Bapu Jena is an associate professor of health care policy at Harvard Medical School, a physician in the Department of Medicine at Massachusetts General Hospital and a member of the 2021 Tradeoffs Research Council. His research interests include the economics of physician behavior and the physician workforce. 

Physicians regularly have to make quick decisions in high-pressure situations. Ideally, these decisions are based on years of experience and evidence, but a number of studies have shown that physicians often rely on simple mental shortcuts known as heuristics to make these life-or-death calls. A new study looks at whether these mental shortcuts influence one of the most important medical events — childbirth.

In a paper published in the journal Science, University of Massachusetts Amherst Professor Manasvini Singh used electronic health record data on more than 80,000 births to see whether one of the most critical choices obstetricians make when choosing how to deliver a baby — cesarean section or vaginal birth — is susceptible to what’s known as the “win-stay, lose-switch” heuristic. 

She found obstetricians were more likely to switch delivery modes after having a complication in the previous delivery. In other words, if there were complications with a vaginal birth, obstetricians were less likely to attempt a vaginal birth in the next patient (and vice-versa for c-sections), even though switching may not have been appropriate clinically. (The effect was only statistically significant for the birth immediately following the complication.)

There are some important limitations to this study. It only uses data from two academic hospitals in one state, and it cannot definitively tell us whether the complication was the reason the obstetrician switched delivery modes, since it can’t tell us what the original birth plan was. But Singh’s rigorous analysis is pretty convincing. 

The results are interesting to me for two reasons. First, they suggest that in a high-stakes environment like childbirth where we might think (or hope) that mental shortcuts are used less, they still occur. This could have troubling implications if these heuristics reinforce physician biases and/or lead to suboptimal decisions. Singh suggests algorithms could help guide physician decision making, but algorithms can also reinforce biases.

Second, I think this study (along with others like it) illustrate a way to use behavioral economics to answer other questions in health care. Heuristics like the one in this paper could create “natural experiments” for comparing the effectiveness of different forms of care. In this case, the heuristic could be seen as effectively randomizing which patients got a c-section or vaginal delivery, which could allow researchers to compare which delivery mode works best in different scenarios. While there is good evidence that c-sections are overused, there are many other areas of medicine — like oncology, surgery and nearly every medical specialty — where we often lack high quality data on how various treatments or services compare to one another.