Value and disvalue of the pregnancy checkbox on death certificates in the United States -- Impact on newly released 2018 maternal mortality data.

Abstract Maternal mortality is a sentinel health indicator. To improve identification of maternal deaths, a pregnancy question was added during the 2003 revision of the standard death certificate in the United States. Its adoption across U.S. states took 16 years (2003-2018), and therefore, the National Center for Health Statistics (NCHS) has not provided a national maternal mortality rate (MMR) between 2007 and 2018. During this time, researchers raised questions about the accuracy of the checkbox information, particularly its contribution to over-reporting of maternal deaths in the United States. Checkbox errors were especially evident among women >40 years and for nonspecific causes of death. In January 2020, NCHS resumed reporting of maternal mortality and provided 2018 figures using a new coding method (i.e. 2018 method). Despite internal analyses suggesting both high false positive and high false negative pregnancy checkbox errors, NCHS reported identification of 658 maternal deaths nationwide and an MMR of 17.4 deaths per 100,000 live births for 2018. The 2018 coding method restricts application of the checkbox information to decedents 10–44 years – the information is, therefore, not applied to those >45 years when no pregnancy-related cause of death information is indicated on the death certificate; deaths with a pregnancy or obstetric condition reported in the cause of death section of the death certificate continue to be coded as maternal deaths regardless of age. The 2018 method likely corrects errors introduced by the use of the checkbox among women >45 years, but whether it provides accurate maternal mortality figures remains unknown. We call for efforts to urgently and systematically validate the pregnancy checkbox information. Post-hoc coding adjustments cannot substitute for providing accurate and actionable maternal mortality data.