Background
Delivery of non‐invasive ventilation commonly occurs in the pediatric intensive care unit (PICU). With the advent of high‐flow nasal cannula (HFNC), patients with respiratory distress may be rescued on the ward without a PICU admission. We evaluated our ward HFNC algorithm to determine its safety profile and independent predictors for non‐responders, defined as requiring subsequent PICU admission.
Methods
A retrospective chart review of patients <17 years of age admitted with respiratory distress between 2016 and 2017 was carried out. Pediatric Early Warning System (PEWS) respiratory score was used to assess the clinical response of patients requiring HFNC. Variables associated with non‐responders were evaluated, and their PICU admission was studied for escalation of care and criticality.
Results
Patients with comorbidities (P = 0.02) were more likely to require HFNC. Of the 18 patients initiated on HFNC, 44% (n = 8) remained on the ward. Non‐responders (n = 10; 56%) had higher (2.7 vs 1.8; P = 0.03) and worsening (−0.1 vs 0.3; P = 0.05) PEWS respiratory scores 90 min after HFNC initiation. Eighty percent (n = 8) of non‐responders required escalation to continuous positive airway pressure or bilevel positive airway pressure in the PICU. For both HFNC responders and non‐responders, there were no requirements for intubation, evidence of air leak or difference in days of respiratory support.
Conclusions
High and worsening PEWS scores 90 min after HFNC initiation may indicate non‐response when coupled with a standardized ward HFNC algorithm for respiratory distress. Further improvements may be seen with an earlier initiation of HFNC in the emergency department and more aggressive flow escalation on the ward.