UCLA

EMTALA requires Medicare-participating hospitals to provide emergency care to all patients regardless of payer. We examine the effect of EMTALA in three studies.

(1) In key informant interviews, we examine causes for and solutions to EMTALA noncompliance. We find that hospitals may violate EMTALA for five reasons: financial pressure, complexity/lack of knowledge about the law, a high referral burden that makes it difficult to comply with EMTALA, inter-hospital relationships that discourage reporting on borderline inappropriate transfers, and a principal-agent problem with differing physician and hospital priorities. We propose several ways to strengthen the Act, including requiring Medicaid to fully reimburse required EMTALA screening exams and amending EMTALA to permit informal mediation sessions between hospitals.

(2) We measure changes in hospital utilization and readmissions when EMTALA is extended to inpatients. In 2009, the Federal Court of Appeals for the Sixth Circuit ruled that EMTALA obligations continued until a patient was stabilized, regardless of whether s/he was admitted. However, hospitals outside the court’s jurisdiction continued to follow 2003 regulations that EMTALA obligations cease after a good faith admission. This study uses a difference-in-difference-in-differences design, comparing Medicaid/uninsured with commercially-insured patients before and after the case in hospitals inside and outside the Sixth Circuit. We find that although more unprofitable inpatients are discharged with a short length of stay after extending EMTALA to inpatients, they are substantially less likely to be readmitted. These results suggest that extending EMTALA to inpatients may encourage hospitals to fully stabilize unprofitable patients admitted from the emergency department (ED).

(3) We explore whether hospitals strategically avoid treating uninsured and Medicaid patients by temporarily closing part of their EDs (through an ambulance diversion) when nearby safety net hospitals declare diversion. We find that hospitals are more likely to declare diversions when nearby safety net hospitals go on diversion, as compared to when nearby non-safety safety net hospitals (matched by size and distance) do so. Furthermore, hospitals that divert when a nearby safety net hospital diverts have a slightly lower ED occupancy than hospitals that divert when a nearby non-safety net hospital diverts. In addition, we theorized that, like musical chairs, hospitals do not want to be last one with an open ED after a nearby safety net hospital declares a diversion. Consistent with this theory, when multiple hospitals in a market are on diversion, the third hospital in a market to declare a diversion does so sooner if the first hospital declaring a diversion is a safety net hospital than if the first hospital is a non-safety net hospital.

Hospitals also end their diversions differently depending on whether the nearby diverting hospital was a safety net hospital or non-safety net hospital. Specifically, hospitals are on diversion longer and end their diversions later after a nearby safety net hospital ended its diversion than after a nearby non-safety net hospital ends its diversion. Perhaps hospitals that are strategically diverting are waiting to make sure that the nearby safety net hospital will not go back on diversion.

Our results suggest that hospitals engage in strategic diversions, reducing access to emergency services to unprofitable patients and circumventing the goals of EMTALA.