The SpO2/FiO2 Ratio and ARDS: The Evidence Behind the Tool

It is two in the morning on the medical ward. A 45-year-old paddy farmer from Brahmavar, on day 5 of a febrile illness, has become breathless. His chest radiograph shows shadows in both lungs. He is on nasal prongs at 4 L/min, his saturation is 93 per cent, and his respiratory rate is 30. The blood gas machine is in the ICU, and the ICU is full. Your senior asks on the phone, “Is it ARDS?” This is an illustrative case, not a real patient.

By the end of this article you will know how to answer that without a blood gas, what the saturation ratio can and cannot tell you, and why three respected definitions can give three different answers about the same man.

1. A Ratio Without a Blood Gas

The classic measure of oxygenation in ARDS is the P/F ratio: the arterial PaO₂ divided by the fraction of inspired oxygen, the FiO₂. It needs an arterial sample. The S/F ratio replaces the PaO₂ with the pulse oximeter’s saturation, the SpO₂, so it can be worked out anywhere there is an oximeter.

“Then why not always use it?” Because it has one firm limit. The Global Definition of 2023 says so plainly: SpO₂/FiO₂ is not valid above a saturation of 97 per cent. The reason lies in the shape of the oxyhaemoglobin dissociation curve.

The oxyhaemoglobin dissociation curve flattens above a saturation of about 97 per cent. At a reading of 100 per cent the PaO₂ may be 100 mmHg or 400, and the oximeter does not distinguish them.

Think of a measuring jug that is already full. Pour in a little more or a lot more, and the level on the side reads the same. A patient on a non-rebreathing mask reading 100 per cent is that full jug. A common mistake is to divide 100 by 0.8 and write the result in the notes. Instead, reduce the oxygen until the saturation is 96 per cent or below and read the ratio then, or take a blood gas.

2. Turning One Ratio Into the Other

Rice and colleagues worked out the relation between the two ratios in patients of the ARDS Network trial: S/F = 64 + 0.84 × (P/F). On that line an S/F of 315 corresponds to a P/F of 300, and an S/F of 235 to a P/F of 200. The ratio rules out mild hypoxaemia more reliably than it rules it in: at the 315 line it has 91 per cent sensitivity but only 56 per cent specificity.

Our patient’s FiO₂ on prongs is uncertain, and we will return to that. With the common rule of thumb, 21 per cent plus four points for each litre, 4 L/min gives about 37 per cent. His S/F is then 93 divided by 0.37, about 250.

3. Three Definitions, One Patient

Is that ARDS? It depends on which definition you hold in your hand. Three are in use, and they differ in the measure of oxygenation they accept, the support the patient must be on, and the imaging that counts.

Now read our patient against all three.

Take one patient on nasal prongs, with an S/F of 250, bilateral shadows and a pneumonia. By Kigali this is ARDS. By the Global Definition it is ARDS in a resource-limited setting and not ARDS anywhere else, because prongs are not a support it admits. By Berlin it cannot be assessed at all until a blood gas is drawn. All three readings are correct on the same patient at the same moment.

“So which one is right?” All of them, each on its own terms. The answer is not to choose the one you like. It is to write down which definition you used. The tool prints all three side by side for that reason.

4. How Much Oxygen Is He Really Getting?

The rule of thumb we used is teaching shorthand, and no consensus document underwrites it. What a patient on prongs actually inspires depends on the flow, his tidal volume, his respiratory rate and whether he breathes through his mouth. A breathless man entrains more room air, so he receives less oxygen than the rule predicts, and his true S/F is higher than the one we calculated.

That is why the Global Definition sets a floor of 30 L/min for high flow and 5 cm H₂O for non-invasive ventilation: at those settings the delivered FiO₂ is known rather than estimated. On low flow oxygen, the S/F is a trend to follow, not a number to diagnose on. A Venturi device, which fixes the FiO₂, is another way to know what you are giving.

5. When the Oximeter Misleads

The European Society of Intensive Care Medicine (ESICM) guideline of 2023 considered adding high flow oxygen and the saturation ratio to the Berlin criteria, and made no recommendation to do so. One of its reasons was that SpO₂ readings are inaccurate in patients with darker skin. The figures behind that worry come from a very large American dataset.

• In more than 30,000 paired readings from the United States Veterans Health Administration, occult hypoxaemia, meaning an arterial saturation below 88 per cent while the oximeter read 92 or above, occurred in 19.6 per cent of Black patients against 15.6 per cent of white patients.

No comparable Indian dataset has been identified, and the direction of the effect across Indian skin tones is not known. So when the oximeter reads 92 and the patient looks worse than that, believe the patient and send the blood gas.

6. On High Flow: Watch the ROX Trend

At 3 a.m. our patient is moved to high flow nasal oxygen at 40 L/min with an FiO₂ of 0.6. His saturation is 94 per cent and his respiratory rate 30. His S/F is now 94 divided by 0.6, about 157. On the Rice line that is a P/F of about 110.

The ROX index is the S/F ratio divided by the respiratory rate. It answers one question on high flow: will this patient need intubation? In the validation cohort of 191 patients with pneumonia, a ROX of 4.88 or above at 2, 6 or 12 hours was associated with a lower risk of intubation, and values below 2.85 at 2 hours, below 3.47 at 6 hours and below 3.85 at 12 hours predicted failure. Two hours after starting, his numbers are unchanged, so his ROX is 157 divided by 30, about 5.2.

“Above 4.88. So he is safe?” Not yet. One value read once tells you little.

• Prediction improves with time, from an area under the curve of 0.679 at 2 hours to 0.759 at 12.

At 6 hours he needs an FiO₂ of 0.75 for a saturation of 88 per cent, and his respiratory rate is 36. His S/F is about 117 and his ROX about 3.3, below the 6-hour line of 3.47. A ROX that does not rise over the first twelve hours is the warning, and his has fallen. This is the time to call the intensivist, not the time to try one more mask.

7. The Indian Numbers

A prospective study at Bharati Vidyapeeth in Pune took 250 paired samples from 125 adults with acute hypoxaemic respiratory failure, across every mode of oxygen supplementation. It found S/F thresholds of 321 for a P/F of 300 and 252 for a P/F of 200, at 69 per cent sensitivity and 95 per cent specificity, against Rice’s 315 and 235.

• Correlation was weaker than in the ARDS Network cohort, r of 0.622 against 0.89, as expected of a mixed ward population compared with a trial cohort on protocolised ventilation.

The tool prints both sets and does not choose between them. Neither should you, in the notes: give the ratio, and say which thresholds you read it against.

8. Non-Invasive Ventilation: The National Position

The ISCCM guideline of 2020 allows a trial of non-invasive ventilation over conventional oxygen in mild, early acute hypoxaemic respiratory failure with a P/F between 200 and 300, under close supervision. Below a P/F of 150 it recommends strongly against a trial of NIV at all. Where only a saturation ratio is available, an S/F of about 190 corresponds to a P/F of 150. Our patient, at an S/F of 157 and falling, is past that line: he is not a patient for a trial of NIV on the ward.

Back to Our Patient

On the ward at 2 a.m. he had bilateral shadows and an S/F of about 250 on prongs: ARDS by Kigali, ARDS by the Global Definition only in a resource-limited setting, and not assessable by Berlin. On high flow at 40 L/min his S/F of 157 meets the Global Definition’s threshold. The tool would label any severity band for him as indicative only, because the Global Definition grades severity only in intubated patients. His falling ROX, not the label, is what sends him to the ICU.

Try This

A woman on a non-rebreathing mask at 15 L/min has a saturation of 100 per cent. The resident takes the FiO₂ as 0.8 and writes “S/F 125, ARDS”. What is wrong? Illustrative case.

Answer. Her saturation is above 97 per cent, so the S/F is not valid. Her PaO₂ could be anywhere from 100 mmHg to 400. Reduce the oxygen until her saturation is 96 per cent or below and work out the ratio then, or take a blood gas.

Take-Home Points

• Do not calculate S/F above a saturation of 97 per cent: lower the oxygen first, or take a blood gas.
• On nasal prongs the FiO₂ is a guess, so the S/F is a trend, not a diagnosis.
• Write down which ARDS definition, and which thresholds, you used.
• On high flow oxygen, follow the ROX index over hours; a ROX that does not rise is the warning.
• Below a P/F of 150, or an S/F of about 190, the ISCCM advises against a trial of NIV.

For how a Venturi device delivers a known FiO₂, see Oxygen Therapy via Venturi Devices: A Practical Guide. For oxygen in chronic lung disease, see COPD Management: The Evidence Behind the Tool.

Declaration: the author built and maintains the tool. It is free to use. The figures in this article are those the tool carried in its Evidence tab on 5 November 2026, each from the source the tool names. The illustrative case follows the tool’s rules and the rule of thumb the tool itself calls shorthand; check the tool’s result for a real patient.

References

1. Matthay MA, Arabi Y, Arroliga AC, et al. A New Global Definition of Acute Respiratory Distress Syndrome. Am J Respir Crit Care Med. 2024;209(1):37-47.
2. ARDS Definition Task Force; Ranieri VM, Rubenfeld GD, Thompson BT, et al. Acute respiratory distress syndrome: the Berlin Definition. JAMA. 2012;307(23):2526-2533.
3. Rice TW, Wheeler AP, Bernard GR, Hayden DL, Schoenfeld DA, Ware LB. Comparison of the SpO2/FIO2 ratio and the PaO2/FIO2 ratio in patients with acute lung injury or ARDS. Chest. 2007;132(2):410-417.
4. Riviello ED, Kiviri W, Twagirumugabe T, et al. Hospital incidence and outcomes of the acute respiratory distress syndrome using the Kigali modification of the Berlin definition. Am J Respir Crit Care Med. 2016;193(1):52-59.
5. Brown SM, Grissom CK, Moss M, et al. Nonlinear imputation of PaO2/FiO2 from SpO2/FiO2 among patients with acute respiratory distress syndrome. Chest. 2016;150(2):307-313.
6. Grasselli G, Calfee CS, Camporota L, et al. ESICM guidelines on acute respiratory distress syndrome: definition, phenotyping and respiratory support strategies. Intensive Care Med. 2023;49(7):727-759.
7. Alur TR, Iyer SS, Shah JN, Kulkarni S, Jedge P, Patil V. A prospective observational study comparing oxygen saturation/fraction of inspired oxygen ratio with partial pressure of oxygen in arterial blood/fraction of inspired oxygen ratio among critically ill patients requiring different modes of oxygen supplementation in intensive care unit. Indian J Crit Care Med. 2024;28(3):251-255.
8. Chawla R, Dixit SB, Zirpe KG, et al. ISCCM guidelines for the use of non-invasive ventilation in acute respiratory failure in adult ICUs. Indian J Crit Care Med. 2020;24(Suppl 1):S61-S81.
9. Roca O, Caralt B, Messika J, et al. An index combining respiratory rate and oxygenation to predict outcome of nasal high-flow therapy. Am J Respir Crit Care Med. 2019;199(11):1368-1376.
10. Guérin C, Reignier J, Richard JC, et al. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med. 2013;368(23):2159-2168.
11. Acute Respiratory Distress Syndrome Network; Brower RG, Matthay MA, Morris A, et al. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med. 2000;342(18):1301-1308.
12. Valbuena VSM, Seelye S, Sjoding MW, et al. Racial bias and reproducibility in pulse oximetry among medical and surgical inpatients in general care in the Veterans Health Administration 2013-19: multicenter, retrospective cohort study. BMJ. 2022;378:e069775.