Prospective interventional study with crossover design.Setting
Urban pediatric emergency department.Participants
32 well-hydrated children aged 1 month to 12 years brought to the emergency department for care of minor illness or injury.Interventions
Participants were assigned in random order to a 15-minute waiting period in each of two rooms, with and without air-conditioning (cool and warm rooms, respectively). At the end of each waiting period, fingertip CR was measured with a stopwatch, three times by each of one or more three trained observers.Results
Mean CR time was 0.85 +- 0.45 seconds in the warm room (mean ambient temperature 25.7 degrees C) vs 2.39 +- 0.76 seconds in the cool room (mean temperature 19.4 degrees C). The mean overall difference in CR time between the two environments was 1.53 seconds (95% confidence interval (CI): 1.31, 1.75; P < .001); the difference was significant regardless of age or sequence of exposure. 100% of patients were considered to have normal CR (less than 2 seconds) in the warm room, whereas only 31% were considered normal in the cool room. In the 16 patients with CR measured by three observers, interobserver reliability was fair, with an intraclass correlation coefficient of 0.70 (95% CI: 0.56, 0.85), and kappa of 0.54 (95% CI: 0.33, 0.73).Conclusions
Decreases in ambient temperature within a range found in typical office/emergency department settings may cause significant prolongation of CR time in children with normal circulatory status. There is marked interobserver variability in the measurement of CR even when performed by experienced observers. These findings suggest limitations to the use of CR in the assessment of ill or injured children. Pediatrics 1993;92:699-702; capillary refill, ambient temperature, blood circulation.