ObjectiveSeveral indexes are used to quantify the severity of hypoxemia, including the arterial to alveolar oxygen ratio (a/APO2), the alveolar-arterial difference P(a−a)o2, the ratio Pao2/Fio2, and the oxygenation index (OI = mean airway pressure × Pao2/Fio2). This study was carried out to test how stable these indexes are when small changes in Fio2are made in ventilated neonates.DesignOpen prospective clinical study.SettingLevel III neonatal intensive care unit of a teaching hospital.PatientsForty studies were performed in 31 clinically stable ventilated neonates (median birth weight, 1450 g; median gestation, 30.6 wks), monitored by transcutaneous Pao2-oxygen saturation (Sao2).InterventionsIf hyper- or hypoxemia without derangements of Paco2or pH were detected in a blood sample taken from an indwelling arterial catheter, Fio2was changed (median change, 0.05; range, −0.3 to 0.25) and another arterial blood sample was obtained 26–83 mins (median, 42) after. The indexes were calculated in the two blood samples, and for each index the changes between baseline and the value after Fio2change were analyzed.Measurements and Main ResultsMedian baseline P(a−a)o2was 211.7 torr, median a/APO2was 0.24, median Pao2/Fio2was 161 torr, and median OI was 6.14. After the Fio2change, the coefficients of variation (sd/mean) were calculated, and they were 27.5%, 23.8%, 24.5%, and 31.6% for P(a−a)o2, a/APO2, Pao2/Fio2, and OI, respectively. Changes in the value of each index were correlated to changes in Fio2, indicating a dependency on Fio2. When data were analyzed as “high Fio2” (approximate Sao295%) vs. “low Fio2” (approximate Sao290%), differences were statistically significant for all indexes except for a/APO2.ConclusionsAll the indexes tested showed a dependency on the value of Fio2: increasing Fio2spuriously made neonates appear less hypoxemic. The a/APO2appeared to perform better than other indexes in this study, with a lower variability and a lower oxygen dependency.