Frequently Asked Questions

Summary of 2018 Findings

What are the key points in NCHS’ reports on maternal mortality statistics?

• The National Center for Health Statistics (NCHS) has released three reports that provide significant new insight into maternal mortality measurement in the United States, along with detailed data files for researchers.
• NCHS has released the first national estimate of maternal mortality in more than a decade. The new rate is based on the use of a checkbox that helps identify deaths related to pregnancy.
• The national maternal mortality rate (MMR) in 2018 was 17.4.
• The rate for non-Hispanic black persons in 2018 was more than twice that for non-Hispanic white persons, 37.3 compared with 14.9. Rates also increased substantially by age, with rates for women aged 40 and over roughly eight times the rate for women under 25 (81.9 and 10.6, respectively).
• 2018 data for selected states can be found at Maternal Mortality by State, 2018 pdf icon[PDF – 228 KB].
• The MMR is more than double the rate reported before the checkbox was added, but a rigorous evaluation confirms that the increase in reported rates is almost entirely because of changes in reporting methods. After evaluating more comparable data, the rate has not significantly changed since 1999.
• Based on this evaluation, NCHS has also identified instances where application of the checkbox information according to coding rules led to misclassification of maternal deaths, and NCHS is making changes in coding rules and reporting to make data more accurate.
• In addition to the reports, NCHS has released a public-use data file for 2018, and will be releasing detailed data files for researchers from 2015 through 2018. These 2015–2018 data files for researchers will include all of the details included on the death certificate, so that researchers can independently assess the influence of the checkbox.

Process: Changes in Recoding Maternal Deaths

Why did it take so long for states to adopt the new checkbox?

• Vital records are governed by state laws and procedures, and states make independent judgements on when to change certificates.
• Outdated technology presented challenges to updating vital registration systems, and some states deferred changes in content until technology could be upgraded.
• No new funding was provided to states when the U.S. Department of Health and Human Services recommended a new approach to data collection. States relied on funding from diverse sources and variable cycles in order to move forward.

Why was this detailed evaluation done now, rather than when NCHS started to see state practices diverging?

• NCHS recognized the growing lack of comparability across states and the inability to aggregate state data to make national estimates. This led NCHS to suspend publication of a national MMR in 2007 and to note the comparability issues in annual vital statistics reports.
• Once all states had added a checkbox, it was possible to fully assess the impact of the change and to calculate a national rate.

Are there lessons from the way the checkbox was adopted and used?

• For any future revision of the standard certificates, NCHS and state partners will need to more explicitly address implementation issues, including technology and funding availability. NCHS will also need to develop more explicit criteria for when new items are used to produce official statistics.
• Data quality questions need to be addressed as they are recognized, and clear guidance needs to be provided to data users. NCHS can be more responsive to evolving research and commentary about quality issues like those raised about maternal mortality.
• NCHS’ evaluation of maternal mortality reinforced the need for modernizing the underlying vital statistics system and the electronic systems that support vital registration. These systems need to be updated to be more flexible and responsive to changing content, as well as to provide for greater real-time quality checks as deaths are being recorded and causes of death defined.
• Short of a fundamental modernization of the system, NCHS will explore options for adding quality to the system, including approaches such as adding pop-up prompts that alert certifying physicians to validate information that may seem to be in error.

Trends and Changes in Understanding

Are we overcounting now, or were we undercounting then?

• NCHS concluded that there are errors in both directions, making it important to update procedures and reporting.
• There was clear evidence in the early 2000s that many maternal deaths were being missed. The addition of the checkbox had the intended effect of a significant increase in reporting. But in NCHS’ recent evaluation, and other independent research, instances have been identified where deaths were inappropriately being coded as maternal. NCHS had previously recognized that the checkbox was likely to be incorrectly checked for women aged 55 and over. NCHS’ review led to the conclusion that there was a high likelihood this was occurring in many deaths at a younger age.
• NCHS has taken steps to minimize overcounting with women of older ages.

Many have raised alarm about increasing rates. Were they wrong?

• Previous reports were based on the best available evidence at the time, but as noted, most of these studies were based on incomplete data and evolving data collection methods from the states.
• New NCHS work does not fundamentally change our understanding of the magnitude of the problem and reinforces the significant black–white differential in maternal mortality.
• The observed increase in MMR reflects changes in the way the data was collected and reported, and was not likely to be the result of an underlying increase in maternal deaths.
• NCHS has also found instances where the checkbox was used in error, inaccurately elevating the MMR to a degree.

Changes in Coding Methods

The reports identify some incorrect use of the checkbox for older women. Why are you sure these are not real pregnancies?

• Multiple studies have evaluated the quality of reporting on the checkbox and concluded that errors increase with increasing age, as seen in a recent studyexternal icon by staff from the Centers for Disease Control and Prevention’s Division of Reproductive Health.
• Relying solely on the checkbox results in a clearly implausible number of pregnancies and related deaths at older age groups. For example, it was found that hundreds of women over 75 had checkbox entries showing them as pregnant within the year preceding death.

Why did you select 44 as the cutoff age for sole reliance on the checkbox?

• The cutoff was based on NCHS’ assessment of errors.
• More importantly, death certificates for women of all ages are evaluated for indications of pregnancy. Any mention of pregnancy or a pregnancy-related condition in the reported cause of death would result in the death being coded as a maternal death, regardless of the checkbox.

State-Level Data on Maternal Mortality

Is NCHS releasing data by state?

• For selected states, data can be found at Maternal Mortality by State, 2018 pdf icon[PDF – 228 KB].
• To protect confidentiality, the rate is not shown for states with fewer than 10 maternal deaths.
• There are significant limitations in the quality of subnational data. NCHS does not advise using the 2018 MMR data to make comparisons across states because of small numbers. Readers are advised to read the notes accompanying this table to understand these limitations.
• Detailed data files for multiple years released by NCHS allow further analysis of state-level data.
• State-level estimates from NCHS rely on coding that makes data more comparable across states. As a result, estimates from NCHS may differ from those published by the states themselves.
• Detailed files to be released for 2015 through 2018 enable researchers to better understand state-specific data.

Public Access to Underlying Data

Can researchers and advocates have access to the underlying data so they can validate NCHS’ work, or make different estimates using different approaches?

• NCHS’ public-use data file has the full information from the certificate that is provided by states. As a result, it is possible for researchers and advocates to make choices on how to analyze and track maternal mortality.
• By resuming the publication of an official MMR, NCHS intends to clarify the understanding of maternal mortality and allow for the tracking of trends using a standardized approach over multiple years.

For more information, visit Data Files and Resources.