Journal of Healthcare Management - September/October 2016 - 375

M edI cal S taff c onf I guratIonS models be compared with MA-heavy models? These findings also suggest that some community colleges that traditionally have trained nurses for acute care settings may have a significant opportunity to train nurses for primary care. CEOs' interest in retention is also consistent with the commitment of local workforce boards (charged with creating new jobs) to nursing as an economic opportunity in rural communities (Lauder, Reel, Farmer, & Griggs, 2006), especially through "stackable" degree programs in community colleges. Local wage gaps among RNs, LPNs, and MAs were also a major concern for CEOs in our study, and workforce planners and researchers should consider them to be a local variable. Understanding what drives variation in wage differences is also important; is variation simply a matter of constrained supply and competition from hospitals? If so, should schools in regions with lower wages for nurses and an adequate supply be particularly attuned to training LPNs and RNs for primary care settings? Across the provider and clinical support domains, we found that, with two exceptions, payment policies were not a dominant factor in CEOs' thinking about staffing choices. The first relevant exception pertained to private insurance carriers that reimburse for nursing services. In areas with a high supply of nurses, encouraging payers to reimburse for nursing services could lead to more nurses practicing in CHCs. The second payment-related variable that emerged from our study was PCMH accreditation. We found three staffing In c o MMunIty H ealt H c enter S changes attributed to PCMH accreditation: (1) the opportunity to hire more APPs to complement the physician staff, (2) hiring of MAs to manage data, and (3) increased use of nurses in quality improvement roles. As more primary care practices strive to attain PCMH accreditation, tracking the effects of PCMH on staffing configurations will be important, both to anticipate changes in demand and to begin to understand the associations between staffing changes related to PCMH status and care outcomes. S T U D Y L I M I TAT I O N S Our study was limited by several factors. First, the qualitative design and use of maximum variety sampling do not allow us to generate hypotheses regarding correlations between participant- or site-level variables and CEO viewpoints. On the other hand, our study provides a foundation for further research to explore these questions. Second, the study's focus on CHCs, which provide approximately 10% of primary care visits in the United States (Hing & Uddin, 2010), may confine its relevance to those settings. Third, because the UDS data are aggregated across sites for each grantee, some of the staff configuration statistics we used to identify study sites reflected single sites, whereas others reflected multiple sites. However, we assumed that the CEOs' decision criteria about medical staffing configurations would likely be similar for single- or multiple-site CHCs. Fourth, our focus on outlier CHCs that have high proportions of the types of providers and clinical support staff being studied may not represent the experiences of all 375

Table of Contents for the Digital Edition of Journal of Healthcare Management - September/October 2016