Are We Calculating Maternal Mortality Correctly?
January 7, 2020 2:34 PM Subscribe
Our narrow focus ignores mental health, substance abuse, and other issues that shouldn’t necessarily be separated from pregnancy. But when you look at the overall deaths for women in and around pregnancy, without regard to whether the cause is traditionally considered “pregnancy-related,” the number of dead women goes up a lot. And the list of causes of death becomes a very different one: motor vehicle collisions, homicide, suicide, substance abuse. There’s more to the maternal mortality crisis in this country then just physical health and our racist healthcare system.
What's interesting and frustrating about the Slate article is how close it comes to actually showing what it says it does.
If you look at the top causes of death [PDF] amongst women in their late teens to early 30s - whether pregnant or not - you see: motor vehicle collisions, homicide, suicide, substance abuse. Accidents (which includes both motor vehicle collisions and unintentional ODs as well as much less common things like falls, fires and drowning) is the number one cause of deaths for women in all reported age brackets from 15-44; for women aged 15-24, the top three causes are in order accidents, suicide, homicide. Even in the older age brackets up to 44, suicide is the #4 cause of death and homicide #8 at the lowest.
I found the underlying study, and they did an admirable job going through the calculations to show that homicide rates for women aged 15-29 in and around pregnancy (pregnant, or within a year of the end of pregnancy) were about double the rates for women in the same age cohort who were not. The study also showed a small, not-statistically-significant reduction in homicide mortality rates for women over 30. But they didn't do or say anything for any of the other red bars on the graph, presumably because that wasn't the goal of the study. (The study has the exact same graph as Slate, in blue and red, except with error bars because science.) They conclude that around 13% of the pregnancy-related deaths were due to this increased homicide risk.
I did a CDC Wonder request for causes of death for women aged 15-34 in Illinois since 1999, and found that the death rate to homicide and suicide were about the same (3.6 and 3.5 per 100K respectively); that the death rate from motor vehicles was about double that of either homicide or suicide (7.3 per 100K), and that the death rate from accidental overdose about 1.5x that of the homicide or suicide rates (5.5 per 100K).
Looking again at the red bars in the chart, it seems like if pregnancy-adjacent women were twice as likely as other women to die from homicide, then they were actually no more likely than other women in the same age groups to die from motor vehicle collisions or from suicide, and they are actually less likely than other women to die from substance abuse. You'd need an actual epidemiologist rather than someone who is desperately procrastinating to show this with more rigour, of course.
So in the end what happened is there's this journal article, entitled Higher Risk of Homicide Among Pregnant and Postpartum Females Aged 10–29 Years in Illinois, 2002–2011 that shows a clear increased risk to pregnant/postpartum women of homicide; 2.2x; and then makes no claims about other common causes of death amongst the women in this age cohort (but it seems like there may not be any increase at all.) And it gets cited in a Slate story - by a doctor! - that is more than anything else about depression, and contains the statement: This probably contributes to that long red bar of homicide risk, but it’s hard to know by how much.
posted by Homeboy Trouble at 4:31 PM on January 7, 2020 [29 favorites]
If you look at the top causes of death [PDF] amongst women in their late teens to early 30s - whether pregnant or not - you see: motor vehicle collisions, homicide, suicide, substance abuse. Accidents (which includes both motor vehicle collisions and unintentional ODs as well as much less common things like falls, fires and drowning) is the number one cause of deaths for women in all reported age brackets from 15-44; for women aged 15-24, the top three causes are in order accidents, suicide, homicide. Even in the older age brackets up to 44, suicide is the #4 cause of death and homicide #8 at the lowest.
I found the underlying study, and they did an admirable job going through the calculations to show that homicide rates for women aged 15-29 in and around pregnancy (pregnant, or within a year of the end of pregnancy) were about double the rates for women in the same age cohort who were not. The study also showed a small, not-statistically-significant reduction in homicide mortality rates for women over 30. But they didn't do or say anything for any of the other red bars on the graph, presumably because that wasn't the goal of the study. (The study has the exact same graph as Slate, in blue and red, except with error bars because science.) They conclude that around 13% of the pregnancy-related deaths were due to this increased homicide risk.
I did a CDC Wonder request for causes of death for women aged 15-34 in Illinois since 1999, and found that the death rate to homicide and suicide were about the same (3.6 and 3.5 per 100K respectively); that the death rate from motor vehicles was about double that of either homicide or suicide (7.3 per 100K), and that the death rate from accidental overdose about 1.5x that of the homicide or suicide rates (5.5 per 100K).
Looking again at the red bars in the chart, it seems like if pregnancy-adjacent women were twice as likely as other women to die from homicide, then they were actually no more likely than other women in the same age groups to die from motor vehicle collisions or from suicide, and they are actually less likely than other women to die from substance abuse. You'd need an actual epidemiologist rather than someone who is desperately procrastinating to show this with more rigour, of course.
So in the end what happened is there's this journal article, entitled Higher Risk of Homicide Among Pregnant and Postpartum Females Aged 10–29 Years in Illinois, 2002–2011 that shows a clear increased risk to pregnant/postpartum women of homicide; 2.2x; and then makes no claims about other common causes of death amongst the women in this age cohort (but it seems like there may not be any increase at all.) And it gets cited in a Slate story - by a doctor! - that is more than anything else about depression, and contains the statement: This probably contributes to that long red bar of homicide risk, but it’s hard to know by how much.
posted by Homeboy Trouble at 4:31 PM on January 7, 2020 [29 favorites]
Homeboy Trouble, I will slightly counter your argument with different data. In Texas, ProPublica looked at the data around maternal death in greater detail than is typical and unsurprisingly discovered that there are many more deaths that can be attributed to pregnancy than currently are. In fact, what they found was a huge number of women who died in the '4th trimester', aka after pregnancy (they quote the CDC in saying that, "more than half of all maternal deaths in the U.S. now occur following delivery") with pre-existing conditions being by far the bigger factor:
Then again, it's a hell of a tougher sell to argue that we need health insurance for everyone so babies aren't born with holes in their hearts than it is for society to judge pregnant people who end up victims of domestic homicide.
posted by librarylis at 6:23 PM on January 7, 2020 [6 favorites]
In the last few years, maternal health researchers in the U.S. have seen a dramatic shift in what kills, or nearly kills, pregnant women and new mothers. Historically, the leading medical causes were acute crises: hemorrhages, massive infections, blood clots. These days, the greatest risks come from chronic conditions diabetes, high blood pressure, heart disease that can escalate during pregnancy. According to the most recent report of the Texas maternal mortality review task force, preexisting health problems are the most common contributing factor to maternal death in the state.So while we cannot necessarily analyze maternal risk of death due to mental health, homicide, and vehicular death with this data, I do think we can safely say that we are not calculating maternal mortality correctly. Where I think you agree with me is that the 'other causes' that the Slate author mentions might actually merit more attention than homicide and vehicular death.
Then again, it's a hell of a tougher sell to argue that we need health insurance for everyone so babies aren't born with holes in their hearts than it is for society to judge pregnant people who end up victims of domestic homicide.
posted by librarylis at 6:23 PM on January 7, 2020 [6 favorites]
« Older One woman’s quest to read all of Proust | "...it’s about feeling vibrations and frequencies... Newer »
This thread has been archived and is closed to new comments
But, beyond that, she has likely stumbled upon some effects of the utterly crap-tastic (lack of) support system we have for expecting and new parents, which she also carefully doesn't talk about. I wonder how many of those automobile deaths are from driving while over-exhausted from not getting more than 80 minutes of sleep in a row for the past 6 weeks and now also going back to working a full-time job. How much does the strain of trying to figure out the logistics and finances of feeding and caring for a child increase depression and relationship stress and violence? Let alone exacerbating substance use problems? I'm in a ridiculously good position as a new mom and it's forking hard, and I'm a little skeptical that doing more depression screenings is going to help a lot. It'll help some, don't get me wrong, but there are some of these issues that are well beyond the ability of meds and counselling to meaningfully address.
posted by DebetEsse at 4:08 PM on January 7, 2020 [14 favorites]