Effects of depression screening on diagnosing and treating mood disorders among older adults in office-based primary care outpatient settings: An instrumental variable analysis

Existing literature shows mixed findings regarding the efficacy and effectiveness of depression screening, and relatively little is known about the effectiveness of depression screening among older adults in primary care visits in the U.S. This study examines the effects of depression screening on the three following outcomes: mood disorder diagnoses, overall antidepressant prescriptions, and potentially inappropriate antidepressant prescriptions among older adults ages 65 or older in office-based outpatient primary care settings. We used data from 2010-2012 National Ambulatory Medical Care Survey (NAMCS), a nationally representative sample of office-based primary care outpatient visits among older adults (n=9,313 unweighted). We employed an instrumental variable approach to control for selection bias in our repeated cross-sectional population-based study. Injury prevention and stress management were selected as instrumental variables, as they were considered completely exogenous to outcomes of interests using conceptual and statistical criteria. We conducted multivariate bivariate probit (biprobit) regression analyses to investigate the effect of depression screening on each outcome, when controlled for other covariates. We found that depression screening was negatively associated with potentially inappropriate antidepressant prescriptions (β=-2.17; 95% CI -2.80 to -1.53; p<0.001). However, no significant effect of depression screening on diagnosis of mood disorders and overall antidepressant prescriptions was found. Overall, depression screening had a negative effect on potentially inappropriate antidepressant prescriptions. Primary care physicians and other healthcare providers should actively utilize depression screening to minimize potentially inappropriate antidepressant prescriptions in older adult patients.