Familial Prevalence of Gallstones in First-Degree Relatives of Patients With Gallstones patients should always be screened for gallstone occur-To the Editor: rence.'' Why? The most likely consequence is that their asymptomatic gallstones will be removed, exposing We write in response to the recent article by Shiv et them to immediate postsurgical complications, late al., ''High Familial Prevalence of Gallstones in the postcholecystectomy diarrhea, or gastroduodenal re-First-Degree Relatives of Gallstone Patients.'' 1 Flaws flux. The authors themselves state that of the ''eleven in the study design may have easily biased the results, of . . . 51 (21.6%) relatives'' who ''had symptoms of leading to costly recommendations without adequate upper abdominal discomfort and dyspepsia . . . these evidence.
symptoms were not characteristic of gallstone disease.'' With regard to study design, a high degree of selec-We find the endorsement of expensive and possibly deltion bias was present from the outset. For example, eterious recommendations frightening, especially only subjects with gallstones who had two living firstwhen based on the results of a single study. degree relatives willing to participate in the study were To be clear, we are not dismissing the possibility that included. It is easy to conceive that subjects whose famcertain families are at higher risk for the development of ily members did not welcome a familial trend toward cholelithiasis. However, this study, because of design ercholelithiasis had negative personal interests in the rors, cannot be relied upon to provide the correct answer. project and would be less likely to enroll (or vice versa). In addition, which first-degree family members partici-YVONNE ROMERO, M.D. pated was also a nonrandom, self-selection process.
Gastroenterology Fellow Overall, the study design favors enrollment of family members with gastrointestinal symptoms or those SIDNEY F. PHILLIPS, M.D. members who were curious to their gallstone status.
Professor of Medicine The control group was selected specifically for ''not Gastroenterology Research Unit having any family history of gallstones or any systemic Mayo Clinic disease known to predispose to gallstone formation.'' Rochester, MN With the intentional exclusion of control families who REFERENCE may have had a trend toward cholelithiasis, is it really