You go into Starbucks, and order your preferred brew. What you get is what you expect: a tall, dark roast. The next day, you go into a different Starbucks location and place the same order. Though you gave your usual instructions and the only thing that changed was the address of the coffee shop, you get something wildly different.
Now imagine if, instead of your morning coffee, we were talking about whether you were going to get surgery or not: If where you sought care determined what kind of care you received, such as the difference between having your abdomen cut open for a caesarean-section birth or a less risky (and less costly) vaginal delivery.
“You would think if two women who looked pretty similar walked into two hospitals to deliver,” said Dr. Katy Backes Kozhimannil (PhD), a public health professor at the University of Minnesota, “they would have a similar chance of a cesarean delivery.” But in a new study, Dr. Kozhimannil found quite the opposite. Looking at 800,000 deliveries in 593 U.S. hospitals, she and her co-investigators discovered that seven per cent of mothers got a C-section in some hospitals, while at others, 70 per cent did. (The World Health Organization says 10 to 15 per cent is the appropriate rate of cesarean delivery.)
To make sure the variation wasn’t driven by hospital profile—different patient populations, hospitals that specialize in complex cases versus those that don’t—they also crunched a sub-set of the data: women with low-risk births. To their surprise, here too they found a massive range, from two per cent to 36 per cent of deliveries.
This finding was disturbing, Dr. Kozhimannil said, “because it signals an overuse or underuse of services, both of which have an impact on health and cost outcomes.” In the context of delivery, she added, the conversation immediately turns to overuse but what’s sometimes forgotten is that there may also be women who needed C-sections and didn’t get them.
In health care, the term “unwarranted variation” describes these differences, when the care patients receive is determined by things like a doctor’s preferences or policies within a local hospital instead of the best-available evidence or patient need. In this case, the researchers thought “vast differences in practice patterns”—not patient characteristics—were driving the fluctuation in C-sections. After all, they found requests by mothers accounted for only a very tiny percentage of cesarean deliveries. (The same is true in Canada.)
The thing that’s science-ish about all of this is that it’s not a new health-care problem. With C-sections, unwarranted variation has been documented for decades at the country level (Finland has a lower rate compared to Australia and the U.S.) and even within provinces and states. It’s also a problem that goes beyond child delivery. In Canada, this medical lottery has been seen in everything from mastectomies and hysterectomies, to knee-replacement and coronary-bypass surgeries.
An intractable problem?
And yet, even though we have reams of data about unwarranted variation, the issue persists, and is the subject of very little public discourse and political attention. For Dr. Kozhimannil, part of the challenge is that addressing variations in health care involves confronting vested interests and the health system status quo. “We’re on a particular path and getting off that path is very difficult,” she said.
Her co-author, UBC’s Dr. Michael Law (PhD), pointed out that the health-care community’s hands-off approach hasn’t helped. “We assume by putting out practice guidelines and publishing on appropriateness, that’s going to promote change,” he said, explaining that changing physicians’ practice patterns requires intensive training and education.
Beyond transforming medical practice, there’s another obstacle: the ‘more is better’ narrative still dominates in medicine, and unwarranted variation upends that. Sometimes more is worse, and just more expensive.
Still, there are solutions. As Drs. Law and Kozhimannil outline in their paper on C-sections, getting policymakers to introduce statewide quality-improvement programs or incentives that promote hospital-level change can help. But first, we need strong voices to raise awareness about unwarranted variation and the political will to address it. We wouldn’t tolerate inconsistency within a coffee chain. Why would we tolerate such variability when it comes to our bodies and our health?
Science-ish is a joint project of Maclean’s, the Medical Post and the McMaster Health Forum. Julia Belluz is the senior editor at the Medical Post. Got a tip? Message her at firstname.lastname@example.org or @juliaoftoronto on Twitter.