Provider-Associated Measurement Error in Routine Outcome Monitoring in Community Mental Health

Despite the strengths of routine outcome monitoring (ROM) in community mental health settings, there are a number of barriers to effective implementation of ROM, including measurement error due to provider factors (e.g., training level) and non-target client factors (i.e., client characteristics which have no meaningful relationship to the outcome of interest). In this study, ROM data from 80 client-provider dyads were examined for sources of variance due to provider factors and non-target client factors. Results indicated that provider factors and non-target client factors accounted for between 9.6 and 54% of the variance in the ROM measures. Our findings supported past research that provider characteristics impact ROM, and added the novel finding that client gender, age, diagnosis, and cognition also impact ROM. Methods to increase accuracy and utility of ROM in community mental health are discussed.