I spend a fair amount of time talking about how social determinants of health come into play, with respect to health care outcomes. I also talk a lot about things that don't work, and therefore constitute waste in that they don't affect outcomes positively, and cost money. But there are some aspects of health care (many, in fact) that do matter, and when research points that out, it's worth highlighting. In a recent BMJ paper, a number of researchers investigated how surgeon specialization was associated with operative mortality in the United States:
Objective: To measure the association between a surgeon’s degree of specialization in a specific procedure and patient mortality.
Design: Retrospective analysis of Medicare data.
Setting: US patients aged 66 or older enrolled in traditional fee for service Medicare.
Participants: 25 152 US surgeons who performed one of eight procedures (carotid endarterectomy, coronary artery bypass grafting, valve replacement, abdominal aortic aneurysm repair, lung resection, cystectomy, pancreatic resection, or esophagectomy) on 695 987 patients in 2008-13.
Main outcome measure: Relative risk reduction in risk adjusted and volume adjusted 30 day operative mortality between surgeons in the bottom quarter and top quarter of surgeon specialization (defined as the number of times the surgeon performed the specific procedure divided by his/her total operative volume across all procedures).
Eight procedures were studied among more than 25,000 surgeons and almost 700,000 patients. Surgeons were ranked by the number of times they performed a specific procedure as a percentage of their total procedures. This was called "specialization", and was a measure of how much surgeons devoted their activities towards a specific procedure.
The hypothesis was that surgeons who specialize more on a specific procedure might be better at it. The outcome of interest was the relative risk reduction in 30-day operative mortality between the most-specialized and least-specialized surgeons, after adjusting for risk and volume.
There were four cardiovascular procedures studied, and in all of them, a surgeon's specialization predicted operative mortality. The risk reductions ranged from 46% (valve replacement), to abdominal aortic aneurysm repair (42%), to carotid endarterectomy (28%), to coronary artery bypass grafting (15%). For two of the four cancer resection procedures, specialization was also associated with reduced operative mortality - lung resection (28%) and cystectomy (41%).
What was interesting was that the predictive power of specialization was independent of the number of times that a surgeon performed a procedure. In fact, for five of the procedures - carotid endarterectomy, valve replacement, lung resection, cystectomy, and esophagectomy - the risk reduction from specialization was greater than that from volume of procedures. In other words, it mattered more that surgeons focused their time on a specific procedure than that they did a lot of them.
It's not that volume didn't matter. It did. But specialization mattered more than volume a great deal of the time. And even when volume mattered, specialization accounted for much, if not all, of the variability of 30-day mortality outside of volume.
This is important, both at an individual level and at a societal one. Often, when selecting a surgeon to perform a specific procedure, we are concerned with how many of them they do each year. It makes sense to us that surgeons who do a lot of one type of procedure will be better at them. This study would caution us to consider further how much they specialize in that procedure. It's not just how many times they do it each year, but also how much they focus on that procedure as a percentage of their total practice.
Moreover, as we think about how we make changes in delivery systems to improve outcomes, we might consider encouraging physicians to specialize in certain types of procedures more. We certainly do that already at some level (ie not everyone does heart transplants). But we might consider doing it more, at least as something we might study in the future.
–Aaron