The Unintended Consequences of the Texas “Heartbeat Law”

Abortion restrictions may have led to a dramatic increase in infant mortality.

I’m highlighting a study this week that shows us how public policy has effects that can extend well beyond what was intended. It’s about how the choices we make, or perhaps the choices our representatives make, can harm people that they did not intend to harm. It’s a study about Texas SB-8, the so-called “heartbeat bill”. It’s a study about abortion restrictions. And how the strict abortion rules in Texas may have led to a significant increase in the infant mortality rate.

Let’s start with the timeline. May 19, 2021 – Texas Senate Bill 8 (SB8) passes the Texas legislature and is signed into law by Governor Greg Abbott. The bill takes effect on September 1st, 2021, and restricts abortion in the lone-star state once a fetal heartbeat is detectable – about 6 weeks into a pregnancy.

It further provides a mechanism for public individuals to sue those who performs or facilitates an illegal abortion for a minimum of $10,000. The law makes no exceptions for rape or incest, and, importantly for this study, no exceptions for congenital anomalies. The only exception is if there is imminent risk of death or severe bodily harm to the mother.

Researchers from Johns Hopkins wanted to understand some of the ramifications of this law. As reported in this paper in JAMA Pediatrics, they hypothesized that with restrictions on abortions due to congenital anomalies, there would be a higher rate of neonatal and infant mortality once the law was enacted.

Source: Gemmill et al. JAMA Pediatrics. 2024

Congenital anomalies are, after all, the number one cause of infant mortality in the US.

Source: CDC

This is somewhat tricky data to analyze. Their primary data source was the Provisional Multiple Cause of Death Database hosted by the CDC.  This is essentially a dataset of death certificates which include a primary cause of death, multiple secondary causes of death, and demographic data, including age, but not gestational age at birth – an important limitation.

In fact, since the researchers couldn’t tell whether the infants who died were born full-term or not, they took a conservative approach, treating any death that occurred from March 2022 on as deaths that occurred after the policy change – this is 6 months after the law went into effect. Since common congenital anomalies are typically detected at a 12-week ultrasound, this seems reasonable, but we should be aware that some infants who, in this framework, are classified as having died before the law took effect actually were exposed to the consequences of the law. A bit of misclassification bias here which would tend to bias the overall results towards the null.

And those overall results?  I’ll give you the numbers, but a picture is worth quite a bit more here.  What you are seeing is neonatal (that’s within 28 days of birth) and infant (within 1 year of birth) mortality rates over time in Texas.

Source: Gemmill et al. JAMA Pediatrics. 2024

There’s a pretty clear increase after the law went into effect.

Putting some numbers to it, in 2022 – the infant mortality rate increased by 12.9% in Texas, while it increased by 1.8% in the rest of the US. The neonatal mortality rate went up by 10.4% in Texas and 1.6% in the rest of the US.

Source: Gemmill et al. JAMA Pediatrics. 2024

As expected, congenital anomalies were the #1 cause of infant mortality in Texas and the broader US – but the rate increased by 22.9% in Texas in 2022 compared to a decrease of 2.9% in the rest of the US. Sudden infant death syndrome increased by 10.5% in Texas, and 3.3% in the rest of the US. Necrotizing enterocolitis – a horrible complication – remained rare but increased by 73% in Texas, compared to 6.4% in the rest of the US.

Source: Gemmill et al. JAMA Pediatrics. 2024

That is simply the data. There are certainly things to take issue with. Does this mean the SB8 caused these deaths?  No – correlation is not causation. Let’s not forget that COVID was happening during this period as well, which may play a significant role, though the use of others states as controls would mitigate this somewhat.

It seems reasonable, if not provable with this data, to imagine that some number of women carrying fetuses with congenital anomalies – remember, typically detected around 12 weeks - were prevented from obtaining abortions. And that some of this excess infant mortality is due to them being forced to carry those pregnancies to term.

Of course, while we all agree that infant mortality is a terrible thing, your view on how terrible this increase in infant mortality is is likely dependent on your views on abortion in general. But regardless of that opinion, it is worth recognizing that every policy decision has multiple outcomes – some that are intended and some that aren’t. As a physician, I measure the quality of a policy decision based on its ability to minimize suffering. And I certainly feel the suffering that any family that loses and infant goes through.

Regardless, I appreciate the effort to bring science to a debate that hinges on issues that are, admittedly, bigger than science. Sometimes progress on an issue can begin once there is an agreed upon set of facts. For now, then, we’ll let the facts speak for themselves.

A version of this commentary first appeared on Medscape.com.