Data Strategy

Every maternal mortality review committee (MMRC) should have a comprehensive data strategy documented in its policies and procedures. This data strategy should include the following:

  • Methods for identifying maternal death cases (e.g. through an agency or agreement)
  • Methods for assigning cases to abstractors
  • Methods for collecting, abstracting, and storing records and data
  • Methods for recording, collecting, and storing MMRC deliberations and recommendations
  • Methods for analyzing, summarizing, and reporting data on maternal death cases

An MMRC’s data strategy should employ best practices in the collection, abstraction, and analysis of maternal mortality data. Adhering to these standards ensures data collection is consistent and comparable across cases within the review and with other jurisdictions. This offers a solution toward solving the existing national dilemma: as most reviews collect and analyze data in a nonstandard way, data cannot be aggregated to create a national picture of maternal mortality in the United States.

MMRCs manage data separately for two distinct parts of the review process: case identification and case review. Potential cases are identified through a partnership between the MMRC, the state Vital Records office, and epidemiologists. This partnership conducts a routine review of deaths to determine if any of the deaths are pregnancy-associated. Death certificates may indicate a pregnancy-associated death through the pregnancy checkbox or a cause of death code related to pregnancy. However, death certificates are not sufficient to comprehensively identify all pregnancy-associated deaths. To identify further pregnancy-associated deaths, the MMRC should coordinate with the state vital records office and epidemiologists to ensure routine linkage is conducted of death certificates of females of reproductive age for a given time frame with infant birth or fetal death records during a related time frame. For a pregnancy-associated death identified from the death certificate alone, there may be value to the committee for a representative of Vital Records or the MMRC to confirm that the death occurred during pregnancy or within one year of the end of pregnancy before sending the case for abstraction and committee review. See below for an outline of how to create a file which includes your starting set of pregnancy-associated deaths.

  1. Pull death file for women of reproductive age (For example, 10-55 years)
  2. Link fetal deaths and births for the 365 days that preceded the death
    1. 2017 deaths linked to 2017 and 2016 births and fetal deaths
    2. 2016 deaths linked to 2016 and 2015 births and fetal deaths
    3. Etc.
  3. Collect those that link into a file (Linked File)
  4. Identify all deaths to women with a pregnancy checkbox marked indicating:
    1. Pregnant at the time of death
    2. Not pregnant, but pregnant within 42 days of death
    3. Not pregnant, but pregnant 43-days to 1 year before death
    4. Unknown if pregnant within the past year
  5. Collect those into a file (Checkbox file)
  6. Compare Linked file to Checkbox file and remove duplicates from the Checkbox file
  7. Examine the remaining Checkbox deaths with “Unknown if pregnant within the past year” marked
    1. Look at the text from Part I and Part II of the death certificate’s cause of death section for words that would indicate a relationship to pregnancy, such as “postpartum, ectopic, eclampsia, pregnancy, preeclampsia”
    2. Remove the deaths that have the unknown checkbox marked, but there are no words that indicate a relationship to pregnancy in the cause of death text
  8. Combine the remaining Checkbox file with the Linked file. This is your starting set of pregnancy-associated deaths.

Note that some MMRCs have additional protocols for identifying pregnancy-associated deaths, including through direct hospital reporting and obituary searches or social media. If the death was not already represented in the pregnancy-associated death file, then it would be added by the designated MMRC contact. MMRCs use the created pregnancy-associated death file to populate a Pregnancy-Associated Deaths Preliminary Case Classification Spreadsheet. To view a Case Identification Flow Chart, click here.

The MMRC may designate a subcommittee to preliminarily classify pregnancy-associated cases to be sent for abstraction. The subcommittee is typically comprised of a clinician, an epidemiologist, an abstractor, and a coordinator. Once the list of cases to be sent for abstraction is determined, a comprehensive database, i.e. MMRIA, should support standardized case abstraction, case narrative development, documentation of committee decisions, and routine analyses. To view a Standard Committee Review form, click here.

Until recently, MMRCs lacked a standard approach for collecting data. Without consistent data collection, jurisdictions could not share their findings to more fully understand the causes of maternal mortality and develop shared approaches to eliminate preventable maternal deaths. To address the need for a common data language, the Centers for Disease Control and Prevention (CDC) partnered with maternal mortality review committees and subject matter experts to create the Maternal Mortality Review Information Application (MMRIA).

MMRIA supports standard case abstraction, case narrative development, documentation of MMRC decisions around preventability and contributing factors to a death, and routine analyses. MMRIA reflects lessons learned from implementing a previous version of the system, the Maternal Mortality Review Data System (MMRDS), among 13 state MMRCs.