Characteristics of Maternal Mortality Missed by Vital Statistics in Hong Kong, 2000-2019.

Importance: Reducing maternal mortality is a global objective. The maternal mortality ratio (MMR) is low in Hong Kong, China, but there has been no local confidential enquiry into maternal death, and underreporting is likely. Objective: To determine the causes and timing of maternal death in Hong Kong and identify deaths and their causes that were missed by the Hong Kong vital statistics database. Design, Setting, and Participants: This cross-sectional study was conducted among all 8 public maternity hospitals in Hong Kong. Maternal deaths were identified using prespecified search criteria, including a registered delivery episode between 2000 to 2019 and a registered death episode within 365 days after delivery. Cases as reported by the vital statistics were then compared with the deaths found in the hospital-based cohort. Data were analyzed from June to July 2022. Main Outcomes and Measures: The outcomes of interest were maternal mortality, defined as death during pregnancy or within 42 days after ending the pregnancy, and late maternal death, defined as death more than 42 days but less than 1 year after end of the pregnancy. Results: A total of 173 maternal deaths (median [IQR] age at childbirth, 33 [29-36] years) were found, including 74 maternal mortality events (45 direct deaths and 29 indirect deaths) and 99 late maternal deaths. Of 173 maternal deaths, 66 women (38.2%) of individuals had preexisting medical conditions. For maternal mortality, the MMR ranged from 1.63 to 16.78 deaths per 100 000 live births. Suicide was the leading cause of direct death (15 of 45 deaths [33.3%]). Stroke and cancer deaths were the most common causes of indirect death (8 of 29 deaths [27.6%] each). A total of 63 individuals (85.1%) died during the postpartum period. In the theme-based approach analysis, the leading causes of death were suicide (15 of 74 deaths [20.3%]) and hypertensive disorders (10 of 74 deaths [13.5%]). The vital statistics in Hong Kong missed 67 maternal mortality events (90.5%). All suicides and amniotic fluid embolisms, 90.0% of hypertensive disorders, 50.0% of obstetric hemorrhages, and 96.6% of indirect deaths were missed by the vital statistics. The late maternal death ratio ranged from 0 to 16.36 deaths per 100 000 live births. The leading causes of late maternal death were cancer (40 of 99 deaths [40.4%]) and suicide (22 of 99 deaths [22.2%]). Conclusions and Relevance: In this cross-sectional study of maternal mortality in Hong Kong, suicide and hypertensive disorder were the dominant causes of death. The current vital statistics methods were unable to capture most of the maternal mortality events found in this hospital-based cohort. Adding a pregnancy checkbox to death certificates and setting up a confidential enquiry into maternal death could be possible solutions to reveal the hidden deaths.