(12-17-2016, 07:25 PM)marcus10 Wrote: Hi all
Last night's AHI on Sleepy Head was 4.23. On analysing the data in SH this was mainly caused by 22 hypopneas, 6 OAs and 4 RERAs. However, a lot of the events scored in SH do not appear to be genuine events.
Out of those 22 events, I'd say only 2 or 3 appear to be genuine hypopneas and the rest appear to be false events. The link is http://imgur.com/a/cu3hS and the link shows 2 typical hypopnea scored events along with a more genuine hypopnea.
Have other people experienced this in SH and how do you deal with them? Do you delete events or "un-score" them to get a more realistic AHI?
I feel pretty good this morning and better than a 4.23 AHI would suggest.
Thanks
SleepyHead does not manufacture data ... the basic raw data comes from your machine and is scored according to the criteria set by the manufacturer. And, while you can set reporting criteria, there does not appear to be any way to alter or filter the incoming data easily -- nor would you want to change a medical record.
I see much the same things, especially where I am convinced that there are false hypopneas in the data. Using the UF2 flag (custom event logging by flow restriction and duration) I am able to get a clearer picture of reality as produced by my particular machine. These custom events are not included in the AHI computation.
As for deleting that data which you believe is erroneous and rescoring the AHI is possible with a lot of effort, why bother? Individual data points are almost meaningless in the greater context of how you feel and how therapy is progressing.
When I was considering becoming a physician, my approach was conditioned by my experience in data processing. Some of the best advice came from my mentor, a rather well-known cardiologist. He said "don't treat the machine, treat the damn patient." Which he then proved by showing it is possible for a heart to display false information (PEA - pulseless electrical activity) where the patient has no pulse and is dying while an idiot looks at the monitor and not the patient. (BTW -- I did not become a physician because I could not divorce myself from the patient and the outcome.)
Use your data as indicative of a trend and keep trying to optimize your therapy using machine settings, not trying to improve your scores by manipulating data. It is what it is!
Best of luck with continued good numbers.
"The object in life is not to be on the side of the majority, but to escape finding oneself in the ranks of the insane." -- Marcus Aurelius