To the Editor:

One cohort study in South Korea conducted by Choi et al.¹ revealed that the cancer risk was increased in people who underwent cholecystectomy than the aged- and sex-matched control group (adjusted hazard ratio=1.19). Some points not mentioned in the original article are discussed.

First, gallstone disease is the leading cause of cholecystectomy.² Thus, we assume the cholecystectomized group found in the study of Choi et al.¹ should be those people with gallstone initially. The control group consisted of non-cholecystectomized people including those with gallstone and those without gallstone. The literature revealed that gallstone itself, not cholecystectomy, correlates with an increased risk of cancer.³ The readers hope to know the sub-analysis investigating: (1) the cancer risk between cholecystectomized people and non-cholecystectomized people having gallstone; (2) the cancer risk between cholecystectomized people and non-cholecystectomized people who did not have gallstone; and (3) the cancer risk between non-cholecystectomized people having gallstone and non-cholecystectomized people who did not have gallstone (Fig. 1). Then the role of cholecystectomy and gallstone on the risk of cancer might be clarified. Choi et al.¹ commented that the operational definition of gallstone based on the claimed data was not accurate, so they did not analyze the relationship between carrying gallstone and cancer. We think that given the good quality of the Korean National Health Insurance Service database, the diagnosis of gallstone based on ICD-10 codes can be trusted. Choi et al.¹ have the ability to conduct the sub-analysis to clarify the puzzle. In the absence of such a sub-analysis, it is difficult to persuade the readers that cholecystectomy itself would increase the cancer risk. The readers would believe that the increased cancer risk related to cholecystectomy is distorted by selection of an inappropriate control group. Second, surveillance bias should be considered. Persons of interest could be more closely monitored than the general population.⁴ Therefore, disease of interest may be early found in these monitored persons. Choi et al.¹ also commented that cholecystectomized people might receive more health screening than non-cholecystectomized people. Thus cholecystectomized people had more chances to discover cancer. However, surveillance bias is difficult to avoid in observational studies clinically.⁵ Finally, we agree with Choi et al.¹’s conclusion that further studies are warranted to clarify this issue.

Figure 1. The sub-analysis investigates: (1) the cancer risk between cholecystectomized people and non-cholecystectomized people having gallstone, (2) the cancer risk between cholecystectomized people and non-cholecystectomized people who did not have gallstone, and (3) the cancer risk between non-cholecystectomized people having gallstone and non-cholecystectomized people who did not have gallstone.

No potential conflict of interest relevant to this article was reported.