A multi-objective model for optimizing staffing across geographically distributed patient-centered medical homes

Abstract

As demand for medical services increases, it may become necessary to centralize provision of medical care, particularly specialty services, to certain locations within a geographic region. Due to tradeoffs between operating costs, the risk of limited access, and decreased availability of patient-centered care, there is a need to help healthcare administrators make informed decisions. This article examines the issue of staffing clinician care in a region of locally distributed patient-centered clinics. As previous literature suggests, there are some benefits, as well as drawbacks, to centralizing the provision of care. Therefore, an administrator must understand the tradeoffs between the risk of not meeting patient demand while considering staffing expenses, patient travel time, and lack of continuity with large centralized clinics. We propose a multi-objective mixed-integer program to minimize the risk of insufficient staffing as well as minimize an aggregated penalty function that incorporates staffing expenses, patient travel time, and a discontinuity penalty term. The methodology provides an efficient frontier of risk versus penalty and we demonstrate the approach with numerical results using data sampled from a typical demand distribution. Furthermore, the numerical examples demonstrate how optimal decisions could vary, depending on the distribution that characterizes demand, particularly when the demand distribution has non-normal or heavy-tailed effects.

Keywords:

• Healthcare management
• integer programming
• outpatient clinics
• risk management
• stochastic programming

Additional information

Funding

This research was supported in part by the National Science Foundation, Grant CMMI-1632793.