Hospital Boarding Crises: The Impact of Urgent Versus Prevention Responses on Length of Stay

Abstract

Healthcare policy makers use wait-time metrics to encourage hospital managers to improve patient experience. For example, in 2002, Massachusetts mandated that hospital managers develop processes to identify and respond to boarding crises, which occur when emergency department patients experience long waits for inpatient beds. Performance improvement theory suggests that patients would be better served by preventing boarding crises rather than responding urgently after they occur.
To empirically test this theory, we use data from a Massachusetts hospital with two different physician-based processes related to boarding and patient flow. First, to comply with the state mandate, the hospital developed processes to identify when the hospital is in a boarding crisis, and subsequently requests that physicians prioritize patient discharge (urgent response). Second, physicians can use pre-discharge orders (PDO)—optional written communication about discharge barriers—to avoid unnecessary discharge delays for patients approaching discharge. Ensuring that beds are freed up as soon as possible might prevent boarding crises from occurring (prevention response).
Our data supports the existence of a tradeoff between these two responses. Surprisingly, and counter to our hypothesis, the state-mandated urgent response is associated with an increased inpatient length of stay (LOS) of 49.7%. We suspect that the LOS-increasing effect of high levels of hospital occupancy that occur for all patients during a CY outweigh any reduction in LOS gained by the few patients who might be discharged early due to the CY. We also find that CY does not improve ED outcomes. The prevention response is associated with a 24.7% reduction in inpatient LOS. We conduct counterfactual analyses at the day-level to find that if physicians use PDOs for all of their patients, it would increase the number of discharges per day and substantially reduce the number of boarding crises. We conclude that the state policy has unintended negative consequences that stymie hospital efforts to create longer-term improvement in their hospitals.

Keywords: Health Care; Hospitals; Information Systems; Application Contexts/Sectors; Government; Regulations; Empirical Operations; Discharge Coordination