697 - Is inspired oxygen in neonates on bubble CPAP truly reliable? Evidence from an observational study in Hanoi, Vietnam

Publication Number: 4681.697

Hien Thi Thu. Vu, Karolinska Institutet, Stockholm, Stockholms Lan, Sweden; Mårten Larsson, Sahlgrenska Academy, Gothenburg University, Göteborg, Vastra Gotaland, Sweden; Vy Tran, University of Medicine and Pharmacy, Vietnam National University, Ha Noi, Ha Noi, Vietnam; Ngoc Anh Tran, University of Medicine and Pharmacy, Vietnam, Ha Noi, Ha Noi, Vietnam; Huong Lien. Nguyen, Ha Noi Obstetrics and Gynecology Hospital, Ha Noi, Ha Noi, Vietnam; Linh Thi. Nguyen, Ha Noi Obstetric and Gynecology, Ha Noi, Ha Noi, Vietnam; Tobias Alfvén, Karolinska Institutet, Stockholm, Stockholms Lan, Sweden; Nicolas J. Pejovic, Karorlinska Institutet, Stockholm, Stockholms Lan, Sweden; Susanna Myrnerts Höök, Sachs' Children and Youth Hospital, Stockholm, Sweden Department of Global Public Health, Karolinska Institutet,Stockholm, Sweden Sachs' Children and Youth Hospital, Stockholm, Sweden, Stockholm, Stockholms Lan, Sweden; Thomas Drevhammar, Karolinska Institutet, Östersund, Jamtlands Lan, Sweden

Background: Bubble Continuous Positive Airway Pressure (bCPAP) systems with separate flow meters for oxygen and air are commonly used in low- and middle-income (LMIC) settings. The fraction of inspired oxygen (FiO2) is estimated by reading flow meters and using a conversion table. Precis estimation depends on flowmeter function and accuracy. Assessment of FiO2 influence clinical decisions, such as surfactant therapy or need for mechanical ventilation. We don't know from previous studies how FiO₂ estimated by clinicians from rotameters compares to the actual delivered values.

Objective: The aim of this study was to investigate the accuracy of FiO2 as flow-based estimation by clinicians compared to FiO2 measured using an oxygen sensor.

Design/Methods: This observational study was conducted in the neonatal department of Phu San Hanoi Hospital, Hanoi, Vietnam on 25 bubble CPAP machines and for 100 comparisons. Clinicians estimated FiO₂ by reading separate flowmeters and using a conversion table. They were blinded to FiO₂ measured with a calibrated Teledyne oxygen sensor. Eligible infants (≥28 weeks postmenstrual age, stable on bCPAP) were enrolled when investigators were available, and each infant could be measured multiple times on different days. Data were collected in REDCap and de-identified before analysis.

Results: A total of 100 FiO₂ comparisons were performed, based on 17 clinicians' estimates in 66 neonates with an average birth weight of 1150 g (±341). One comparison was excluded from analysis due to clinician reading O2 as air. There were discrepancies between estimated and measured FiO2 with clinicians overestimated and underestimated FiO2 with equal frequency. The largest discrepancy is found at an average FiO₂ of 30-50%. At high FiO₂ levels (>50%), many physicians underestimate the true value.

Conclusion(s): This study found frequent under- and overestimation of FiO2 in neonates on bubble CPAP with two rotameters. The low accuracy may influence important clinical decisions on respiratory support, such as the indication for surfactant treatment and escalation to mechanical ventilation. This is not only important for everyday clinical decisions in many LMIC settings but also for researchers to expand scientific understanding and to provide a foundation for improved clinical application as well as future researches.

Figure 1: Scatter plots of clinician-estimated vs sensor-measured FiO₂ (n=99)
Dashed line indicates concordance; over- and underestimation lie above and below. Right panel shows 30% FiO₂ surfactant threshold with quadrants indicating unnecessary or missed treatment


Figure 2: Measuring FiO₂ with a calibrated Teledyne oxygen sensor at bCPAP machine


Table 1: Demography and primary diagnosis of neonates on bCPAP at Phu San Hanoi Hospital