Estimate hypoxemic respiratory failure severity using FiO2, mean airway pressure (MAP), and arterial oxygenation (PaO2), then apply a trend-based escalation workflow.
Last reviewed: February 26, 2026
Use FiO₂ as a fraction (0.21–1.0), not %.
Mean airway pressure from the ventilator or device.
Use an arterial blood gas value; we convert kPa to mmHg when selected.
Oxygenation Index (OI) describes the severity of hypoxemic respiratory failure by relating the level of respiratory support to the achieved arterial oxygenation. It is defined as:
OI = (FiO2 x MAP x 100) / PaO2(mmHg). If PaO2 is entered in kPa, we convert to mmHg by multiplying by 7.50062 before computing the index.
Before trusting the number, verify that FiO2, MAP, and PaO2 reflect the same clinical window and not mixed pre-change and post-change values.
OI is used in NICU and PICU workflows to track whether oxygen delivery is improving, plateauing, or deteriorating as support is adjusted. Rising OI despite higher support suggests worsening oxygenation efficiency and should trigger a structured bedside reassessment.
OI has important limitations. It does not directly represent lung compliance, dead space, CO2 clearance, or cardiac contribution to oxygen delivery. Sedation changes, recruitment maneuvers, ventilator mode shifts, and sampling timing can alter OI without representing a true baseline change.
Use OI trends with a parallel review of ventilator prescription, imaging, perfusion/hemodynamics, and bedside exam. Document the number and the change decision made from it so handoffs remain reproducible.
OI should be interpreted as a trend with the child's exam, ventilator trajectory, and response to interventions, not as a single detached threshold.
Institution-specific thresholds and timing vary. Follow your PICU/NICU pathway and involve respiratory therapy and ECMO-capable teams early when OI remains high despite optimization.
OI is most useful in intubated, mechanically ventilated infants and children with significant hypoxemic respiratory failure, where FiO2, MAP, and PaO2 are measured at the same bedside timepoint and trended serially.
No. OI is defined with arterial PaO2. If you only have pulse oximetry, use oxygen saturation index workflows separately; do not mix SpO2 into the OI formula.
Many centers consider sustained OI ≥ 40 despite maximal conventional support as a threshold to evaluate for ECMO, but exact triggers vary by institution and patient context; always follow your local ECMO guidelines.
Recalculate after meaningful ventilator or oxygen changes and with each new arterial blood gas in unstable periods. A reproducible trend over hours is more actionable than one isolated value.
Yes. OI increases when MAP rises, so higher airway pressures can worsen OI even if PaO2 has not fallen. Interpret OI with ventilator strategy, hemodynamics, and exam rather than as a standalone trigger.
Only with caution. Mode changes can alter MAP and oxygenation dynamics, so compare serial OI values alongside mode, settings, and timing to avoid false trend interpretation.
No. ECMO decisions are based on sustained severity, reversibility, trajectory, comorbidities, and center-specific criteria. OI is one major input, not a stand-alone decision.
Document OI with timestamp, synchronized ventilator settings, trend direction, and action taken. This prevents handoff confusion from mixed-timepoint values.