To the Editor We read the study by Li et al1 with great interest and have been inspired to consider new prophylactic strategies against Clostridium difficile infection because the surgical patients in our departments are similar to those in their study. However, we have 2 questions for the authors.
First, the authors mentioned that missing data on preoperative demographics and clinical characteristics in the Veterans Affairs Surgical Quality Improvement Program were rare, but item 12(c) in the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement recommends that one should “[e]xplain how missing data were addressed.”2(p346) Precisely how much data were missing? We would like to know whether these missing data had the prior condition for multiple imputation3 and why they were missing.
Wada Y, Tsunemori H, Kikuchi M. Question About the Analysis of Missing Data and Identification of an Error. JAMA Surg. 2016;151(8):781-782. doi:10.1001/jamasurg.2016.0127