Acute lower respiratory infections (ALRI) are the main cause of hospitalization during the winter season. High-flow nasal catheter (HFNC) has been established as part of the treatment of these infections.
Objective: To characterize the population of children with acute hypoxemic respiratory failure treated with HFNC and to determine the predictors of failure of this therapy.
Patients and method: Prospective study in children under 2 years of age admitted due to ALRI treated with HFNC at the Pereira Rossell Hospital Center (CHPR), from 20/5 to 20/09 2019. Data were collected from medical records and analyzed with SPSS 21.0 software. The following quantitative variables were analyzed: gestational age, birth weight, current age and weight, respiratory rate, pH, pCO2, HCO3-, BE, and pre-connection white blood cell count. The qualitative variables evaluated were: sex, comorbidities, radiological alterations, viral antigens, and history of similar episodes. Two groups were defined according to the success or failure of treatment with HFNC, where the following variables were evaluated: sex, weight, age, gasometric alterations, and white blood cell count.
Results: 449 patients were included, of which 274 (61%) were successful with the treatment. Those who failed had lower gestational age (p = 0.003), birth weight (p = 0.046), and pre-connection higher respiratory frequency (RF) (p = 0.031). In the multiple logistic regression model, pre-connection RF predicted failure for HFNC treatment (OR: 1.035 IC 95%: (1.01-1.07)). In the analysis with ROC curves, the area under the curve for RF was 0.558 (CI 0.503-0.613). The RF with the highest sensitivity (53%) and specificity (55%) was 59 breaths per minute.
Conclusion: Pre-connection RF to HFNC was the main predictor of technique failure. It is necessary to deepen other characteristics to find reliable predictors of failure.