Are Hospitals Really Reducing Maternal Mortality Rates?

Darrell Huff wrote his classic How to Lie with Statistics long before many of today’s researchers were born, but it remains a goldmine of information for the non-statistician to ferret out problems with data and conclusions in scientific studies. His perspective is sorely needed in the world of today’s metaverse where sound bites substitute for reasonable conclusions.

Almost the first thing to look for is bias—the laboratory with something to prove for the sake of a theory, a reputation, or a fee; the newspaper whose aim is a good story; labor or management with a wage level at stake.

The CDC reports the maternal mortality rate was 32.9  per 100,000 live births in 2021. In 2020, it had been 23.1 per 100,000 live births, and in 2019, 20.1 per 100,000 births. The Commonwealth Fund reports that 31 percent of maternal deaths occur during pregnancy, 17 percent around time of delivery, and 52 percent up to a year postpartum. 

There has been a lot of discussion about how we can reduce these maternal mortality numbers. In June 22, 2023, JAMA published “Trends in Maternal Mortality and Severe Maternal Morbidity During Delivery–Related Hospitalizations, 2008 to 2021.” This study included more than 11.6 million delivery-related hospitalizations from 2008 to 2021. The study concluded that the in-hospital maternal mortality per 100,000 discharges declined from 10.6 in 2008 to 4.6 in 2021. How can a study return this kind of conclusion when we all know the maternal mortality rate is continuing to go up?

First of all, the hospitals from which data was collected are not given. How many of these hospitals were metropolitan and how many rural? Indeed, collecting data as far back as 2008, one wonders if some of these hospitals are even still operating. Or whether the results of the study would have been so low if the time span were, say, 2018 to 2021, when maternal mortality rates have risen markedly. And finally, the hospital stay is a small segment of the data on maternal mortality, covering only a few days as opposed to data for the prenatal and postnatal periods. With the maternal mortality rate rising as rapidly as it is, one can only wonder why anybody would measure this very small segment of maternal deaths, a relatively small and hence potentially misleading segment regarding the maternal mortality data.

There were some other limitations to the study. The data was taken from a “large, all-payer (including Medicaid), geographically diverse administrative database comprising more than 1200 US hospitals and health systems,” the PINC AI database. The information about the database indicates it includes rural as well as urban hospitals, but no indication of numbers is given. This database, according to the authors, is made up of more than 1200 hospitals and health systems which represent approximately 25 percent of U.S. inpatient admissions. The data is collected from hospitals in U.S. census geographic divisions, but the percentage of data from one census area to another varies widely. The Northwest Central data accounts for 7 percent of the data collected while the South Atlantic area provides 23.5 percent of the data. Maternal mortality rates vary considerably from state to state. If certain areas of the U.S. supply significantly more of the hospital information in the PINC AI database, bias is inevitable.

And finally, there was a significant conflict of interest in that the two principal investigators actually worked for the Office on Women’s Health, US Department of Health & Human Services during this study.

The following paragraph is a quote from the US Department of Health and Human Services Article June 22, 2023.

HHS continues to implement important initiatives focused on improving maternal and infant health. The Centers for Medicare and Medicaid Services recently finalized two new hospital quality measures on severe obstetric complications and the rate of low-risk Cesarean deliveries, as well as a new “Birthing-Friendly” designation that requires hospitals engage in such quality improvement activities. This new “Birthing-Friendly” designation and other initiatives included in the White House Blueprint for Addressing the Maternal Health Crisis-PDF demonstrate that HHS and the entire Biden-Harris administration are prioritizing efforts to improve maternal health outcomes.

         The Birthing Friendly designation is concerned primarily with helping mothers breastfeed successfully, and waves its hand in the direction of giving more attention to low income women. One of the goals is to establish a checklist of hospital services to gain certification as a birthing-friendly hospital. What always seems to be missed by those studying the high maternal mortality rate in the U.S. is that all women need the attention which only seems to be considered needed for women with high risk pregnancies or limited access to healthcare.

         This article, then, appears to be what Darrell Huff tells us carefully selected data is meant to broadcast. Hype on how the current administration is successfully addressing maternal mortality issues. As Darrell Huff says:

Look for conscious bias….It may be selection of favorable data and suppression of unfavorable.

Unfortunately, we often use explanations or excuses for our high maternal mortality rate by citing comorbidities such as diabetes, age, smoking, or obesity. The causes of maternal mortality seldom appear suddenly. But someone needs to be watching for the early signs of problems in a pregnancy and our current healthcare system does not provide the needed oversight which keeps a low-risk pregnancy from developing into a high-risk one. The stories of maternal deaths in the media almost always involve the complaints of the patient being dismissed by the nurses and doctors providing care.

The U.S. maternal mortality rate is higher than any other developed country, and the rate is increasing. No amount of looking at a narrow segment of the problem such as research on a narrow set of in-hospital deaths or the creation of “birthing friendly” systems will fix the maternal mortality problem in the United States. None of these studies look at the data from countries with low maternal mortality rates to see how it is done.

Darrell Huff ends his book with a quote from Mark Twain:

There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact.