An infant may remain hypoxemic after FiO₂ is raised to 1.0, which can raise concern for cyanotic congenital heart disease (CHD). But an oxygen response alone neither diagnoses the cause nor shows how much respiratory support was needed. The oxygenation index (OI) addresses the second question: it puts PaO₂ in context with FiO₂ and mean airway pressure. OI tracks severity, not cause.
One PaO₂, three different questions
| Tool | Question it helps answer | What it cannot establish alone |
|---|---|---|
| Hyperoxia challenge | How does PaO₂ respond when FiO₂ is raised to 1.0? | Whether CHD, persistent pulmonary hypertension of the newborn (PPHN), or severe lung disease is the cause |
| Oxygenation index | How much oxygen and ventilator pressure are being used relative to the PaO₂ achieved? | The underlying diagnosis or systemic oxygen delivery |
| Echocardiography | What are the cardiac anatomy, shunt direction, estimated pressures, and ventricular function? | Every aspect of lung disease or respiratory support needs |
These tools answer different questions. A poor oxygen response does not replace evaluation of the underlying physiology, while OI quantifies support relative to PaO₂.
Build the oxygenation index correctly
For conventionally expressed OI, use:
OI = (FiO₂ × mean airway pressure × 100) ÷ PaO₂
Enter FiO₂ as a fraction—for example, 0.80, not 80. Use the ventilator’s mean airway pressure in cm H₂O and arterial PaO₂ in mmHg. Here, MAP means mean airway pressure, not mean arterial pressure.
Consider two infants—or two moments in the same infant—with a PaO₂ of 50 mmHg:
| FiO₂ | Mean airway pressure | PaO₂ | OI |
|---|---|---|---|
| 0.50 | 8 cm H₂O | 50 mmHg | 8 |
| 0.90 | 16 cm H₂O | 50 mmHg | 28.8 |
The PaO₂ is identical, but the second value was achieved with substantially more oxygen and pressure support. That difference is clinically important. An isolated PaO₂ does not reveal the effort required to maintain it; OI combines those support requirements with the result.
Read the trend beside the ventilator settings
OI is most useful as a trend. If PaO₂ falls while FiO₂ or mean airway pressure rises, oxygenation is worsening despite escalating support. If the PaO₂ improves while FiO₂ and pressure needs fall, the trajectory is more reassuring than a PaO₂ rise achieved only by increasing support.
But OI is not a pure measure of disease. It changes when clinicians change ventilator settings, so interpret each value alongside the settings and the timing of the arterial sample. Be cautious when comparing values across a major ventilator-mode change or when the sample and recorded settings were not obtained close together.
Sampling site matters in a newborn with ductal shunting. A right radial arterial sample is preductal; an umbilical arterial catheter (UAC) sample is generally postductal. Label the site and use a consistent approach for serial comparisons. If preductal and postductal oxygenation differ, that difference is a separate physiologic clue—not a reason to treat the two PaO₂ values as interchangeable.
OI requires an arterial PaO₂. Do not substitute pulse-oximeter SpO₂ into the OI formula. The oxygen saturation index (OSI) is a separate measure that uses SpO₂; it may be useful when arterial sampling is unavailable, but it is not simply another way to calculate OI, and its role in guiding treatment is less established.
Use OI to guide attention—not to name the diagnosis
A high or rising OI can help a neonatal team communicate the severity and trajectory of hypoxemic respiratory failure, seek specialist input, and consider escalation within the local protocol. It does not distinguish PPHN from parenchymal lung disease or CHD. Different causes can produce similar oxygenation impairment, and the right next step depends on the infant’s whole physiology.
For example, in a late-preterm or term infant with hypoxemic respiratory failure despite optimized oxygenation and ventilation, OI may help inform consideration of inhaled nitric oxide within local guidance. It does not diagnose PPHN or automatically trigger treatment. Gestational age, lung recruitment, hemodynamics, cardiac function, and the suspected underlying lesion all matter. When feasible, echocardiography should assess the cardiac physiology, ventricular function, and whether a ductal-dependent systemic lesion is present before inhaled nitric oxide is started. If clinical instability requires treatment before imaging, obtain echocardiography as soon as feasible.
Keep the roles separate: PaO₂ describes the arterial oxygen tension; OI relates that result to respiratory support; echocardiography helps define the cardiac and pulmonary vascular physiology. Confusing those questions can turn a useful measurement into a false diagnosis.
Practical takeaways
- A hyperoxia response raises a diagnostic question; OI answers a separate severity question by relating oxygenation to respiratory support.
- OI combines FiO₂, mean airway pressure, and arterial PaO₂ to put oxygenation in support context.
- Interpret OI serially with contemporaneous ventilator settings and a clearly identified arterial sampling site.
- OI helps describe severity and may inform treatment decisions; it does not diagnose PPHN or CHD, and it is not a stand-alone treatment trigger.
