Cost-effectiveness of a nurse practitioner-family physician model of care in a nursing home: controlled before and after study.

Aims To examine the cost-effectiveness of a nurse practitioner-family physician model of care compared with family physician-only care in a Canadian nursing home. Background As demand for long-term care increases, alternative care models including nurse practitioners are being explored. Design Cost-effectiveness analysis using a controlled before-after design. Methods The study included an 18-month ‘before’ period (2005–2006) and a 21-month ‘after’ time period (2007–2009). Data were abstracted from charts from 2008–2010. We calculated incremental cost-effectiveness ratios comparing the intervention (nurse practitioner-family physician model; n = 45) to internal (n = 65), external (n = 70) and combined internal/external family physician-only control groups, measured as the change in healthcare costs divided by the change in emergency department transfers/person-month. We assessed joint uncertainty around costs and effects using non-parametric bootstrapping and cost-effectiveness acceptability curves. Results Point estimates of the incremental cost-effectiveness ratio demonstrated the nurse practitioner-family physician model dominated the internal and combined control groups (i.e. was associated with smaller increases in costs and emergency department transfers/person-month). Compared with the external control, the intervention resulted in a smaller increase in costs and larger increase in emergency department transfers. Using a willingness-to-pay threshold of $1000 CAD/emergency department transfer, the probability the intervention was cost-effective compared with the internal, external and combined control groups was 26%, 21% and 25%. Conclusion Due to uncertainty around the distribution of costs and effects, we were unable to make a definitive conclusion regarding the cost-effectiveness of the nurse practitioner-family physician model; however, these results suggest benefits that could be confirmed in a larger study.