Abstract
This report discusses issues of validity and reliability that arose when activity of daily living (ADL) items from Residential Assessment Instrument (RAI) were used by case managers in the State of Washington to determine care needs of elders in a community-based long-term care system funded by Medicaid. It describes methods for and results from an interrater reliability study of these ADL items and discusses strengths and limitations of the study. Also, it recommends ways (1) to improve the study methods, (2) to integrate reliability studies into routine operations, and (3) for case managers to improve the accuracy of assessments.