— Rep. Don Beyer (D-Va.), in remarks at a House Ways and Means Committee hearing on maternal mortality, May 16, 2019
Many times, the corrections receive far less attention than the original headline. Here’s a good example.
In arguing that laws that limit abortion rights increase maternal and infant mortality, Rep. Beyer cited some alarming data about Texas and asked a witness about the connection between the two.
The witness, Lisa Hollier, the immediate past president and interim chief executive of the American College of Obstetricians and Gynecologists, sidestepped the question. But, in as polite a way as possible, she told Beyer that his data was bunk. She should know: She chairs the Texas task force that examined the original findings that were quoted by Beyer.
“When we analyzed the data, we identified there was significant over-reporting of deaths based on the death certificate data,” she testified, saying the death certificates were not accurate and so the maternal mortality rate was “not as high as previously reported.”
When The Fact Checker contacted Beyer’s office, the lawmaker immediately admitted error. “We weren’t trying to deceive people here, we just hadn’t seen the subsequent reporting on the study in question,” said Aaron Fritschner, Beyer’s communications director. “An honest mistake.”
Regular readers know that we generally do not award Pinocchios when a politician admits an error, but this is still worthy of a fact check. In fact, we found a witness at the hearing submitted testimony with the same error, so clearly the news about the new data has not traveled far enough.
The original study, by a team of researchers headed by Marian F. MacDorman and published in Obstetrics & Gynecology in 2016, made headlines. The maternal mortality rate in Texas had climbed sharply – in fact, doubled – between 2010 and 2012, even though it had only modestly increased from 2000 and 2010. The study found that 148 Texas mothers died of pregnancy-related complications in 2012, compared with 72 in 2010.
The researchers admitted they found the result puzzling and even dubious.
“There were some changes in the provision of women’s health services in Texas from 2011–2015, including the closing of several women’s health clinics,” the report said. “Still, in the absence of war, natural disaster, or severe economic upheaval, the doubling of a mortality rate within a 2-year period in a state with almost 400,000 annual births seems unlikely.”
The reference to the closing of women’s health clinics – with a footnote to Texas barring Planned Parenthood affiliates from participation in a state health program for the poor – spawned speculation by abortion rights advocates that the startling jump in deaths was tied to the Texas law. Beyer did more than speculate; he made a causal connection between the funding cut and the maternal mortality rate.
But when the task force headed by Hollier examined and cross-referenced the actual records of 147 women listed as having died of pregnancy-related complications, they found many had been erroneously counted. Only 56 of the deaths fell under the definition of maternal mortality — any pregnancy-related death while a woman is pregnant or within 42 days of giving birth, excluding accidental or incidental causes such as car crashes or homicides.
That changed the 2012 Texas maternal mortality rate from 38.4 deaths per 100,000 live births to 14.6 per 100,000 live births. The task force only examined data from that year, so it’s still unclear how much the mortality rate changed from 2010, but the task force is continuing its research.
"We will use this new method to confirm maternal deaths and calculate maternal mortality ratios for additional years so that trends can be assessed,” the report said. “We do know that when using the standard method for identifying maternal deaths, the Texas maternal mortality ratio has decreased from 2012 to 2015.”
The task force’s study on the revised Texas data was also published in Obstetrics & Gynecology in May 2018 – a year ago. So it’s not as if this is recent news.
As ProPublica has documented, maternal mortality statistics in the United States are incomplete and highly suspect, prone to undercounting and overcounting. Since 2003, states have added a checkbox to death certificates, asking whether a person was pregnant when they died or had been in the previous year. But the data collection is not consistent, and the addition of a checkbox appears to have resulted in many false positives.
As a result, since 2007 the United States hasn’t had an official annual count of pregnancy-related fatalities, or even an official maternal mortality rate. So anyone trying to tie a link between abortion laws and mortality rates is going to run into trouble.
Interestingly, on the same congressional panel as Hollier was Patrice Harris, president-elect of the American Medical Association Board of Trustees. Her prepared testimony also contained a reference to the original flawed numbers – and, like Beyer, drew a direct connection to the cut in Texas funding to Planned Parenthood.
“In Texas, maternal mortality rates were stable from 2006-2010. The passage of Texas House Bill 2 led to the defunding of health clinics offering family planning and the closure of half of family planning clinics relied on by poor women and women of color (MacDornan). Following this, the rate of maternal death in Texas doubled, with other states’ rates increasing more slowly and California’s rates diminishing. Conclusions of this paper hypothesized a connection with these two events (Ravich).”
Chuck Donovan, president of the Charlotte Lozier Institute, which opposes abortion rights, sent a letter the day before the hearing to members of the committee, warning that witness testimony had been submitted citing the flawed data. He pointed out to lawmakers the original study had been corrected. But the missive apparently got lost in the mail room.
A spokesman for AMA said the testimony had relied on a CNN report about the original study, but staff had missed an editor’s note at the top of the article saying that a subsequent study found the numbers had been inflated by two-thirds. He said the AMA would be contacting the Ways and Means Committee to correct the record.
The Pinocchio Test
In ordinary circumstances, Beyer would have received Four Pinocchios. We try not to play gotcha here at The Fact Checker, and he quickly admitted his mistake. In fact, he was corrected on the spot, during the hearing, given the unusual circumstance that he directed his question to the very person who had debunked the data he cited.
Still, the fact that Beyer’s staff originally missed the revision, as did the AMA in its testimony, demonstrates that the news that the original study was based on flawed data has not traveled far enough. It’s been a year since Hollier’s task force published the revised numbers – more than enough time to eliminate this talking point. Lawmakers should not try to make public policy by relying on bad data.
Send us facts to check by filling out this form
Sign up for The Fact Checker weekly newsletter
The Fact Checker is a verified signatory to the International Fact-Checking Network code of principles