Bad Data Is Partly to Blame in Texas’ Maternal Mortality Spike


Connecting state and local government leaders

Maternal health data in the U.S. is so imprecise the federal government hasn’t released an official annual count of pregnancy-linked fatalities since 2007.

A national study published in the September 2016 issue of the journal Obstetrics and Gynecology dropped a bombshell on the Texas public health world. That study’s authors had found that within a two-year period—from 2010 to 2014—the maternal mortality rate in the state had doubled.

The researchers were so surprised by their findings they dove back into the Texas data with the hope of explaining how or why the mortality numbers had climbed so dramatically. Their follow-up research, published Thursday in the journal Birth, found that at least part of that spike can be attributed to over-reporting of maternal mortality data.

In the new study, the authors isolated the deaths into two time periods—from 2006 to 2010 and from 2011 to 2015. When they did that, they found that the number of deaths had almost doubled, from 366 in the first time period to 668 in the second and, notably, that deaths in women over the age of 40 made up a disproportionate share of the increase. The maternal mortality rate for those women over the age of 40 had more than doubled—a rate that the study’s authors found to be “implausibly high.”

Imprecision on death certificates may be to blame for some of this over-reporting. When someone dies, the immediate, intermediate and underlying causes of death are marked on the death certificate. When the deceased is a pregnant woman and the underlying cause is listed as “other” or similarly nonspecific causes, it makes it difficult for researchers to ascertain whether or not the death was directly linked to pregnancy or childbirth.

The researchers found that from 2011 to 2015, deaths attributed to nonspecific causes accounted accounted for 197 of the 302 additional deaths. They also found that the deaths of women over the age of 40 were more likely to be ascribed to vague causes than the deaths of younger women.

Imprecise data does not explain away Texas’ rise in maternal mortality entirely. When the authors of the study eliminated the deaths that were listed as having unspecified causes, and the deaths of women over 40 from the data set, they were still presented with a 36 percent increase in mortality from the first 5-year period to the next.

The study’s authors also want to be clear that the results from this follow-up research should not be interpreted as good news. Rather, they highlight a twofold public health problem.

“One is a sharp increase in the maternal mortality rate in recent years,” Marian MacDorman, a research professor at the University of Maryland Population Research Center and one of the study’s authors, told KUT, the National Public Radio affiliate in Austin.

The second problem, MacDorman said, is “a lack of reliable data to better characterize and further understand the increase.”

“You don’t know if the women are dying because of hemorrhage after the birth, or opioid addiction or high blood pressure,” MacDorman added. “You don’t know what’s going on, so you can’t design a program to prevent those deaths.”

Texas isn’t the only state with a maternal mortality data problem. In fact, maternal data in the U.S. as a whole is so imprecise and incomplete that the federal government hasn’t released an official annual count of pregnancy-linked fatalities since 2007.

“Our maternal data is embarrassing,” said Stacie Geller, a professor of obstetrics and gynecology at the University of Illinois College of Medicine during a conversation with ProPublica in October 2017. “Maternal health in the U.S. is simply still not a priority. It’s not interesting. Preventable maternal deaths are not in the basement of our priorities, they are in the sub-basement.”

Quinn Libson is a Staff Correspondent for Government Executive’s Route Fifty and is based in Washington, D.C.

NEXT STORY: California Says the Only Safe Way to Talk on Your Cell Phone Is to Text