The US doesn’t care about mothers —

A crummy drop-down menu appeared to kill dozens of mothers in Texas

This doesn’t mean maternal mortality in Texas or the US is OK. It’s still horrific.

A crummy drop-down menu appeared to kill dozens of mothers in Texas

In the US, the rate of women dying from pregnancy and childbirth is higher than in any other developed country—much, much higher. And we’re bucking the global trend of improving the situation. While the rest of the world largely saw its maternal mortality rates drop by more than a third between 2000 and 2015, the US was one of the few countries that seemed to experience increases in the rate of women dying from pregnancy-related causes.

The state of maternal health in the US is so grim that researchers can’t even get quality data on the deaths. In fact, the country has not published an official maternal mortality rate since 2007 due to the lack of accurate data from individual states. In 2016, a group of researchers didn’t mince words about the situation: “It is an international embarrassment that the United States, since 2007, has not been able to provide a national maternal mortality rate to international data repositories,” the researchers concluded in a study published in the journal Obstetrics & Gynecology.

Now, a new study in the same journal goes further to highlight just how bad the state of maternal health data is in the US. The study links a dramatic rise in maternal deaths in Texas to errors from a poorly designed drop-down menu in the state’s electronic death records system. While the discovery drags down the state’s stratospheric maternal mortality rate, the corrected numbers are still extremely high for a developed country. Moreover, having to make these types of corrections squanders precious resources, experts note.

In an accompanying editorial, maternal health experts wrote that the maternal mortality review committee responsible for the discovery—and others like it— are “spending too much time simply identifying cases and eliminating false positives.” Instead, those groups need to “get back to doing the job they were designed to do—investigating and preventing maternal deaths.”

Labored statistics

The new study on Texas’ data is a direct response to the 2016 study, which was led by Marian MacDorman of the Maryland Population Research Center at the University of Maryland. In it, MacDorman and colleagues tried to amass the ragtag data from individual states to come up with a country-wide estimate of maternal mortality rates.

They determined that between 2000 and 2014, the country’s maternal mortality rate (maternal deaths per 100,000 live births) went from roughly 18.1 to 23.8—a 26.6 percent increase. For comparison, that would put the US second to last out of 31 countries that reported maternal mortality rates to the Organization for Economic Cooperation and Development (OECD). We would be above only Mexico, while Italy reported a rate of 1.2, Spain reported 2.1, Japan reported 3.3, and the UK reported 6.7 in 2014.

Those numbers are a bit low compared with other estimates, which can differ depending on methodology and timeframes for death after pregnancy (from 42 days to 18 months). For instance, a study published in The Lancet (PDF) later in 2016 came up with higher estimates overall, but the US still fared terribly. In that study, the US maternal mortality rate for 2015 was 26.4, while Italy, Spain, Japan and the UK had rates of 4.2, 5.6, 6.4, and 9.2, respectively. Similarly, the Institute of Health Metrics and Evaluation—part of the University of Washington—pegged the US’s 2015 rate at 29.4, while Italy, Spain, Japan, and the UK had rates of 3.9, 5.0, 5.9, and 6.8, respectively.

Part of the reason the US estimate calculated by MacDorman and her colleagues in 2016 may have been a bit different was that the researchers excluded data from California—which saw a unique decrease in maternal mortality rates—and Texas, which saw a “puzzling” and dramatic increase between 2010 and 2012. Texas’ rate went from an abysmal 18.6 in 2010 to a jaw-dropping 38.4 in 2012, a leap the authors struggled to understand.

“In the absence of war, natural disaster, or severe economic upheaval, the doubling of a mortality rate within a two-year period in a state with almost 400,000 annual births seems unlikely,” the authors concluded.

The new study explains what happened—user error.

Birthing defects

The new study was authored by the Texas Maternal Mortality and Morbidity Task Force, a type of review committee that several states have set up in recent years to try to understand the country’s high maternal death rates. These committees can help shape up state stats, but as an investigation by ProPublica pointed out last year, they have trouble making progress due to a lack of resources. A third of the states that have such committees have no budget for them, and they rely solely on volunteer efforts from health practitioners who focus on maternal health.

The Texas committee set out to try to understand the extremely high maternal death rate in just 2012—at which point the rate had doubled from 2010. There were 147 maternal deaths in Texas during 2012, based on medical codes on death records that indicated maternal death while pregnant or within 42 days of postpartum. The researchers looked into those 147 cases, plus medical records of all women’s deaths during the year, to see if any maternal deaths were missed.

The committee tried to match up women’s death records with birth or fetal death records as well as individual medical records—which could show prenatal care or not—plus any information from an autopsy or death certifier that would indicate pregnancy-related death.

With this method, they found that only 47 of the 147 deaths were clearly linked to pregnancy. Of the other 100, 74 had zero evidence of pregnancy in their medical records, 15 had insufficient medical records to make a call, and 11 deaths occurred beyond the 42-day window after pregnancy to count for this study.

Looking through the death records of all women, the committee found nine other maternal deaths that were previously missed, bringing the 2012 total to 56 confirmed maternal deaths. That works out to a maternal mortality rate of 14.6 deaths per 100,000 live births. That’s still very high, but it's significantly lower than the 38.4 reported before the corrections. If the researchers add in the 15 deaths that didn’t have enough medical data to make a call, the rate jumps to 18.6.

Nursing numbers

The researchers note a possible explanation for why many of the 74 women had obstetric codes without any other evidence of pregnancy. Fifty-six of the 74 (76 percent) had a listing of “pregnant at the time of death” on their death record. This label is listed directly below the “not pregnant within the past year” option on a drop-down menu in Texas’ electronic death-registration system. Health professionals filling out the record may have simply mis-clicked. To add to that concern: between 2010 and 2012, use of electronic death records rose from 63 percent to 91 percent. The authors suggest that the system should have separate boxes for pregnancy status—as well as more training for personnel—to avoid the potential click-fail in the future.

The accompanying editorial response—penned by MacDorman and colleagues—says that “the work of the Texas maternal mortality review committee is laudable, necessary, and entirely appropriate.” But they note that it’s just one year’s worth of data and doesn’t explain the higher rate overall or the recent trend upward.

The solution, they say, is that:

We need to improve data quality in the National Vital Statistics System by working at the national, state, and local levels to better train physicians, medical examiners, and coroners on the importance of the pregnancy checkbox and completing the cause-of-death section; enhance query systems and internal consistency checks in real-time; and review cause-of-death coding procedures in relation to maternal deaths.

All the experts agree that such information and systems are critical to preventing pregnant women and mothers from dying needless deaths from pregnancy-related issues and complications, such as hemorrhaging and infections. A recent investigation by Vox found that California reversed its trend—dropping maternal mortality rate from 21.5 in 2003 to 15.1 in 2014—by doing simple things like providing a "toolkit" for hospitals on safe births and having "hemorrhage carts” at hand that are packed with everything a medical team needs to handle hemorrhaging. During childbirth, a woman can bleed to death in as little as five minutes, Vox notes.

And additional data can help address the massive racial disparities in maternal mortality rates. For instance, even after the Texas death rates were corrected, the maternal mortality rate for black women was still a stunning 27.8. And older mothers, too, face grave statistics. Those over 35 years of age had a mortality rate of 32.2 in Texas after corrections.

As MacDorman and colleagues concluded in their 2016 study: “There is a need to redouble efforts to prevent maternal deaths and improve maternity care for the 4 million US women giving birth each year.”

Obstetrics & Gynecology, 2018. DOI: 10.1097/AOG.0000000000002565  (About DOIs).

Channel Ars Technica