Surgery Really Is a Numbers Game

Your first question should be: "How many have you done?"


I never get used to it, no matter how often it happens. Somebody tells me he needs surgery. I suggest asking the surgeon how many times she's done the procedure. "Really?" he says, clearly surprised. "You can ask a doctor to give you numbers?" No matter what kind of surgery or whether the setting is a hospital, an outpatient clinic, or a doctor's office, the principle is no different for surgeons than for musicians or tennis players: Practice counts. If it didn't, we'd have medical students do the surgery when the veteran cutters are out of town. They'd charge less, too.

No, thanks. Give me a surgeon who's done a lot of hernia repairs, cyst removals, heart bypasses, whatever it is I need. Unfortunately, for most procedures, nobody's gotten around to defining "a lot." (Heart bypass surgery is an exception: The number is at least 100 a year.)

A large part of the reason is that few health statisticians have gone to the trouble of figuring out the learning curve for different procedures. If they did, they could see where the curve starts to flatten out. Doing more procedures above that point wouldn't make much difference to a surgeon's success.

This made me pay close attention to a study about prostate surgery, published last week in the Journal of the National Cancer Institute. It was intriguing because it did look at surgeons' learning curve in performing a radical prostatectomy, in which the entire prostate gland is removed. And two points stood out. One was that the curve is quite steep. After five years, cancer had returned in about 18 percent of the patients whose operation was performed by surgeons who previously had done 10 or fewer during their career. The rate for patients whose surgeons had previously performed about 250 was slightly under 11 percent. That's a huge difference.

The other point was that "a lot" for radical prostate surgery is no small number. The curve didn't begin to flatten out until the surgeons had done about 250 of the operations. And these were no slouches fumbling their way through. They practiced at respected, even stellar, hospitals—Memorial Sloan-Kettering Cancer Center and Baylor College of Medicine were two. I also saw that even at 250, the curve wasn't flat. Survival rates continued to improve even after 500 procedures. The real-world suggestion to physicians sending a man for surgery, the authors concluded, was "advising patients to choose high-volume surgeons." You see? You can (and should) ask for numbers.