Nocturnal oximetry in infants: Reference values and diagnostic accuracy for upper airway obstruction

To present reference data for nocturnal oximetry (NOx) in infants without respiratory disorders and evaluate their diagnostic accuracy for distinguishing moderate-to-severe upper airway obstruction (UAO) from mild/no UAO.

Infants (aged 1-12 months) without respiratory disease hospitalized for common disorders (e.g. gastroenteritis) underwent NOx (reference). Infants with and without UAO who had NOx were included as validation group. Abnormal basal SpO₂ (<10th percentile for age), oxygen desaturation (≥3 %) index (ODI3) (>90th percentile), cumulative nocturnal hypoxemia score (>90th percentile) and McGill oximetry score (MOS) (>1) were assessed as predictors of UAO requiring intervention (moderate-to-severe UAO).

Of 187 infants without respiratory disorder and with NOx, 138 (73.8 %) had acceptable tracings. Basal SpO₂ was not age-related (P = .877; 10th percentile >95 %), while ODI3 and cumulative nocturnal hypoxemia score decreased with age (P < .001). MOS>1 likelihood diminished with age (OR 0.72 [0.59-0.86]; P < .001). For reference validation, 20 infants without respiratory disorder underwent NOx, and NOx of 77 infants with UAO were analyzed retrospectively (32.5 % required intervention post-NOx). Although NOx parameters values in UAO and reference groups partially overlapped, abnormal ODI3, cumulative nocturnal hypoxemia score and MOS were associated with increased odds of moderate-to-severe UAO (OR [95 %CI]: 5.33 [1.93-14.71] for ODI3-10s (≥10s desaturation duration); 3.57 [1.27-9.99]; and 5.98 [1.82-19.69], respectively). MOS>1 outweighed ODI3-10s in sensitivity for moderate-to-severe UAO detection (80.0 % [95 %CI: 60.9 %-91.1 %) vs. 52.0 % [95 %CI: 32.4 %-71.6 %]; P = .040).

Frequent desaturations are common in infants with UAO, but also in those without respiratory disorder. Clusters of desaturations (MOS>1) represent an acceptable index of moderate-to-severe UAO.