Portal de Periódicos da CAPES

The Quasi-Experimental Effects of Maternal Health, Severe Maternal Morbidity, and Extreme Climate Disasters: A Difference-In Difference Analysis of the South Carolina 2015 Extreme Floods

2022; RELX Group (Netherlands); Linguagem: Inglês

10.2139/ssrn.4136258

1556-5068

Margaret M. Sugg, Jennifer D. Runkle, Sophie Ryan, Luke Wertis,

Food Security and Health in Diverse Populations

Research documenting the public health impacts of natural disasters often focuses on adults and children. Little research has examined the influence of extreme events, like floods, on maternal health, and less has examined the effect of disasters on maternal indicators like severe maternal morbidity (SMM), unexpected outcomes of labor and delivery that result in significant short- or long-term consequences to a woman's health. SMM provides a surveillance composite indicator of near-miss mortality events used by the U.S. Centers for Disease Control to assess maternal health trends. The aim of this study is to identify the impacts of the 2015 flood events on maternal health outcomes and SMM in South Carolina, USA. We employ a quasi-experimental design using a difference-in-difference analysis with log-binomial regressions to evaluate maternal outcomes for impacted and control locations during the disaster event. We did not find evidence of statistically significant main effects on maternal health from the 2015 flood events related to preterm birth, gestational diabetes, mental disorders of pregnancy, depression, and generalized anxiety. However, we did find a statistically significant effect for increases in SMM and low birth weight during the flood event for women in select trimester periods who were directly exposed. Our work provides new evidence on the effects of extreme flood events, like the 2015 floods, which can impact maternal health during specific exposure periods of pregnancy. Additional research is needed across other disasters and extreme weather events as the unique context of the 2015 floods limit the generalizability of our findings.