What does my Sleepifi data look like?

What is in the sleep-test report imported: which nights, how often, which quantities, and anything that looks wrong -- so you can check it matches what you expected. Nothing here is tested or judged.

Answer to the Presenting Question: 39 nights, Jun-Oct 2026

Time range: 39 nights, from the night of Jun 2 to the night of Oct 30 2026 (151 days).
Gaps: 2 long breaks: Jun 28-Aug 13 (47 nights); Sep 30-Oct 29 (30 nights).
Frequency: 39 of 151 nights recorded; per month: Jun 16, Jul 0, Aug 9, Sep 13, Oct 1.
Quantities: 21 measures per night, every one filled on all 39 nights (breathing events, oxygen, sleep stages, pulse).
Problems: sleep efficiency partly reflects when recording was started (more than an hour before first sleep on 2 nights); and every value is a home-device estimate, not a sleep-lab measurement.

Do you have a similar question? Join the waitlist.

Questions asked of this data

None yet. Tell us what you want to know, and the answer will be linked here.

When is there data?

One mark per recorded night; hover for the times.

Data Source(s) and Date(s)

Sleepifi home sleep test (a service run by VirtuOx): a wearable on the finger that records blood oxygen and pulse waveform; breathing events and sleep stages are estimated from those, which is why the measures carry an "e".
One PDF report of 78 pages: one summary page per night, then the same nights again as waveform charts. Shared with PerSci by Google Drive on Nov 2 2026.
Anonymised on Bill's computer before anything else read it: the top 5.6 cm and bottom 2 cm of every page (name, date of birth, phone, patient number, device serial) were cut off, and the remainder was checked for those details. The original stays on that computer and is not read again.
Read from the report's own text, not from images of it, so every number here is exactly the one printed. The anonymised text was processed with Claude (Anthropic) to write this page.

Time range and gaps

Sun Jun 28 to Thu Aug 13: 47 nights with no recording.
Wed Sep 30 to Thu Oct 29: 30 nights with no recording.
Otherwise the longest break is 5 nights.

Frequency

39 nights recorded out of 151; recorded nights per month: Jun 16, Jul 0, Aug 9, Sep 13, Oct 1.
Recordings started between 01:10 and 03:00 (clock time as printed), median 02:01; each night is dated by the evening before.
Recordings lasted 5.0 to 12.0 hours.

Quantities reported

Measure	What it is	Nights filled	Range
eAHI, 3% rule	estimated breathing events (pauses or shallow breaths) per hour of sleep, counting an event when oxygen dips 3% or more	39 of 39	2.5-13.4/h
eAHI, 4% rule	the same, counting only dips of 4% or more	39 of 39	1.1-8.8/h
obstructive eAHI (3%)	events where the airway narrows or closes	39 of 39	2.3-12/h
central eAHI (3%)	events where no effort to breathe is made	39 of 39	0-1.5/h
eRDI (3%)	eAHI plus smaller breathing disturbances	39 of 39	4.9-15.6/h
ODI (3%)	oxygen dips of 3% or more per hour	39 of 39	2.3-14.3/h
eAHI in REM sleep (3%)	events per hour during REM (dreaming) sleep	39 of 39	1.6-18.9/h
eAHI in non-REM sleep (3%)	events per hour in all other sleep	39 of 39	0.7-8.3/h
lowest oxygen (SpO2)	lowest blood-oxygen reading of the night	39 of 39	70-92%
average oxygen (SpO2)	average blood-oxygen reading	39 of 39	94-97%
time at or below 90% oxygen	minutes with SpO2 at or below 90%	39 of 39	0-38.37 min
hypoxic burden	depth times length of oxygen dips, per hour of sleep	39 of 39	2.77-22.79 %min/h
total sleep	minutes the device scored as asleep	39 of 39	167.5-494.5 min
REM sleep	minutes of REM sleep	39 of 39	13-137.5 min
deep sleep	minutes of deep sleep	39 of 39	19.5-92 min
awake after first falling asleep	wake after sleep onset	39 of 39	8-367 min
time to fall asleep	from the start of recording to first sleep	39 of 39	4-160 min
sleep efficiency	sleep as a share of the recording	39 of 39	38.2-96.2%
average pulse	mean heart rate	39 of 39	53-65 bpm
lowest pulse	lowest heart rate	39 of 39	40-52 bpm
fragmentation	number of sleep-stage changes in the night	39 of 39	14-42

Each breathing measure comes in two versions, the 3% and 4% rule; the report prints both, and the 4% rule always scores lower. The report also prints its own reference, "<5/h", beside each.
Not recorded: body position, snoring, airflow at the nose, and brain waves. The questionnaire field "Epworth Score" (daytime sleepiness) is blank on every page.

The current stretch

The record so far ends with 23 nights since Aug 14, the most recent on Oct 30. Figures for it are in Browse the data; this page does not compare it with earlier nights.

Browse the data

Show

Every night, every main column:

Night of	Start	Sleep min	eAHI 3%	eAHI 4%	Obstr.	Central	REM	NREM	ODI 3%	Min SpO2	≤90% min	Hyp. burden	Pulse
Tue Jun 2	01:10	280.5	6.2	4.7	6.2	0	9	1.4	6.4	88	5.32	10.73	59
Wed Jun 3	01:45	392.75	6.9	3.7	5.5	1.4	11.2	1.9	6.7	84	1.32	8.57	57
Thu Jun 4	01:19	440.5	3.7	1.5	3.4	0.3	1.6	2.4	3.5	84	1.07	4.78	56
Fri Jun 5	01:03	379.5	6	2.4	5.7	0.3	11.3	1.3	5.8	89	0.3	7.94	57
Sat Jun 6	04:59	426.5	4.8	2.3	4.2	0.6	6.5	2.1	4.5	92	0	6.5	61
Sun Jun 7	02:59	343.5	8	4.9	7.7	0.3	11.7	2	8.6	87	2.25	11.78	55
Mon Jun 8	02:15	284	2.5	1.1	2.3	0.2	6.1	0.7	2.3	91	0	2.77	57
Tue Jun 9	01:23	329.5	2.7	1.5	2.4	0.4	2.9	1.6	3.1	90	0.13	4.65	57
Fri Jun 12	02:25	365	3.1	2.1	2.8	0.3	4.9	0.8	3	91	0	4.33	61
Mon Jun 15	01:43	352	4.6	2.2	4.6	0	8.3	1.6	6	85	5.2	10.38	57
Wed Jun 17	02:27	297	5.1	2.2	4.6	0.4	5.7	2.3	4.6	89	0.25	6.51	56
Thu Jun 18	02:18	345	4.3	1.9	4.2	0.2	4.4	2.1	4.2	89	0.42	5.32	58
Fri Jun 19	02:07	356.5	5.7	2.9	5.6	0.2	7.1	2	7.2	87	5.37	9.39	58
Sat Jun 20	05:31	366.5	6.9	3.3	6.4	0.5	8.1	2.3	9.7	70	21.57	14.19	62
Fri Jun 26	01:45	429	4.3	1.8	4.1	0.3	7.6	1.2	5.9	83	1.58	8.21	59
Sat Jun 27	06:09	167.5	11.8	8.2	11.8	0	18.6	5.8	11.8	86	3.12	19.18	63
Fri Aug 14	02:20	331	13.4	6.3	12	1.5	7.4	8.2	13.1	88	0.45	17.76	62
Sun Aug 16	01:45	433.5	5.8	2.6	5.4	0.4	6.2	3	5.4	83	0.55	8.27	54
Thu Aug 20	01:04	332	10.3	5.2	9.8	0.5	11.8	5.9	10.7	86	13.75	17.46	58
Sun Aug 23	02:27	334.5	11.5	5.4	11.5	0	13	5.4	13.6	85	2.83	18.7	55
Thu Aug 27	01:21	280.5	5.3	1.9	5.3	0	1.9	3.7	5.8	88	0.5	8.75	63
Fri Aug 28	02:46	182.5	6.9	3.9	6.9	0	12.4	3	7.6	90	0.45	10.45	63
Sat Aug 29	03:33	317.5	7.6	3.4	7.4	0.2	5.8	5.1	10	88	2.43	13.46	63
Sun Aug 30	00:55	239.5	2.8	1.3	2.8	0	1.8	1.8	2.8	90	0.22	3.8	57
Mon Aug 31	00:03	385.5	12	6.1	12	0	15	6.1	13.7	84	20.62	21.43	57
Tue Sep 1	01:29	321	9.3	6.9	8.8	0.6	13.6	4.6	10.3	83	38.37	20.06	57
Wed Sep 2	01:09	396.6	4.4	2.9	4.1	0.3	8.2	1.9	6.5	90	0.23	6.65	60
Fri Sep 4	02:45	230.5	9.1	6.5	9.1	0	9.8	5	11.2	85	1.48	17.25	60
Mon Sep 7	02:27	304	5.1	2.2	5.1	0	6.5	3.3	4.9	89	0.13	6.68	54
Wed Sep 9	01:07	399	3.8	2.4	3.3	0.5	7.7	1.8	4.1	84	1.12	7.35	59
Thu Sep 10	02:02	309.5	5.8	2.9	4.8	1	8	3	5.2	88	0.62	8.61	60
Sun Sep 13	00:27	316	7.4	4.4	7.2	0.2	18.9	3.4	8.2	86	3.95	14.62	53
Fri Sep 18	02:55	494.5	9.2	5.5	7.9	1.3	8.8	4.7	11	89	2.27	13.97	58
Sat Sep 19	04:39	205	3.5	1.2	3.5	0	5.2	2.3	4.7	92	0	6.01	57
Wed Sep 23	01:04	424.92	3.8	2.7	3.7	0.1	10.7	0.9	3.1	86	1.17	4.82	57
Thu Sep 24	02:03	185	12	8.8	12	0	11.4	8.3	14.3	84	3.13	22.79	65
Sun Sep 27	00:32	267.5	3.1	2	3.1	0	3.1	1.8	4	85	0.55	6.57	54
Tue Sep 29	00:54	226.5	7.7	4.8	7.2	0.5	8.5	3.9	8.5	90	0.07	9.4	54
Fri Oct 30	03:00	360	9.3	4.2	8.5	0.8	14	4	10.3	86	1.7	12.82	65

Problematic issues

Efficiency depends on when recording started. The device counts from the start of recording; on 2 of 39 nights that was more than an hour before first sleep (longest 2.7 h). Efficiency ranges from 38.2% to 96.2%, and on those nights it says more about when the device was switched on than about sleep. Total sleep does not have this problem.
Estimates, not a lab study. Every breathing figure is estimated from a finger sensor, which is why the report calls it eAHI rather than AHI.
Two breaks of 47 and 30 nights; nothing is known about those nights.
Each night appears twice in the PDF (summary, then waveforms); the waveform pages' few printed numbers match the summary's, so this page reads the summaries only.

Actions Recommended by PerSci

Tell us your question: how serious this is, whether something helps, whether to get a lab study -- so the next page answers it.
Tell us about the two breaks: a pause in testing, or recordings that exist but were not in this report?
If the sleep-test service can export the nightly oxygen and pulse readings as a file, ask for it: it would allow looking inside each night.

If you have a health-related question about your own data, join the waitlist for your own custom analysis.

Act soon: it is free while we stand this service up and refine it. All we ask in return is permission to publish a results page like the ones we have made for our other clients. We never publish your surname or your date of birth. The page shows a first name or a pseudonym, whichever you prefer, and your age in years and your sex.

Made by PerSci from your Sleepifi report, Nov 6 2026. It describes your data; it is not medical advice.

PerSci is not medical advice. Analyses describe one person's own data and are not a diagnosis or a treatment recommendation. Full disclaimer · Privacy Policy