A case-control study of cholecystectomy and right-side colon cancer: the influence of alternative data sources and differential interview participation proportions on odds ratio estimates.

One hundred fifty patients with right-side colon cancer (i.e., patients with adenocarcinoma of the cecum or ascending colon) were compared to 150 matched left-side colon cancer controls (i.e., patients with adenocarcinoma of the descending or sigmoid colon) and to 123 neighborhood controls, Pittsburgh, Pennsylvania, Standard Metropolitan Statistical Area, 1975-1978. The gastrointestinal surgical history was ascertained for all study subjects so that the presence or absence of a history of cholecystectomy could be noted. Cholecystectomy history was obtained through telephone interviews and whenever possible subsequently validated from operative and pathology reports at time of cholecystectomy. Cholecystectomy history for the colon cancer patients was also abstracted from hospital records at time of colon cancer diagnosis with an attempt to confirm the gallbladder's status through operative reports, cholecystograms, and physical examinations. Hospital records and interviews for the colon cancer patients appeared to provide accurate exposure history. Point estimates of the odds ratios and confidence intervals for intra- and inter-data source comparisons (i.e., hospital records, interviews, and hospital records and interviews combined) were comparable with similar measures of effect. Consistent odds ratio estimates appeared in both left-side colon cancer controls (1.9) and neighborhood controls (1.89). The authors suggest that changes in bile acid metabolism following cholecystectomy may be associated an increased risk of right-side colon cancer.