Clinical quantification of SpO₂ instability using a new histogram classification system: a clinical study

Oxygenation instability is not quantified or documented despite being common and correlated with neonatal morbidities, retinopathy of prematurity, and adverse 18-month outcomes.

We developed a five-type SpO₂ histogram classification system based on the SpO₂ difference within the 10-90th cumulative time percentile (A) and the time percentage with SpO₂ ≤80% (B). In type 1, A is <5% and in type 5, A and B are ≥10%. We then studied consecutive 12-h SpO₂ frequency histograms in all infants ≤34 weeks gestation receiving respiratory support on day 1, over 6 months.

Six thousand and sixteen histograms were obtained in 73 infants, 28.9 ± 3.0 weeks gestation, and birth weight (BW) 1318.5 ± 495 g. All types were common and did not overlap. Type 3-5 ("unstable") histograms were more common in oxygen or any intubated support. Time in SpO₂ <85% and <80% progressively increased in types 3-5. Among histograms in oxygen, the mean (±SD) of SpO₂ medians was 92.8 ± 1.9. Infants ≤28 weeks exhibited three phases of SpO₂ instability (stable-unstable-stable). Those developing unstable histograms during the first week received longer ventilatory support (median [IQR], 101 [66] vs. 62 [28] days) and supplemental oxygen (62.5 [72] vs. 40.5 [40] days), and more were on ventilatory support at 40 weeks (7/15 vs. 0/10).

Classified SpO₂ histograms quantify and document SpO₂ instability and identify early infants at risk of prolonged respiratory support, while median SpO₂ does not.