Preventing hypothermia in pediatric neurosurgery in Africa-A randomized controlled non-inferiority trial of insulation versus active warming.

Abstract

The objective of this study was to compare the efficacy of a low-cost heat-preserving method in preventing intraoperative hypothermia with that of forced-air warming in a resource-limited setting. In this randomized controlled non-inferiority trial, we recruited children younger than 12 years scheduled for cranial neurosurgery in a large East-African hospital. Patients were block-randomized by age to intraoperative warming measures using Hibler's method (intervention) or warm air (comparator). Hibler's group patients were circumferentially wrapped in transparent plastic sheeting (providing a vapor-trap) over a layer of cotton blankets, then laid on an insulating foam mattress. Warm air group patients were treated with forced-air convection via an underlying Snuggle Warm™ Pediatric Full Body mattress. Allocated warming measures were initiated in the operating theatre and discontinued upon anesthesia emergence. Perioperative temperatures were measured using noninvasive forehead probes (SpotOn™). The primary outcome was incidence of hypothermia (core temperature < 36.0° for longer than 5 min). Our null hypothesis was that Hibler's method is inferior in efficacy to the warm air method by a margin exceeding 20%. Among secondary outcomes were duration of hypothermia as proportion of surgical duration, incidence of postoperative shivering and rescue measure requirements. We analyzed data for 77 participants (Hibler's = 38; warm air = 39). There was no significant difference between the Hibler's and warm air arms of the study in the primary outcome of incidence of hypothermia (59.0% vs. 60.5% respectively; OR 1.07; 95% CI 0.43-2.65; p = .890). However, the risk difference (1.55%; 95% CI -0.20 to -0.24) exceeded the 0.2 margin and non-inferiority could not be declared. There was considerable need for rescue measures in both groups (71.1 0% vs. 69.2%; OR 1.09; 95% CI 0.41-2.90; p = .861). There was no statistically significant difference between groups for any prespecified secondary outcome. Although perioperative core temperatures were not significantly different, we could not declare an inexpensive heat-preserving method non-inferior to warm air convection in preventing intraoperative hypothermia in children undergoing anesthesia for cranial neurosurgery in a resource-limited setting. The extensive need for rescue measures may have masked important differences. US National Institutes of Health Clinicaltrials.gov database (ID no. NCT02975817).