Abstract
BACKGROUND: Noninvasive respiratory support (NRS) is widely used in pediatric ICUs (PICUs). However, there is limited experience regarding the utilization of NRS in non-PICU settings. We aimed to evaluate the success rate of NRS in pediatric high-dependency units (PHDUs), identify predictors of NRS failure, quantify adverse events, and assess outcomes. METHODS: We included infants and children (> 7 d to < 13 y old) admitted to PHDU in 2 tertiary hospitals for acute respiratory distress over a 19-month period. Collected data included diagnosis, type and duration of NRS, adverse events, and the need for PICU transfer or invasive ventilation. RESULTS: Two hundred and ninety-nine children were included, with a median age of 7 (interquartile range [IQR] 3-25) months and a median weight of 6.1 (IQR 4.3-10.5) kg. Bronchiolitis (37.5%), pneumonia (34.1%), and asthma (12.7%) were the most frequent diagnoses. Median NRS duration was 2 (IQR 1-3) d. At baseline, median S(pO(2)) was 96% (IQR 90-99); median pH was 7.36 (IQR 7.31-7.41), and median P(CO(2)) was 44 (IQR 36-53) mm Hg. Overall, 234 (78.3%) children were successfully managed in PHDU, whereas 65 (21.7%) required transfer to PICU. Thirty-eight (12.7%) needed invasive ventilation on a median time of 43.5 (IQR 13.5-108.0) h. On multivariable analysis, maximum F(IO(2)) > 0.5 (odds ratio 4.49 [95% CI 1.36-14.9], P = .01) and PEEP > 7 cm H(2)O (odds ratio 3.37 [95% CI 1.49-7.61], P = .004) were predictors for NRS failure. Significant apnea, cardiopulmonary resuscitation, and air leak syndrome were reported in 0.3, 0.7, and 0.7% children, respectively. CONCLUSIONS: In our cohort, we found NRS in PHDU safe and effective; however, maximum F(IO(2)) > 0.5 post treatment and PEEP > 7 cm H(2)O were associated with NRS failure.