations. Two individuals observed 39 nursing home residents four times per hour during daytime and nighttime hours. Results indicated high interrater reliability for both sleep/wake and the presence of apnea during sleep. Both day and night observations were made reliably. We found only 101 discrepancies of 1.160 tandem observations. These results suggests that behavioral observations are a viable approach in the study of the sleep/wake cycle in nursing home settings.