Letter to the editor: From Min Gong et al: "Risk for infections during treatment with denosumab for osteoporosis: a systematic review and meta-analysis"

Abstract

To the editor: We read with great interest the systematic review by Diker-Cohen et al (1) about the risk of infection under the treatment of denosumab for osteoporosis. This is a well-conducted systematic review that provided us with valuable information about the safety of denosumab. We would like to point out some concerns about the analytic methods used in the systematic review. We noticed that the main outcome of this systematic review was the serious adverse events of infections (SAEIs), and the authors reported the results measured by risk ratio (RR) and risk difference (RD). The authors claimed that they used RD with the Mantel-Haenszel method to deal with studies with no events in both arms. We totally agree with this. However, for RR, they failed to use a valid method to deal with studies with no events; instead, they discarded such studies in the meta-analysis. This is problematic as such studies generally indicate no difference for treatment effects for balanced trials, and discarding them is expected to result in an overestimate of the effects (2). In their systematic review of SAEI outcome, 9 out of 34 studies had no events in both arms, and 5 of them had balanced sample size in treatment and control arms.