350 Enhanced maternal mortality surveillance identifies higher mortality ratios and greater racial/ethnic disparity than death certificates

Abstract

To describe California’s Pregnancy Mortality Surveillance System (CA-PMSS), a rapid case review method for tracking pregnancy-related deaths statewide. To compare death counts and maternal mortality measures from CA-PMSS with those based on two conventional case ascertainment methods: death certificate data only and linked vital records. CA-PMSS identified deaths that occurred while pregnant (including early pregnancy) or within 1 year of pregnancy via linkage of vital records and hospitalization data. Coroner reports and medical records were obtained to confirm pregnancy status, verify timing to death, and supplement case information. An expert committee determined the cause of death and relationship to pregnancy based on review of these records. For deaths identified by death certificate data, pregnancy-relatedness was determined using obstetric codes. CA-PMSS death counts and mortality ratios were compared to those generated by (1) death certificate data alone and (2) linked vital records data (i.e., death, birth, and fetal death files) to assess misclassification of obstetric deaths. Data were disaggregated by race/ethnicity to examine differences in disparities by case ascertainment method. In 2009-2016, 254 maternal deaths were identified by death certificate data; 341 deaths by linked vital records data; and CA-PMSS identified 446 pregnancy-related deaths within 42 days postpartum. As a result, mortality ratios generated by CA-PMSS were 76% higher than those based on death certificate data alone and 31% higher than those using linked vital records. The 2014-2016 mortality ratio generated by CA-PMSS revealed greater racial/ethnic disparities than the one based on death certificate data. CA-PMSS, a record review method for maternal mortality surveillance, provides a more complete and accurate accounting of pregnancy-related deaths than death certificate data alone or linked vital records. Accurate pregnancy-related mortality data are critical for public health surveillance.