Effect of Usual Source of Care on Depression among Medicare Beneficiaries

Background: A usual source of care (USOC), as a structural indicator of access to care, has been widely found to be associated with higher healthcare utilization, lower costs, and better health outcomes. Most previous studies have failed to control for the potential endogeneity between USOC and health care utilization or outcomes. Few studies have examined the effect of access to care on depression among older adults.

Objective: To investigate the effect of having a usual source of care (USOC) on depression prevalence in a representative sample of the U.S. community-dwelling elderly.

Design: A multiple-year cross-sectional study.

Participants: 20,455 community-dwelling person-years were identified for respondents age 65+ in Medicare Fee-for-Service for a full year.

Methods: The Medicare Current Beneficiaries Survey Cost and Use files for 2001-2003 and Area Resource File for 2002 were used. USOC (yes/no) was defined by the question “Is there a particular medical person or a clinic you usually go to when you are sick or for advice about your health?” Screened depression (yes/no) was identified by a two-item screen (sadness and/or anhedonia). Ambulatory care use (ACU) (yes/no) was defined by having at least one physician office visit and/or hospital outpatient visit based on Medicare claims. All measures were for the past 12 months. First, a probit model was used to estimate depression prevalence controlling for USOC and ACU. Then, a simultaneous-equations (trivariate probit) model including equations for USOC, ACU, and depression were utilized to control for endogeneity bias. All estimates were adjusted by sampling weights and study design effects.

Results: A simultaneous-equations model was found to predict more accurately than a single-equation model. The error terms were found to be significantly correlated between the USOC and depression equations (p=0.002), and the USOC and ACU equations (p=0.01). Based on the trivariate probit model, USOC is associated with a 24.8% decrease in the probability of having screened depression (11.5% as compared to 15.3% for those without a USOC) (p<0.05) among community-dwelling elderly, compared with 9.1% (12.0% for people with a USOC as compared to 13.2% for those without a USOC) in the probit model (p=0.39).

Conclusion: USOC is associated with lower prevalence of screened depression among the U.S. community-dwelling elderly. This suggests that enhancing people's ability to access healthcare appears to improve their mental health and improve health care delivery system efficiency.