In 2014, I had the opportunity to work with the Health Resources and Services Administration (HRSA) Graduate Student Epidemiology Program (GSEP). My project was focused on evaluating the maternal mortality review committee process in Georgia. Early on, most of my work consisted of becoming familiar with the global burden of maternal mortality (starting with the 2010 Amnesty International special report on maternal health care in the United States– This report criticized estimated maternal mortality ratios that gave the United States (U.S.) a subpar global ranking based on data published by the World Health Organization (WHO), United Nations Children’s Fund (UNICEF), United Nations Population Fund (UNPF), and Word Bank in 2005– despite the U.S spending more than any other country on health care), becoming acquainted with appropriate case definitions for surveillance, and reviewing various maternal mortality review data sources to gain situational awareness of the distribution and determinants of maternal mortality in the United States. Ultimately, this led me to realize that, although maternal mortality was seemingly rare and largely preventable in the U.S, the disparities that existed based on race, socioeconomic status, and quality of care were concerning.
As the heart of my project took off, I learned about key aspects of Georgia’s maternal mortality review committee. The committee, like others similar to it, was tasked with identifying maternal deaths using multiple data sources, reviewing medical records and death certificates to classify cases accurately (i.e. not a case, pregnancy-related, or pregnancy-associated), and evaluating which deaths were truly preventable. The process evaluation that I conducted assessed whether the review process was taking place as intended, best practices were being employed, and pregnancy-related deaths were accurately being classified. The evaluation also provided action points and recommendations that included tracking and reporting dissemination efforts intended to inform provider care, having a unified voice on the topic of maternal mortality in the state (due to conflicting local news headlines or statements made), defining process indicators, and developing a feedback loop for policy and educational changes.
After returning to my home state of Texas and joining the public health workforce, I remained interested in maternal mortality and learned that work was being done here too- in the form of a Maternal Mortality and Morbidity Taskforce. That was about a year and a half ago. More recently, there has been renewed media interest concerning how maternal outcomes (including maternal mortality and severe maternal morbidity) in the U.S. compares to other countries. A visit to the Centers for Disease Control and Prevention (CDC) website provided me with some background on how the CDC tracks severe maternal morbidity in the U.S. and highlighted the main risk factors or indicators for complications during pregnancy. According to the CDC, recent trends show an increase in maternal morbidity rates from 1993 to 2014 for the following indicators:
- Blood transfusions at 399%.
- Acute myocardial infarction or aneurysm at 300%.
- Acute renal failure at 300%.
- Adult respiratory distress syndrome at 205%.
- Cardiac arrest, fibrillation, or conversion of cardiac rhythm at 175%.
- Shock at 173%.
- Ventilation/temporary tracheostomy at 93%.
- Sepsis at 75%.
- Hysterectomy at 55%.
At the same time, rates decreased or remained the same for the following indicators from 1993 to 2014:
- Disseminated intravascular coagulation.
- Air and thrombotic embolism.
- Amniotic fluid embolism.
- Acute congestive heart failure or pulmonary edema.
- Puerperal cerebrovascular disorders.
- Heart failure or arrest during surgery or procedure.
- Severe anesthesia complications
The data collected by the CDC suggests that women who are giving birth later in life or were obese before becoming pregnant/have other underlying health issues may be contributing to increases in maternal morbidity due to not being as healthy as the typical population of women that would be giving birth in the U.S. Cesarean sections are also being performed more frequently which greatly increases the risks of complications for women. On November 14th, 2017, the CDC hosted a Grand Rounds session titled, “Meeting the Challenges of Measuring and Preventing Maternal Mortality in the United States“. William M. Callaghan presented data from the National Vital Statistics Registry which showed an increase from 1999 to 2014 in the maternal mortality rate (MMR) in the U.S (from 9.8 to 21.5 maternal deaths per 100,000 live births). America’s Health Rankings indicates that the U.S. had a rate of 19.9 maternal deaths per 100,000 live births in 2016. There was great variation across states, however, ranging from 5.8-40.7 maternal deaths per 100,000 live births.
Vital statistics collected by the National Center for Health Statistics (NCHS) are only one measure of maternal mortality and are limited in their ability to accurately identify cases. The Pregnancy Mortality Surveillance System (PMSS) enhances NCHS data by adding a pregnancy checkbox to maternal death certificates so that epidemiologists can link them to birth or fetal death certificates. PMSS also includes all maternal deaths occurring during pregnancy or within one year of the end of pregnancy, instead of only those that occur during pregnancy or within 42 days of the end of pregnancy. PMSS uses clinical relevance, rather than Cause of Death codes, to classify cases and the Pregnancy-related Mortality Ratio (PRMR) as its unit of measurement- MMR and PRMR have the same denominator. When comparing the MMR (which is calculated using vital statistics data) to the PRMR, there is an increase in maternal mortality but the increase is not as steep (13.2 to 17.3 as opposed to 9.8 to 21.5) and seems to be leveling off over the past few years. PRMR is more comparable to MMR when restricted to maternal deaths occurring during pregnancy or within 42 days of termination. Once this adjustment is made, maternal mortality appears to be relatively flat from 1999 to 2014. Since the PRMR incorporates clinically relevant elements, specific changes in trends can be observed. For example, between 1987 and 2013, there was a decrease in maternal deaths due to hemorrhaging and hypertension as well as an increase in maternal deaths due to heart conditions. Additional data from PMSS indicates that state maternal mortality rates ranged from ~7 to 33 maternal deaths per 100,000 live births from 2006-2013 (again showing an increase that is lower than maternal mortality rates calculated using vital statistics data). At the same, it shows that some states have 3-4 times the number of maternal deaths than other states. These differences are similar when comparing mortality across race/ethnicity- Non-Hispanic Blacks and American Indian/Alaska Natives have 3-4 times more maternal deaths than other groups. Ultimately, this paints a picture of state-level and racial/ethnic disparities that are occurring throughout the U.S.
Review committees and task forces like the one I evaluated in 2014 have been established in different states in order to review cases of maternal death and/or severe morbidity. However, while individual states have a better ability to implement quality improvement initiatives that are relevant to their specific needs and risk factors, it takes significant political and social motivation to pass legislation that establishes these committees. The CDC provides technical assistance to jurisdictions that desire to establish maternal mortality review committees and identify preventable deaths as well as highlight interventions that can save lives. In fact, it is information from these reviews that have led to the realization that 20%-59% of maternal deaths are preventable. The Maternal Mortality Review Information Application (MMRIA) has more recently been established to create a standardized set of best practices that have been gleaned from maternal mortality review committees across the nation.
So, what’s the point? My evaluation in 2014, the surveillance data that is being collected by the CDC, the committees that are being established, best practices that are being compiled, and media stories that are being written should result in the implementation of evidence-based interventions that save lives. Here are some good examples:
- Earlier this year, California was celebrated as being a state that has seen a 55% decrease in maternal deaths between 2006 and 2013 due to the development of the California Maternal Quality Care Collaborative. The collaborative is a product of the review process and has put in place various toolkits, research articles, and collaborative outreach materials to improve the health of mothers.
- A 2016 report from Saving Mothers Giving Lives, an international collaborative, has shown that countries such as Zambia and Uganda are on track to meet their 5-year goal of reducing maternal mortality by 50%
At the end of the day, even one preventable maternal death is one too many.