Session number: 1061
Abstract title: The Demand for primary care physicians for VA Enrollees and its impact on VA inpatient care
Author(s):
Y Shen - Center for Quality, outcomes and economic research
FH Wang - Center for Quality, outcomes and economic research
J Gardner - Center for Quality, outcomes and economic research
A Jayadev - Center for Quality, outcomes and economic research
Objectives: Many VA enrollees consider a non-VA physician as their primary-care doctor. We investigated factors associated with this choice of primary-care physician and how it affected use of VA inpatient care.
Methods: ‘1999 enrollee survey’ (self-reports of a VA/non-VA primary care doctor, trust in doctors, travel time and individual characteristics) were merged with AHA and Area Resource File data (hospital variables). We used PTF 1999 data to identify the number of inpatient stays. Using DRGs we grouped inpatients as: receiving only mental health/substance abuse (MH/SA) care, no MH/SA, MH/SA and other care. Sample size is 102,507. Probit regression estimated the correlation of having a VA primary care doctor to the trust score and travel time, controlling for individual characteristics. Tobit model estimated the impact of having a VA primary care physician on inpatient stays. Multinomial probit examined its impact on MH/SA care, controlling for patient and VA/non-VA hospital characteristics. Maximum-likelihood was applied to examine endogeneity of VA physician choice that might bias estimates.
Results: 65% of enrollees reported having VA primary care physicians, of whom 16% had a VA inpatient stay. Only 1.8% of those with non-VA physicians had any VA inpatient stay. People traveled farther to see VA primary care physicians. Physician-Trust scores were similar for enrollees with VA/non-VA physicians. Having a VA primary care physician increased the number of VA stays by 3 (p<=0.001). Enrollees with non-VA primary care physicians are more likely to come to VA for MH/SA care. The endogeneity of choosing VA physicians was significant.
Conclusions: Patients with VA primary care physicians are more likely to use VA for two reasons: they chose VA physicians because their perceived needs of VA care (endogeniety); and VA physicians refer patients to VA facilities for all/most of their care while non-VA physicians refer patients to non-VA providers for all but MH/SA care which has limited coverage under non-VA systems.
Impact statement: Primary care physicians determine the referrals and utilization for patients, any impact on demand for primary care can affect veterans’ patterns of using other services.