Market Allocation and Health Disparities: The Black-White Survival Gap After Acute Myocardial Infarction

Wednesday, June 26, 2019: 12:00 PM
Madison B (Marriott Wardman Park Hotel)

Presenter: Adam Sacarny

Co-Authors: Amitabh Chandra; Pragya Kakani

Discussant: Marcella Alsan

While black-white health disparities have attenuated meaningfully over the past two decades, black patients continue to face worse health outcomes than their white peers across a wide array of conditions. We identify two channels through which disparities could evolve over time. One “traditional” channel is quality improvement, or a rise in clinical quality at a given hospital with its patient populations held fixed; such improvement can widen or narrow disparities depending on the populations that improving hospitals serve. A second, separate channel that has received less attention in the literature is market allocation, or a change in the assignment of patients to hospitals with clinical quality held fixed; reallocation has ambiguous effects on disparities that depend on the evolution of the sensitivity of patients or their agents to quality.

To disentangle the contributions of quality improvement and market allocation to disparities, we adapt dynamic productivity growth decompositions, typically used to study the evolution of productivity across firms in traditional sectors like manufacturing, to the hospital market. As an application of this technique, we decompose the evolution of black-white survival disparities following acute myocardial infarction (AMI), or heart attack, from 1995-2014 in Medicare. In preliminary results, consistent with the focus of much of the literature, we find that most of the survival gains for black and white heart attack patients are attributable to quality improvement – that is, survival gains within hospitals. Still, our preliminary findings isolate an important role for reallocation of more than 25% of the survival gains for both groups. Our initial results also suggest that changes in allocation raised average hospital quality for black patients relative to white patients over the study period, as measured by the risk-adjusted AMI survival rate of hospitals that patients visit. Specifically, reallocation closed the risk-adjusted survival disparity by 0.7 percentage points over this period, a magnitude comparable to lauded AMI survival-improving technology like reperfusion and angioplasty.

Previous research has suggested that the geographic distribution of patients is a key determinant of hospital market allocation. To better understand the role of this factor in black-white disparities, we next combine the productivity decomposition with reweighting methods to equalize the distribution of patients across areas. According to our preliminary findings, one-half to two-thirds of the allocation gap between black and white patients is due to differences in their geographic distributions. Differences in geography across markets (hospital referral regions) and within markets (ZIP codes conditional on hospital referral regions) both play a role. Given the ability of information, steering, and networks to change allocation within markets, these results suggest that reallocation has potential, if not entirely realized, to reduce socioeconomic health disparities.