Effect of surgeon specialty interest on patient outcome after potentially curative colorectal cancer surgery

OBJECTIVE: There are significant differences in patient outcome after poten tially curative surgery for colorectal cancer that relate to the surgeon pe rforming the procedure. The reasons for these differences remain obscure. T he aim of this study was to examine the effect of the surgeon's specialty o n patient outcome after potentially curative colorectal cancer surgery and to identify factors that may help explain differences in outcome among spec ialty groups. METHODS: Between 1990 and 1993, 378 patients underwent potent ially curative surgery for colorectal cancer by surgeons with different spe cialty interests, vascular or transplant, general, and colorectal surgeons, in a large teaching hospital. Information on operative details, including the length of the resection specimen, resection margins, whether the tumor was removed with en bloc resection of adjacent clinically involved organs, number of lymph nodes removed, and stage was collected. Factors affecting b oth local and overall recurrence rates were analyzed using logistic regress ion analysis at both univariate and multivariate levels. RESULTS: At a medi an follow-up of 45 months the only factors associated with a significantly reduced local recurrence rate were the length of the resection specimen (od ds ratio, 0.56; 95 percent confidence interval, 0.31-0.99) and colorectal s pecialty (P = 0.04). Patients operated on by a general surgeon were 3.42 ti mes (95 percent confidence interval, 1.32-8.9) more likely to develop a loc al recurrence than those operated on by a colorectal surgeon. For overall r ecurrence, early stage disease (P < 0.0001), absence of vascular invasion ( 0.005), and colorectal specialty (0.025) were the only factors associated w ith significantly improved outcome at multivariate analysis. CONCLUSIONS: T hese data show that surgeons with an interest in colorectal cancer achieve lower local and overall recurrence rates com pared with vascular or transpl ant or general surgeons. Differences in local recurrence rates seem to be p redominantly related to the extent of resection performed and demonstrate t he need to remove an adequate specimen when Performing potentially curative colorectal cancer surgery.