Association between perioperative hypothermia and surgical site infection: A meta-analysis

BACKGROUND: A majority of reports in the past decade have demonstrated that perioperative hypothermia increases susceptibility to surgical site infection (SSI). However, in recent years, an increasing number of studies did not find an association between hypothermia and the risk of SSI. These contradictory results have given rise to a conflicting issue on whether perioperative hypothermia is associated with SSI risk in surgical patients. METHODS: We examined the association between perioperative hypothermia and SSI incidence and then integrated available evidence by searching the databases, such as PubMed, Web of Science, Embase, and Cochrane library for potential papers from inception to April 2018. We included studies that reported original data or odds ratio (OR) with 95% confidence intervals (CIs) of the associations. Using fixed-effects models combined the OR with 95% CIs, randomized controlled trials and observational studies were analyzed, respectively, and cohort studies were further analyzed. Sensitivity analyses were performed by omitting each study iteratively, and publication bias was detected using Begg's tests. RESULTS: We screened 384 studies, and identified 8 eligible studies, including 2 randomized controlled trials and 6 observational studies (1 case–control study and 5 cohort studies). The pooled OR results in the randomized controlled studies showed that perioperative hypothermia could increase the risk of SSI without heterogeneity (OR, 1.60; 95% CI, 1.14–2.23; I(2) = 0.0%, P = .845). The fixed-effect meta-analysis indicated no association between perioperative hypothermia and SSI risk in observational studies (OR, 0.98; 95% CI, 0.96–1.01; I(2) = 53.2%, P = .058). Furthermore, cohort studies were performed to pool OR by using the fixed-effect model, and the incorporated results also suggested a similar relationship (OR, 1.13; 95% CI, 0.97–1.33; I(2) = 46.4%, P = .113). CONCLUSION: The meta-analysis suggests that perioperative hypothermia is not associated with SSI in surgical patients. However, the 8 eligible studies were mostly cohort studies. Thus, further randomized controlled trials are required to confirm this finding.