Understanding treatment dropout in outpatient mental health services: An economic framework exploring the relationship between patient satisfaction and appointment noncompliance

Noncompliance in the form of treatment dropouts is a major problem across outpatient mental health settings and can range from 40--50% of all clients. Economic theory suggests noncompliance is consumer signaling about patient preferences. The mental health industry standard for evaluating patient preferences is the patient satisfaction measure. This study examined the relationship between appointment noncompliance and patient satisfaction hypothesizing that, after controlling for sociodemographic and health status factors, patient satisfaction will predict duration of mental health treatment. Using 1996--1999 data from the Agency for Healthcare Research and Quality Medical Expenditure Panel, a satisfaction measure was constructed and regression models developed to examine the impact of satisfaction on treatment duration for patients having at least one mental health visit. Survival analysis was used to examine the hazard rate of patients having high versus low satisfaction. Major findings suggests individuals satisfied with and having confidence in the provider were significantly more likely to have a subsequent visit. The importance of satisfaction in predicting treatment continuation was not constant with respect to number of visits or patient sociodemographic characteristics. Satisfaction was strongly predictive of continuation for college educated and individuals with a mental health diagnosis, but less significantly important for African Americans, high school educated and individuals age 25--44. Practical aspects of care such as access were significantly more important for Hispanics and publicly insured. Evidence suggests there is a distinct group of consumers who seek and are satisfied with consultation only. MEPS design measures satisfaction with 'usual source of care' lacks specificity and does not reflect satisfaction specifically with mental health care. The role of treatment preferences needs to investigated further. Patient satisfaction measures require redesign to account for opportunity costs and preferences. Preference measurement techniques may be are better suited. Treatment noncompliance should be the gold standard on which the validity of satisfaction and preference measures are based.