Preventing perioperative hypothermia in children through implementation of customised guidelines.

Introduction: Perioperative hypothermia (PH) is common in children and associated with adverse clinical outcomes. Guidelines to prevent PH are mainly developed for adults and differ among institutions. We aimed to evaluate the effectiveness of customised guidelines in reducing PH in our paediatric population and the impact of cost considerations on physician practice.

Methods: Patients aged ≤16 years undergoing general anaesthesia in our tertiary paediatric hospital were prospectively recruited in this cohort study. Patient demographics, surgical procedures, anaesthesia details and temperature control measures were recorded. Data collection occurred over four phases: Phases 1 and 2 comprised standard management, while Phases 3 and 4 occurred following guidelines implementation. Sensors for continuous core temperature monitoring were provided free to patients during Phases 1 and 3, but were charged during Phases 2 and 4. The main outcome was occurrence of PH, defined as core temperature <36°C at any point from induction of anaesthesia to discharge from the postanaesthetic care unit. The impact of guidelines implementation and cost considerations influencing physician practice on PH outcomes was also analysed.

Results: Data from 3917 patients was analysed (1766 in Phase 1, 679 in Phase 2, 706 in Phase 3 and 766 in Phase 4). Guidelines implementation decreased PH incidence from 11.0% to 6.79% (odds ratio [OR] 0.63, 95% confidence interval [CI] 0.50-0.80, P = 0.0002). Free sensors increased the odds of detecting PH (OR 1.48, 95% CI 1.17-1.88, P = 0.001). With guidelines implementation, there was greater reduction in PH with free sensors (OR 0.64, 95% CI 0.47-0.88, P = 0.0055) compared to chargeable sensors (OR 0.75, 95% CI 0.50-1.11, P = 0.1471).

Conclusions: Customised guidelines facilitated a sustained reduction of hypothermia in our paediatric surgical patients, although its impact was reduced by cost considerations.