Is Postoperative Hypotension a Risk Factor for Surgical Site Infections After Colorectal Surgery?

To the Editor We read with great interest the article by Yilmaz et al.1 The large sample size in this retrospective study identified that postoperative time-weighted average mean arterial pressure (MAP) was not associated with surgical site infection (estimated odds ratio [95% CI] of 1.03 [0.99–1.08] per 5 mm Hg decrease [P = .16]). However, a significant inverse association between minimum postoperative MAP and surgical site infection was found (estimated ratio odds ratio of 1.08 [1.03–1.12] per 5 mm Hg decrease). We have several concerns with regard to their methodology. In different time-weighted average–MAP groups, significant differences existed in many variables, such as mean age, proportion of females, body mass index, smoking or alcohol abuse history, American Society of Anesthesiologists physical status, Charlson comorbidity index, presence of hypertension, presence of hypothyroidism, presence of metastatic cancer, presence of solid tumor without metastasis, use of immunosuppressive drugs, minutes of intraoperative MAP <55 mm Hg, procedure type, use of vasopressor, blood loss, and last recorded perioperative core temperature. All these differences indicated that there was an unbalanced background for comparisons between groups. Propensity score matching analysis, widely used in retrospective studies, would have provided better balance for all potential risk factors.2 We highly recommend the adoption of propensity score matching analysis before exploring the association between time-weighted average–MAP and surgical site infection, to make this retrospective analysis more statistically convictive. Although the authors compared many risk factors preoperatively and intraoperatively, several important variables were not investigated, such as use of antibiotic perioperatively and baseline blood pressure and serum hemoglobin. These variables may be highly associated with infection or hypotension during perioperative period. Although this investigation analyzed the association between surgical site infection with continuous blood pressure in per 5 mm Hg decrease, the authors did not group the patients according to time-weighted average–MAP intervals of 5 mm Hg. For example, in 95 ≤ MAP < 124 group, the interval of time-weighted average–MAP arrives nearly 30 mm Hg. We recommend analysis of the association between surgical site infection with blood pressure decreases of ≥10 mm Hg because more severe hypotension is more likely to be associated with postoperative adverse outcomes.3 In summary, we suggested that propensity score matching analysis be used in further studies to confirm the consequences of this clinical study. Meanwhile, some potentially important variables on patients’ preoperative or intraoperative characteristics also need to be included. Clarification regarding the above-mentioned omissions would greatly solidify the conclusions of the study by Yilmaz et al.1 Yan-chen Liu, MDXiao-Yan Meng, MDDepartment of AnesthesiologyEastern Hepatobiliary Surgery HospitalThe Second Military Medical UniversityShanghai, China[email protected] Shuai Xu, MDDepartment of Clinical MedicineThe Second Military Medical UniversityShanghai, China