Oxygen Saturation Index (OSI)

OSI = (mean airway pressure × FiO₂ × 100) ÷ SpO₂, with MAP in cm H₂O, FiO₂ as a FRACTION (0.21–1.0), and SpO₂ as a percent. FiO₂ must be the fraction here: the ×100 factor is exactly what converts it to the percentage form PALICC-2 prints, MAP × FiO₂(%) ÷ SpO₂, so the two renderings give the identical number. Applying both conventions, or neither, is a 100-fold error in opposite directions, and a worked example pins the magnitude in the test suite. OSI is interpretable only for SpO₂ 80–97% — values outside that range are rejected as out-of-range and never scored. The raw index is banded against the PALICC-2 (2023) invasive-ventilation cutoffs: OSI < 5 is below the oxygenation criterion, 5 ≤ OSI < 12 is mild–moderate, and OSI ≥ 12 is severe. Bands are matched on the unrounded value even though the index is displayed to one decimal place.

OSI is the mean airway pressure times FiO₂ times 100, divided by the pulse-oximeter oxygen saturation (SpO₂) — higher OSI means a worse oxygenation defect. It substitutes SpO₂ for the arterial PaO₂ used by the Oxygenation Index (OI), sparing an arterial draw, and is defined only on positive-pressure ventilation, where a mean airway pressure exists (conventional IMV or HFOV). SpO₂ is accepted only in the 80–97% window, and the two ends of that window are sourced differently, which is worth knowing before a reading is refused. The CEILING is cited: above ~97% the oxyhemoglobin dissociation curve plateaus and SpO₂ no longer tracks PaO₂, so OSI cannot discriminate severity (Thomas 2010; PALICC-2). The FLOOR is a documented implementation choice: no lower bound specific to OSI has ever been published, and 80% is adopted because it is the lower end of the SpO₂ 80–97% window the pediatric SpO₂-based indices were derived and validated in (Khemani 2009; Khemani 2012) — the same window this platform's S/F ratio enforces. A saturation below 80% is therefore not implausible, it is simply outside the evidence; the score declines to grade it rather than extrapolate. Interpretation bands here are the PALICC-2 (2023) two-tier scheme for invasively ventilated children: the oxygenation criterion is OSI ≥ 5 and severe is OSI ≥ 12. PALICC 2015 used a three-tier scheme — mild (5 ≤ OSI < 7.5), moderate (7.5 ≤ OSI < 12.3), severe (OSI ≥ 12.3) — and PALICC-2 collapsed the two lower tiers into mild–moderate and moved the OSI severe cutoff from 12.3 (2015) to 12 (2023). This implementation applies the PALICC-2 (2023) edition, so a reader who arrives expecting a severe cutoff of 12.3 has met a tertiary source that conflated the two editions; 12 is current. The ×100 factor exactly compensates for FiO₂ being a fraction; the PALICC table rendering MAP × FiO₂% / SpO₂ gives the same number — applying both or neither is a 100× error. The map_awp (5–50 cmH₂O) numeric limits are engineering input-validation bounds, not values from a specific publication [NEEDS SOURCE]. OSI classifies a physiologic defect; it is not an individual-patient outcome prediction.