Pulse Oximetry, Skin Pigmentation, and Occult Hypoxemia in Pediatric Cardiac Critical Care

Pulse oximetry (SpO₂) is central to oxygenation assessment in pediatric and cardiac intensive care. Increasing evidence demonstrates that SpO₂ may systematically overestimate arterial oxygen saturation (SaO₂), particularly in individuals with darker skin pigmentation, thereby increasing the risk of occult hypoxemia-arterial hypoxemia not detected by pulse oximetry. We sought to synthesize current evidence and define implications for pediatric and congenital heart disease populations. We performed a narrative review of experimental physiology studies; observational adult and pediatric cohorts; pediatric COVID-19, ICU, and cardiac ICU investigations; device-comparison analyses; systematic reviews; and regulatory evaluations reporting paired SpO₂-SaO₂ measurements or validated reference standards stratified by race, ethnicity, or objectively measured skin pigmentation. Data sources included PubMed, MEDLINE, and bibliographies of relevant studies. Experimental data demonstrate directional SpO₂ overestimation that increases during hypoxemia. Large adult cohorts confirm higher rates of occult hypoxemia in patients with darker skin pigmentation, findings subsequently replicated in hospitalized children. Emerging pediatric cardiac ICU data suggest that physiologic complexity may further amplify SpO₂-SaO₂ discordance. Device-comparison studies reveal variability across manufacturers, influenced by calibration datasets and signal-processing algorithms. Across populations, misclassification is most clinically relevant near commonly used escalation thresholds. SpO₂ frequently overestimates SaO₂ in darker skin pigmentation and during hypoxemia, increasing vulnerability to occult hypoxemia in hospitalized and critically ill children. In pediatric cardiac populations-where narrow saturation thresholds guide management-contextual, bias-aware interpretation of pulse oximetry is essential to support accurate decision-making and equitable care.