The impact of perioperative hypothermia on surgical site infection risk: a meta-analysis

The connection between perioperative hypothermia and the occurrence of surgical site infections (SSIs) is still not clearly established. This investigation aimed to clarify the potential link between these two factors using meta-analytical techniques.

This investigation examined the potential connection of perioperative hypothermia to the risk of SSI. A comprehensive literature review was conducted utilizing PubMed, Web of Science, and Embase. The primary endpoint was the incidence of SSI. Pooled risk ratios were estimated using fixed- or random-effects meta-analysis. Sensitivity analyses were performed to evaluate the impact of preoperative hypothermia on the pooled risk of SSI. Statistical analyses were performed using Review Manager software and R software.

Our meta-analysis encompassed 25 studies representing a total of 28,761 patients. The analysis revealed no statistically significant association between patients' intraoperative body temperature and their likelihood of developing post-surgical complications [odds ratio (OR), 1.39; 95% confidence interval (CI), 0.98-1.96; I² = 89%, P = 0.06]. However, two notable exceptions emerged from the subgroup analyses. Most importantly, patients undergoing breast surgery demonstrated a significantly higher risk of SSI when experiencing intraoperative hypothermia, with an odds ratio of 1.97 (95% CI: 1.21-3.21, I² = 0%, P < 0.01). Additionally, across all patient groups, a consistent trend was noted: when intraoperative body temperature fell to 35 °C or below, there was a substantial increase in SSI risk (OR: 2.12, 95% CI: 1.42-3.16, I² = 89%, P < 0.01).

The findings suggest that the relationship between perioperative hypothermia and an increased incidence of SSI is not definitive; however, it is significantly associated with breast surgery and when body temperature falls below 35 °C. Future randomized controlled trials (RCTs) should focus on breast surgery and standardize temperature thresholds. Due to the high heterogeneity, the interpretation of this result should be approached with caution, and there is a call for future high-quality research.