Is this a true saturation drop or an artefact?

[drmaestro]

Hi,

I occasionally see some saturation drops which are instantaneous and are generally the worst values for the night. For example:



I find it too sudden to be true and also there is no flow event prior to it, at least not a serious one.  The one below is what I'd normally expect saturation drops to be like:



More gradual and there is a respiratory event before it. Do you think the oxymeter can record false readings or do I need to look into them?

Thanks...

[jaswilliams]

Those sudden drops are likely sensor movement drops the o2 levels don’t drop thatfast

[2SleepBetta]

The most consistent pattern I see in my larger (non-artifact) increases in SpO2 is that those come 30 seconds after the first larger recovery breath after what I consider a  mini-RERA. The larger several second (non-artifact) desats are less clearly marked by FR, TV or MV, but are somewhat consistent. This is what I see using a Contec CMS50I oximeter on my second finger of my left hand. I assume the lag in SpO2 rise is from time it takes better oxygenated (recovery breath) blood from the lung to be pumped to my finger tip. This is interesting, but I don't know a way to change what I see if I wanted to, the VAuto having reduced my FL and SA to minimal levels.

I marked your graphs with pairs of shorter and longer bars to illustrate what I think happened in your desat. The first bar marks the point where the FR has been low and about 75 seconds later the third bar marks the corresponding desat. The second bar marks the largest inhalation after the low FR and it corresponds, in my scheme, with the sharp rise of SpO2 from the local average, again, 75 seconds later. I find that when large (non-artifact) swings are happening FR and SpO2 curves are about 30 seconds out of phase--SpO2 points of rise or drop are about 30 seconds after significant change in FR, TV or MV--FR particularly.

You might look at displays of FR and SpO2 that show the full height and depth of the two curves. Check for a pattern, beginning with checking the time that has elapsed between a sharp increase in FR, TV or MV and the next sharp start of a rise from a low SpO2. In my case that is the most consistent part of a pattern that is less evident in pinpointing the time when FR has fallen and causes a desat.

2SB