I am trying to find out nocturnal variability on one of the indexes of OSA over a course of one month. I am not sure if there are similar studies in other disciplines like hypertension or diabetes. I tried to look for some on Pubmed but was unsuccessful, I can provide more details of needed.
Hi. I'm guessing you will have a single measure of your index of OSA on any one night, and thus several values of this index in the one month?
The most common way to report variability in any field is standard deviation (assumes that data are normally-distributed). You may choose to report the standard deviation as a fraction of the mean (called the coefficient of variation) which can aid comparisons between different people with vastly different mean values.
If there are authentic, gradual physiological changes in your OSA index over time, however, and you are mostly interested in assessing repeatability (e.g. measurement error), you could first detrend your data (e.g. subtract off a linear regression) before calculating the standard deviation. Hope this helps.
- Badges
- Science topic