SleepyHead Leak Plots

System One RemStar Pro with C-Flex+ (460P)

There are two plots, dark green and purple, on the daily leak chart. There are nearly parallel with each other. What is the significance of each? During the sleep study the technician fit the mask so the leak value was less than 30. My system one machine records large leaks, defined as ones adversly affecting the therapy.

I would like to how the two plots, large leaks and the "30" limit relate to each other.

MalabarFlash wrote:System One RemStar Pro with C-Flex+ (460P)

There are two plots, dark green and purple, on the daily leak chart. There are nearly parallel with each other. What is the significance of each? During the sleep study the technician fit the mask so the leak value was less than 30. My system one machine records large leaks, defined as ones adversly affecting the therapy.

I would like to how the two plots, large leaks and the "30" limit relate to each other.

Philips Respironics machines report Total Leaks, which includes the amount the masks are designed to vent naturally.

The lower line on these machines is a SleepyHead generated calculation that attempts to automatically remove the natural leaks out of the equation, leaving just excess leaks. It's not a perfect solution for APAP machines (but it adapts reasonably well with the pressure), however the algorithm works even better on plain CPAP.

For more official results, it's best to look at the top line and look up your mask documentation for a table that shows the natural leakage for your set pressure. (Which is where the 30 figure came from the Sleep Tech gave you) Although using those tables doesn't counter for manufacturing variance, or miscalibrations in a CPAP machines cmH2O sensors, which the lower leak line calculations do counter for.

It was either that or I would have to constantly keep updating a database of every single mask out there's leak profiles/pressure tables, and try to construct a formulae for each one. I enjoy solving maths problems, but that's just too much for me to take on... I was working on a way to enter these in and try to curve fit the data automatically, but that's still not the best solution, as I don't really see everyone really wanting to enter all that in either.

There is definite room for improvement here.. but the current solution does work well for isolating problems with leaks.

Better suggestions on how to handle this are always welcome

As for large leaks, there is an event code that has apparently been identified as "Large Leaks" that is imported, but not displayed.. If I can find some data to help verify this, SleepyHead will be able to show these leak flags for PR System One users. (If anyone could get me a sample set of data and a corresponding Encore report showing Large Leaks for a day, it would help immensely with verifying this)