Methods:
This was a retrospective study conducted in two level III NICUs in Winnipeg, Manitoba, Canada from April 2015 to July 2020. We included infants with HRF who underwent a CAR assessment with integrated monitoring by SpO2 and crRTO by NIRS using the following scheme: 1) a baseline assessment of mean SpO2, and FiO2 for the previous 24 hours before CAR assessment, together with baseline hemoglobin (Hb) and blood gases; 2) assessment of hemodynamics with a predefined targeted neonatal echocardiography protocol, excluding infants with pulmonary hypertension, significant PDA and circulatory shock; 3) a CAR assessment test, which is stepwise reduction of FiO2 by 0.02 every 2 minutes with continuous monitoring of SpO2 by pulse oximetry, and crRTO by NIRS. The infant should be quiet without significant handling for at least 30 minutes before the test, SpO2 should be between 90 to 95 % before starting, and must be maintained within the acceptable range as per unit protocol (86% to 94%) during the test, or the test was discontinued. Following the CAR test, routine saturation monitoring including SpO2 and NIRS was continued with titration of FiO2 with maintaining SpO2 as above and crRTO between 60 to 80% for 72 hours or until weaning FiO2 to <0.3. Institutional ethics board approval was obtained before commencing this study, figure 1 is an algorithm clarifies the steps of ORT and assessment of CAR.