The risk ratio is insufficient for clinical decisions. The case of prophylactic cholecystectomy

The medical literature contains studies that recommend both for and against prophylactic cholecystectomy when cholelithiasis is present. To investigate whether differing analytic techniques contribute to these disparate recommendations, we analyzed the decision to "wait and see" or "operate now," using both risk ratio and life expectancy analysis on the same data base. With intermediate symptom incidence estimates the risk ratio strongly favored an "operate now" approach, because it lessened the risk of death from gallbladder disease by up to 19-fold. In the group most favored by immediate surgery, 25-year-old men, risking death with immediate surgery increased life expectancy by only one percent (168 days). The risk ratio increased to over 30 for this group with the low incidence estimates, but the difference in life expectancy decreased to 120 days. With the high incidence estimates, the risk ratio declined to 16.4 and the life expectancy difference increased to 186 days. These data demonstrate that (1) the risk ratio can imply a course of action that is different from that implied by life expectancy analysis, and (2) the maximum expected return on the risk of immediate death with elective surgery for silent gallstones is only a one percent increase in life expectancy. This analysis suggests that single measures of effect such as the risk ratio and life expectancy do not capture all the information necessary to make an informed decision.