Evidence-Based Surgery: What Evidence?
From Skeptical Scalpel
Here is a study that illustrates everything that is wrong with the current status of evidence-based surgical practice.
Many of the standard procedures we perform in general surgery are based on observational studies, expert opinion or my favorite “That’s the way I was trained,” and not randomized controlled trials. Although some such research has been done, subjecting patients to sham operations raises ethical issues and would expose patients to anesthesia unnecessarily.
But some topics could be studied prospectively. A recent paper [Variation in the use of intraoperative cholangiography during cholecystectomy. Sheffield KM et al. J Am Coll Surg. 2012;214:668-79] highlights the problem of insufficient evidence in some areas of surgery.
A group from the University of Texas Medical Branch in Galveston looked at differences in the rates of operative cholangiography in their state. They analyzed data from 212 hospitals in which more than 160 cholecystectomies had been done including almost 177,000 such operations over the 8 years from 2001-2008. The surgeon was identifiable in 89% of cases.
Despite the known pitfalls of basing clinical research on administrative data, several interesting findings of this paper are worth discussing.
Operative cholangiograms were done on 44.6% of the total cohort. By hospital, the operative cholangiogram rates ranged from 6.0% to 98.2%. The breakdown by surgeon was limited to the 706 who had done at least 40 cases. The range of operative cholangiogram use by individual surgeons was 0% to 100% with a median of 39%. Medians were higher for patients with gallstone pancreatitis (69%) and lower for those with acute cholecystitis (25%).
Uninsured patients were only slightly less likely to have operative cholangiography than those who were insured. A puzzling finding was that of those patients who had both ERCP and operative cholangiography, 37% had undergone the ERCP before the cholecystectomy and operative cholangiogram. Why would an operative cholangiogram be necessary after an ERCP had already been done?
The authors found that the variation in rates of operative cholangiography was attributed much more strongly to the surgeon and the hospital rather than the indication for surgery. They concluded that the likelihood that a patient would undergo operative cholangiography depended on the hospital she arrived at and who the surgeons was.
The extent of the variation in the use of operative cholangiography could hardly be greater. It is difficult to believe that there is no agreement on the indications. I don’t think this is unique to Texas either.
The literature is conflicting. One can find multiple papers to support any position. Some claim that operative cholangiography helps prevent common bile duct injuries and reveals unsuspected stones. Others say false positive operative cholangiograms result in more procedures and that most asymptomatic stones discovered by cholangiography never cause symptoms. Surgeons who routinely perform operative cholangiography say it does not waste time while those who don’t do them say it does.
False negatives can occur. I have seen patients with negative operative cholangiograms readmitted within a few weeks because of symptomatic common bile duct stones.
I firmly reside on the low end of the operative cholangiography spectrum. I never perform one unless there is a specific indication as dictated by the liver function tests or a significant question involving the anatomy in the operating room.
A large, well-designed randomized controlled trial would help settle the issue, but it will probably never be done.
Who would sponsor such a study? The companies that manufacture the equipment for cholangiography certainly would have no incentive to fund it. Maybe the best we can hope for is a consensus statement from a group of expert surgeons.
Will it ever be forthcoming?
A version of this post appeared in General Surgery News in June of 2012.