The U.S. Can Worry Less about Maternal Mortality

According to a recent CDC report, the United States’ maternal mortality rate (MMR) in 2021 was 33 deaths per 100,000 live births. As several media outlets have noted, this is a vastly higher rate than that of other rich countries. To pick out a few, it is 3 times higher than that of Canada and the U.K., more than double that of Russia, and 50% larger than that of China. On top of this, the U.S. rate increased 40% since 2020, which saw a MMR of 24. This all seems really bad! But let’s take a closer look.

First of all, note that each comparison above was relative as opposed to absolute. When it comes to changes in risk, you ought to consider both relative and absolute risks. If the risk of something is very small, a small absolute increase in that risk would not lead to a surge in the number bad outcomes, even if the relative increase is large. Famously, an oral contraceptive pill reportedly doubled the risk of thrombosis in women, leading to public outcry. However, in absolute terms, the risk of thrombosis rose from 1 in 7,000 to 2 in 7,000, meaning the risk was miniscule with or without the contraceptive.

Let’s take this logic to MMRs. Again, there were about 33 maternal deaths per 100,000 live births, a risk of 0.033% for a given pregnancy. First, note that the 40% increase from 2020 was an absolute risk increase of about 0.009 percentage points. Second, repeating this exercise for the countries listed above, you can see that the absolute risk in other nations is only lower than that of the U.S. by at most 0.02 percentage points, or one-fiftieth of a percent. If we think these other countries are doing a good job controlling their MMRs, then this suggests the U.S. is not “ in a maternal health crisis.”

There’s more. While the World Health Organization has a standard international definition for a maternal death, used for all of the countries listed above, the process by which deaths are classified as “maternal” differs both by country and over time. For example, the U.K. has an incredibly thorough system for classifying deaths, involving at least 6 separate reviewers for each potential maternal death. The U.S. had a similar (though less thorough) methodology up until 2003 when, because of research indicating we were undercounting maternal deaths, we added a pregnancy checkbox (box 36) to our federal death certificate template. Information on both a woman’s cause of death and her pregnancy status are now combined to classify a death as maternal or not. Specifically, according to a 2020 CDC report, a death is considered maternal if the woman was pregnant or recently pregnant AND the cause of death was not incidental (e.g., homicide), accidental, nor “due to conditions that are not caused by nor complicated by the pregnancy.” Regarding this final condition, the authors note the following:

[T]here is typically not enough information on the death certificate to determine whether the reported conditions were incidental to pregnancy, and no standard list of medical conditions that are incidental to pregnancy exists for coders (or the automated system) to use. As a result, coders would assume that there is a relationship between pregnancy and the medical condition and assign a maternal code.

Looking at state-by-state variation in adoption of the “pregnancy-checkbox,” researchers at the CDC found that this policy change roughly doubled reported MMRs (see Figure 1 from their report below).

Perhaps we were under-reporting maternal deaths before the 2003 update, or perhaps we are over-reporting now. Regardless, if this process is not similar between countries, then you’re making apples-to-oranges comparisons. Unless we implement a global standard maternal-death classification system, there will always be error in cross-country comparisons. When the numbers you’re comparing are all very small, this is a big deal, relatively speaking. This also means we should be skeptical of headlines like U.S. Maternal Mortality Hits Highest Level Since 1965 since we know 1965 units are not the same as 2021 units.

In the U.S., the absolute risk of maternal mortality in 2021 was low, even if it was lower in 2020. Further, early indications suggest the 2022 MMR will be lower than that of 2021. When it comes to maternal outcomes in the U.S., our time is better spent focusing on maternal morbidity, the longer-term rise in MMRs, links to COVID, and racial disparities.