The number of maternal deaths in the United States is growing. The maternal mortality rate for Indigenous and Black people is two to three times higher than that of white mothers. Previous studies demonstrate that social determinants of health and lack of access to high-quality care are some factors that contribute to these disparities. With the sharp uptick in maternal mortality precipitated by the COVID-19 pandemic, there is a continuing need to understand the complex mechanisms that lead to poor maternal health outcomes. Improving collection of and access to high-quality data is imperative in helping us understand and develop effective solutions to address them.
Strengthening Data Collection and Insights
Maternal health data is collected, stored, and variably shared at the federal, state, local, and even health care system levels. The very complex nature of our health care system introduces many barriers to collecting comprehensive, high-quality maternal health data. For example, the national maternal mortality data does not fully capture deaths associated with pregnancy. The official counts that the National Center for Health Statistics releases capture deaths that occur within six weeks postpartum, but CDC analysis tells us that about a third of maternal deaths happen between six weeks and one year postpartum, after they would be included in most published national statistics. At the state level, the data is equally limited. Some governments include deaths that occur after six weeks, while others don’t. In this way, state-level maternal mortality data often lacks standardization and interoperability, making it more difficult to compare data with other states or national benchmarks.
At the state and local level, Maternal Mortality Review Committees (MMRCs) review pregnancy-associated deaths, defined as those that occur within a year of pregnancy. Members of these committees— physicians, midwives, doulas, nurses, and other experts—look at death certificates, vital statistics data, and institutional records (e.g. medical records) to understand the conditions surrounding deaths. The work of MMRCs has the potential to prevent over 20 percent of maternal deaths, but MMRCs have historically worked independently of each other with little standardization across jurisdictions. Some states and jurisdictions require their MMRCs to determine whether deaths were preventable, while others don’t.
The CDC recently provided awards to 39 states through the Enhancing Reviews and Surveillance to Eliminate Maternal Mortality (ERASE MM) to help them improve how they identify, review, and characterize pregnancy-related deaths, as well as support them in developing recommendations to prevent future mortality. Most states involved in the ERASE MM program have started to implement CDC guidelines for standardized data collection and recommendation development. Despite these efforts, MMRCs continue to face challenges. Errors in death certificates that MMRCs rely on are common. In one study, almost 53 percent of death certificates surveyed had an error. Providers who complete death certificates often have insufficient training on how to complete them properly. Inconsistent, poor-quality data and data use make understanding and preventing future maternal mortality more difficult. To address these data collection challenges, policymakers could:
- Enable the Centers for Disease Control to develop standardized definitions that state agencies use when collecting and reporting maternal health data and ensure that federal agencies, including the National Center for Health Statistics, follow these definitions. This could include expanding the definition of maternal death to include deaths that occur within a year postpartum. The federal government could provide states and agencies with training and peer-to-peer learning opportunities to facilitate this transition.
- Support state vital statistics agencies in providing training programs to ensure that death certificates are completed correctly and consistently.
- Provide sufficient and predictable funding to MMRCs that allow them to continue to standardize investigations of pregnancy-related deaths and improve recommendation development processes.
Improving Access to High-Quality Data
Increasing access to data is just as important as improving how we collect data. Multi-stakeholder interviews with state Medicaid programs, MMRC members, and Perinatal Quality Collaboratives conducted by AcademyHealth and NORC at the University of Chicago revealed high barriers to data sharing across government agencies. States and localities sometimes cannot access Medicaid claims data and Veterans Affairs records or even data from other departments within their jurisdiction. State-to-state data sharing is also important, especially for pregnant people who reside in multiple states during their pregnancy. Without data sharing, it’s difficult for government agencies to capture and respond to underlying contributors to poor maternal health outcomes and disparities throughout the perinatal period.
Currently, researchers do not have ready access to complete, interoperable maternal health data either. Electronic health records often lack standardized places to enter pregnancy start and end dates. Pregnant people may seek services from many different providers, such as midwives, doulas, primary care providers, and OB-GYNs. Different health care and social care providers use different electronic health systems, and data lags can make it difficult for researchers to provide real-time insights. When the systems these providers use aren’t interoperable, it’s difficult to track and analyze health data longitudinally. Standardizing data and data exchange could make this information more accessible to researchers.
During the pandemic, government agencies and health systems started to develop tools to allow health care providers to share sensitive medical information while protecting patient privacy. Databases that aggregate data across providers can provide large sets of timely data to researchers and allow them to understand real-time challenges in maternal care. But it comes at a cost. These datasets are often prohibitively expensive to researchers, especially now that pandemic-era discounts are no longer available. The United States Core Data for Interoperability’s (USCDI) guidelines outline a set of common elements that could allow for easier information sharing across health systems, including pregnancy status. These guidelines could give maternal health researchers access to standardized patient data. To further improve data accessibility, policymakers could:
1) Support states in developing advanced data use agreements (DUA), which could improve intra- and interjurisdictional data sharing. Some states have already succeeded in implementing DUAs. To help more states adopt DUAs, policymakers could provide opportunities for peer-to-peer learning.
2) Continue to fund initiatives to standardize maternal health data elements across EHR systems, such as the USCDI.
Unless researchers and government agencies have access to high-quality and timely maternal health data, they cannot develop interventions that target underlying contributors to maternal morbidity and mortality or find ways of eliminating racial disparities. These suggestions could improve the quality of data that researchers have access to and allow them to develop more effective strategies to address maternal health challenges.
The opinions expressed in this blog post are the author's own and do not necessarily reflect the view of AcademyHealth or of their respective affiliated employers/organizations.