Patient Flow and Capacity Management in Health Services

<p>Growing demand for health services provided by outpatient clinics and hospitals caused health institutions flow and capacity challenges. Health organizations’ poor response to these challenges directly translate into negative patient outcomes and intensified downstream costs. In this study, we investigate dynamics and mechanisms that influence patient wait times and capacity strains and propose strategies and policies that can improve these issues in both ambulatory and inpatient care. </p> <p>First, we investigate the access issue in a multidisciplinary memory clinic, which consists of three practices and six patient types. Considering patient flow and interactions, we develop an empirical simulation model to evaluate the effectiveness of access improvement strategies such as overbooking, repatriation (i.e., referring the patient back to primary care), and increasing provider hours. Our results suggest that despite the increasing wait times in the multidisciplinary memory clinic, increasing provider slots is not always an effective strategy. In fact, overbooking and reducing unnecessary follow-up visits can result in more significant performance improvements.</p> <p>Second, we study the impact of long-stay patients (i.e., patients with discharge barriers that stay in the hospital for non-medical reasons) on flow and capacity. In particular, we focus on the patient flow between Intensive Care Unit (ICU), Step-down Unit (SDU), and Medical Unit (MU) and quantify the impact of long-stay patient volumes on wait time, length of stay (LOS), and 30-day readmission probability of other patients transitioning among these units. We find that larger proportion of long-stay patients in the MU results in shorter LOS for other patients in the MU, and longer wait time for patients leaving the ICU to MU.</p> <p>Third, we examine existing patient grouping system based on the service lines at two hospitals within the same health system and propose a two-step clustering-classification approach to identify new patient clusters. Unlike existing 8 patient clusters (i.e., service lines), our results identified 11 patient clusters in Wilmington hospital and 15 patient clusters in Christiana hospital, indicating the need to further splitting some of the existing service lines such as internal medicine, general surgery, and neurological disorders. </p>