Studying the retention of rural physicians.

Rural communities and policy-makers struggle with efforts to enhance the retention of rural physicians. Research available to guide these efforts is often weak methodologically and thus may be pointing retention efforts in nonproductive directions. This article discusses a range of methodologic issues encountered in rural physician retention studies for the purpose of strengthening future studies. Ideal study approaches to answer causal questions, including questions about the "causes" of rural physician retention, must demonstrate good internal validity, for which chance, bias, and confounding are accounted. Retention studies that rely simply on asking physicians why they stay or leave rural areas can be useful at times, but are too prone to bias and their findings difficult to verify. Simply identifying what physicians find satisfying or dissatisfying about rural work also will not reliably reveal why they stay or leave, a related but still distinct question. Stronger approaches to studying retention include the traditional quantitative study--in which retention factors are identified when they are statistically related to physicians' retention, and the increasingly popular qualitative study--in which retention issues are revealed through prolonged, in-depth interactions with physicians. This article also discusses various definitions of retention, the use of survival curves to present retention findings, and the importance of studying retention in inception cohorts. The benefits and downside of studying retention with prospective and retrospective study designs are described.