Job Market Paper

Understanding Variation in Cesarean Section Use: Supply-Side Drivers and Maternal Health Effects with Helena Roy

A longstanding concern in U.S. healthcare is the large regional variation in spending and treatment intensity, often unaccompanied by improvements in outcomes. We explore this issue in the context of Cesarean sections (C-sections), the most performed inpatient surgery. Notably, C-section rates differ up to 10-fold across hospitals, with significant variation even for clinically similar patients. We use Medicaid administrative claims data, covering 42% of all births in the US, to quantify the contribution of the primary supply-side agents in healthcare to variation in C-section use. Leveraging variation from physicians who perform deliveries at multiple hospitals, we find that differences in physician practice style across hospitals can explain approximately one quarter of the across-hospital differences in C-section usage. This variation in physician practice style has important implications for patient health. Patients quasi-randomly assigned to more C-section-intensive physicians have a significantly higher probability of unplanned C-sections, and these marginal C-sections are associated with worse postpartum maternal health outcomes without measurable benefits to newborns.  Our findings highlight physician practice style as an important driver of C-section use with direct consequences for maternal health. 

Selected Work In Progress

The Role of Physician Training in Racial Disparities in Maternal Healthcare with Helena Roy

Maternal health in the US is poor relative to peer countries and evidences strong racial disparities. In particular, Black patients in the US often face heightened incidence of pregnancy complications, such as preeclampsia. A potential instrument for reducing disparities is physician training in patient heterogeneity by race. Residency provides the bedrock for specialist practice in obstetrics and gynecology, and could be a valuable training period in which to highlight race-specific risks and considerations. In this paper, we examine existent variation in residency training and ask if racial disparities in treatment use and outcomes are evident within residency, based on the practice of their alumni. We find small effects of residency on treatment use and outcomes for patients of all races. Notably, a residency's relative effect on C-section use and delivery complications is not strongly correlated across Black versus White patients, suggesting non-uniform learning across different patient groups. We also examine if training at a residency with a more diverse patient mix, a feature advertised by multiple residencies as an important part of physician education, impacts physicians' deviation from clinical consensus in delivery method diagnosis. Though we document such deviations are more likely for Black patients, we find essentially no impact of training with a diverse patient population on this racial gap.

Birth Centers and Maternal and Infant Health (with Ambar La Forgia, Petra Persson, Maya Rossin-Slater and Helena Roy)

Willingness to Pay for Concordance in Healthcare (with Tamri Matiashvili and Helena Roy)

Shell-Shocked: Long Run Consequences of Wartime Stress  (with Boaz Abramson)