Time Below 90% Saturation Calculator (T90)
Time Below 90% Saturation Calculator (T90)
Convert the T90 percentage a report prints into minutes, and read it against the published MrOS cut-off — the measure the AASM promoted to recommended reporting in version 3 as it demoted the desaturation index.
T90 in minutes, from the percentage
Percentage to minutes, with the MrOS cut-offT90 7.4% of a 357-minute night, nadir 81%
Formula
- why the conversion is needed
- reports almost always print T90 as a percentage; the published cut-off is in minutes. A T90 of 7.4% means 26.4 minutes over a 357-minute night and 35.5 minutes over a 480-minute one, so the percentage alone cannot be compared with the cut-off
- the 12-minute cut-off
- Baumert and colleagues, 2840 men of the MrOS cohort, mean follow-up 8.8 years, 185 cardiovascular deaths: men with T90 above 12 minutes were at elevated risk with a hazard ratio of 1.59 (p = 0.006). Strictly greater than 12, so exactly 12.0 minutes is on the lower side
- what T90 is made of
- the same paper split it into the part from acute desaturations and the part from non-specific drift. Twenty minutes in sixty short respiratory desaturations and twenty minutes continuously at 89% are the same T90 and not the same physiology
- the AASM’s version 3 change
- its own public Summary of Updates records that reporting the time below a specified oxygen saturation threshold became a RECOMMENDED rule in version 3, in the same revision that changed reporting of the oxygen desaturation index from recommended to OPTIONAL. Note ‘a specified threshold’ — the manual does not fix it at 90%, so T88 and T80 appear on reports too and are not interchangeable with this
- the 88% figure is a different thing entirely
- Medicare’s home oxygen documents use a saturation at or below 88% during sleep as a coverage criterion. That is a threshold on the SATURATION, not on the time, and it is quoted here because a reader will meet it rather than as a clinical rule
Worked example
T90 7.4% of a 357-minute night, nadir 81%
T90 = 7.4 ÷ 100 × 357 = 26.4 minutes below 90%
Time at or above 90% = 357 − 26.4 = 330.6 minutes
Per hour of the recording: 0.6 × 7.4 = 4.44 minutes below 90% in every hour
26.4 minutes is above the 12-minute figure from Baumert's 2840 MrOS men, where T90 over 12 minutes carried a hazard ratio of 1.59 for cardiovascular death
The percentage alone would not have told you that. The same 7.4% over a 150-minute night is 11.1 minutes, below the cut-off; over a 480-minute night it is 35.5 minutes, well above it. One percentage, three answers
Set the percentage to 2.5 over a 480-minute night and the result is exactly 12.0 minutes, which reads on the LOWER side — Baumert's cut-off is strictly above 12, so this page's first band owns its own upper bound rather than handing 12.0 to the band above
The nadir of 81% is 7 percentage points below the 88% figure Medicare's oxygen documents use as a during-sleep criterion. That is a statement about a coverage threshold in one jurisdiction and about nothing else
Enter a percentage above 100, or a total time of 0, and the page refuses
The same T90 percentage over three different nights
| T90 (% of recording) | Over 150 min | Over 357 min | Over 480 min |
|---|---|---|---|
| 1.0% | 1.5 min | 3.6 min | 4.8 min |
| 2.5% | 3.8 min | 8.9 min | 12.0 min |
| 5.0% | 7.5 min | 17.9 min | 24.0 min |
| 7.4% | 11.1 min | 26.4 min | 35.5 min |
| 20.0% | 30.0 min | 71.4 min | 96.0 min |
Oxygen-related numbers a sleep report prints, and what each one answers
| Metric | Question it answers | Named source |
|---|---|---|
| T90 | for how long was the saturation low? | reporting time below a specified threshold is RECOMMENDED in AASM manual version 3; the 12-minute cut-off is Baumert’s, in 2840 MrOS men |
| Nadir | how low did it get, once? | no published cut-off located; Medicare’s oxygen documents use 88% during sleep as a coverage criterion, which is not the same thing |
| Oxygen desaturation index | how often did it fall, and by how much? | reporting it became OPTIONAL in AASM manual version 3; it did not reach significance for cardiovascular mortality in MrOS (1.13, p = 0.06) |
| Hypoxic burden | how much desaturation, weighted by depth and duration? | Azarbarzin, in 2743 MrOS men and 5111 Sleep Heart Health Study adults: hazard ratios 1.81, 2.73 and 1.96. NOT computable here — it needs the event-by-event saturation trace, which is not a number on a report |
A percentage on the report, a cut-off in minutes
T90 is the time spent with an oxygen saturation below 90%, and it is the oximetry measure the field has been moving toward. The AASM’s own public Summary of Updates for version 3 of the scoring manual records two changes in the same revision: reporting the time below a specified saturation threshold became a recommended rule, and reporting the oxygen desaturation index was demoted from recommended to optional. There is outcome evidence pointing the same way. Baumert and colleagues, working in 2840 men of the MrOS cohort over a mean 8.8 years, found T90 predicted cardiovascular mortality with a hazard ratio of 1.21, and 1.16 after adjustment, while the desaturation index reached only 1.13 at p = 0.06 and was not significant.
The practical problem is a unit mismatch. Reports print T90 as a percentage of the recording; the published cut-off is in minutes. Baumert’s threshold — men with T90 above 12 minutes carried a hazard ratio of 1.59 — cannot be applied to a percentage at all without knowing how long the patient slept. A T90 of 2.5% is 3.8 minutes on a short study and 12.0 minutes on a full night, which lands on both sides of the figure.
What T90 does not distinguish is where the time came from, and the paper that gave the cut-off made the point itself: it split T90 into the component from acute desaturations and the component from non-specific drift in saturation. Twenty minutes accumulated across sixty brief respiratory desaturations and twenty minutes spent continuously at 89% in obesity hypoventilation are the same T90 and nothing like the same night. A single percentage cannot tell them apart, and neither can this page.
The nadir is a third question again — how low, once — and no published prognostic cut-off for it was located while preparing this page. The 88% figure readers meet is Medicare’s: its home oxygen determinations use a saturation at or below 88% during sleep as a coverage criterion, require at least two hours of recorded oximetry, and no longer carry the five-minute rule that still circulates — the oxygen local coverage determination’s revision history records it as removed at revision 9, effective 1 January 2023. A coverage or regulatory threshold is jurisdiction-specific and is quoted here as what that named authority publishes, not as a clinical standard and not as advice. Every number on a sleep study is scoring-rule-dependent and night-to-night variable: the same recording scored under two rules, or the same patient studied on two nights, gives different figures. This page does the arithmetic and names the body that publishes the threshold. It renders no diagnosis and no clinical decision.
Frequently asked questions
How do I convert a T90 percentage into minutes?
Divide the percentage by 100 and multiply by the time it was taken over. A T90 of 7.4% on a 357-minute night is 26.4 minutes. The conversion matters because reports print the percentage and the published cut-off is in minutes: 2.5% is 3.8 minutes over a 150-minute study and 12.0 minutes over an eight-hour one.
Is there a published threshold for T90?
One, with a named cohort behind it. Baumert and colleagues, in 2840 men of the MrOS cohort, reported that men with a T90 above 12 minutes were at elevated risk of cardiovascular death, hazard ratio 1.59, p = 0.006. That is a prognostic finding in older community-dwelling men, not a severity stratum, not a reference range and not a treatment trigger. It is the only numeric boundary this page carries.
Why does T90 appear more prominently on newer reports?
Because the AASM promoted it. Its public Summary of Updates for version 3 of the scoring manual, February 2023, records that reporting time below a specified oxygen saturation threshold became a recommended rule, in the same revision that made reporting the oxygen desaturation index optional. Note the wording: a specified threshold, not necessarily 90%, so T88 and T80 are also reported and are not interchangeable.
Is T90 the same as the hypoxic burden?
No. T90 counts minutes below a line and weights each equally. The hypoxic burden, as Azarbarzin and colleagues defined it, is the total area under the respiratory event-related desaturation curve divided by sleep duration, so depth and duration both count and only event-related desaturation counts at all. It needs the event-by-event saturation trace, which no report prints, so this page names it rather than calculating it.
What does a nadir of 88% or below mean?
In Medicare’s terms it is a coverage criterion: its home oxygen determinations treat a saturation at or below 88% during sleep as a qualifying finding for a patient who is at or above 89% awake, and then cover oxygen only during sleep. That is one jurisdiction’s coverage rule, quoted as what it publishes. No published prognostic cut-off for the nadir itself was located.
Related calculators
References
- Baumert M, Immanuel SA, Stone KL, et al. Composition of nocturnal hypoxaemic burden and its prognostic value for cardiovascular mortality in older community-dwelling men. Eur Heart J. 2020;41(4):533–41.
- Azarbarzin A, Sands SA, Stone KL, et al. The hypoxic burden of sleep apnoea predicts cardiovascular disease-related mortality in the Osteoporotic Fractures in Men Study and the Sleep Heart Health Study. Eur Heart J. 2019;40(14):1149–57.
- American Academy of Sleep Medicine. Summary of Updates in Version 3 of The AASM Manual for the Scoring of Sleep and Associated Events. Darien, IL: AASM; February 2023.
- Centers for Medicare & Medicaid Services. National Coverage Determination 240.2: Home Use of Oxygen, version 2.
- Centers for Medicare & Medicaid Services. Local Coverage Determination L33797: Oxygen and Oxygen Equipment, including the revision history at R9 (1 January 2023).
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/