Health center clinician staffing patterns and services provided

Purposes: 1) To identify factors explaining variation in clinician staffing patterns across community health centers and 2) to assess the impact of such variation on services provided in CHCs. Data Sources and Design: We used the 2006 Uniform Data Set (UDS) to characterize the staffing patterns of CHCs, specifically the numeric importance of nurse practitioners, nurse midwives and physician assistants relative to the number of physicians. These centers' were geocoded allowing us to combine the UDS data with characteristics of primary care service areas (PCSAs)—levels of poverty and rurality—available from the Dartmouth Project. National Provider Identification (NPI) data are used to estimate clinical staffing mix at the state- and PCSA-level. Finally, 2006 UDS data provide information on approximately 70 different services. Multivariate models are estimated to identify predictors of CHC staffing ratios. Logistic regression models, with controls for staffing ratio and financial resources—measured by percent of patients uninsured—are used to understand how the provision of different services are affected by structural factors. Key Findings: 1) There is substantial variation across states in the relative number of midlevel providers in CHCs: ranging from 70% of all clinical providers in Alaska to 25% in Ohio. This variation at a state-level is closely associated with state workforce specialty mix (r=.77). In general, the ratio of midlevel providers to physicians is much higher in CHCs than the state as a whole. 2) Midlevel providers are relatively more common in CHCs with more uninsured (and fewer Medicaid) patients, CHCs located rural areas and those with more patient encounters per physician. 3) The provision of particular services is influenced by staffing ratios and/or financial resources. For instance, we find that CHCs with relatively more physicians are more likely to provide obstetrical and gynecological care. By contrast, CHCs with relatively more midlevel providers are more likely to offer respite care as well as mental health/substance abuse services. Overall, CHCs with relatively fewer physicians are less likely to offer a full array of services. Implications of findings and policy relevance: Variation across CHCs reflect exigencies of their state and local environment and funding sources. To provide a full “medical home” it is necessary to provide CHCs adequate funding. Given complex scope of practice laws at the state level, it is equally important to determine the extent to which difficulties in recruitment and retention of primary care physicians limits the range of services in CHCs.