Effects of Obstetric Unit Closures on Maternity Care Access and Delivery Outcomes

• In Event: Access to Health Care

Abstract

Wide disparities exist in maternity care access and in maternal and infant health outcomes, and closures of obstetric units may exacerbate these disparities. Existing estimates of maternity care access and effects of obstetric unit closures frequently use county-level data, which mask heterogeneity within counties, including among different racial and ethnic populations. In this paper, we study obstetric unit closures in the State of Illinois to provide novel state-wide evidence on the causal effects of obstetric unit closures in diverse and primarily metropolitan populations. 

We first show that local measures of birthing hospital access (i.e., hospital obstetric unit access) at the ZIP Code level more precisely capture changes in access -- such as among racial and ethnic populations in the City of Chicago -- than county-level birthing hospital access measures. We then investigate the causal effects of Illinois obstetric unit closures using Illinois Department of Public Health hospital discharge data from years 2015-2021 and recent event study methods that allow for heterogeneous treatment effects across obstetric unit closures. 

We find that non-Hispanic Black and Hispanic individuals are more likely to experience nearby obstetric unit closures than non-Hispanic White individuals. Nearby obstetric unit closures increase travel times to chosen (i.e., actual) delivery hospitals on average for all three of these racial and ethnic groups. Focusing on outcomes particularly relevant to non-metropolitan populations, we find that nearby obstetric unit closures increase the share of non-Hispanic White individuals with travel times greater than 40 minutes to chosen delivery hospitals. 

We examine the effects of obstetric unit closures on the characteristics of chosen delivery hospitals. Obstetric unit closures affect the perinatal level of chosen delivery hospitals because the obstetric units that close offer lower levels of perinatal care (i.e., have fewer capabilities) on average. After nearby obstetric unit closures, non-Hispanic Black individuals are more likely to deliver at the highest level (Level III) hospitals, and both non-Hispanic Black and non-Hispanic White individuals are more likely to deliver at the second highest or highest level (Level II-E or III) hospitals. We find suggestive evidence that nearby obstetric unit closures also increase non-birthing hospital deliveries among non-Hispanic Black and Hispanic individuals. 

Finally, we do not find significant effects of nearby obstetric unit closures on maternal and infant health outcomes at delivery, including preterm or early term births, low-risk Cesarean deliveries, and severe maternal morbidity (SMM), on average. However, we examine only several years of obstetric closures in Illinois, and our statistical power is limited for the analysis of population subsamples and relatively rare outcomes like SMM. 

Our results highlight the value of local-level measures in addressing disparities in access to maternity care and may help inform state governments in their design of policies related to hospital service reductions and closures. While obstetric unit closures reduce access to birthing hospitals, our results also underscore the need for other types of policies to address disparities in maternal and infant health outcomes, since reducing or preventing obstetric unit closures alone might not substantially improve delivery health outcomes. 

Author

• Jennifer Helen Kwok, University of Illinois, Chicago
  Presenting Author