Question on Flow Limitations

[super7pilot]

Would the correlation be a percentage of restriction. I.E. my readings last night were a max of .37 for just a very short time. Over all it read .07 @ 95% and .21 @ 99%.

[robysue1]

It's important to understand that Resmed has not made much information public about the vertical "scale" of the Flow Limitation graph.

The official Resmed software for looking at the detailed data from a Resmed xPAP machine is called ResScan. And in ResScan there is not a numerical vertical scale on the flow limitation graph. Instead, there are three icons:

1) a very round "inhalation" shape at the bottom of the graph
2) a somewhat peaked "inhalation" shape in the middle of the graph
3) a flat "inhalation" shape at the top of the graph

The documentation I've been able to find simply says this graph is a visualization of flow limitation from "flat to round" and that the taller a peak is on the graph, the more severe the flow limitation is. "More severe" seems to indicate the more distorted the flow limitation is rather than the size of the inhalations themselves.

Would the correlation be a percentage of restriction. I.E. my readings last night were a max of .37 for just a very short time. Over all it read .07 @ 95% and .21 @ 99%.

Because of the lack of official documentation from Resmed about what the "vertical" scale actually means on the Flow Limitation graph, it's probably not a good idea to assume anything about a precise meaning for the numerical values that Oscar reports for this graph. In other words, I would not assume that a "peak" at .37 means that the flow was "flow limited by 37%".

If anything, the flow limitation numbers reported in Oscar seem to be more to be an intuitive idea of how distorted the shape of the inhalations are. (The idea is that distorted inhalations may indicate the airway is at risk of collapsing, and Resmed Auto machines are programmed to increase the pressure when flow limitations of a certain size are detected.). So what this means is that a flow limitation with a peak at .37 is not as badly misshapen as a flow limitation with a peak at .67 would be.

Similarly, the 95 percentile and 99 percentile numbers that Oscar reports for the Flow Limitation graph are also something that has no well defined mathematical or statistical meaning. In general if these numbers are close to 0, then there is usually almost no activity to speak of in the flow limitation graph and it would be fine for a CPAPer to just ignore those numbers. Conversely, if there are a lot of significant peaks and a lot of activity in the Flow limitation graph, you can see that even before you look at the 95% or 99% percentile number in Oscar for this graph.

And it's also important to keep this in mind: When you start zooming in on flow rate graphs where the flow limitation graph indicates something is "wrong", sometimes you can clearly see misshapen inhalations and sometimes you can't, particularly if the "flow limitation" number is less than about .25 or .3. And conversely, sometimes when you go through the entire night's flow rate graph zoomed in enough to see individual breaths, you'll see inhalations that look "distorted" when compared to their near neighbors, but there's no flow limitation scored in the flow limitation graph.

Whether flow limitations are or are not significant requires some willingness to start analyzing the flow rate graph while zoomed in enough to see the individual inhalations. But it also requires figuring out whether nasal congestion and/or a deviated septum might be causing inhalation shapes that get mis-scored as "flow limitations". In general, if the flow limitation graph gets better as the pressure increases, it's reasonable to assume the flow limitations are real and the pressure increase was worthwhile. On the other hand, if flow limitations persists or get worse as the pressure goes up, it may be an indication that the machine is increasing the pressure inappropriately.

Links to ResScan documentation that I can find that don't require any logins:

ResScan Clinical Guide from 2011. Available at apneaboard.com. The Flow limitation graph is discussed on page 43.

ResScan 6.0 Clinical Guide. Available from an Israeli cpap forum, but the guide is written in English. The Flow limitation graph is discussed on page 25

[super7pilot]

Thanks robysue1, I guess they want it obscure, undefined.

[robysue1]

Thanks robysue1, I guess they want it obscure, undefined.

More likely Resmed has never done the careful analysis to try to quantify what those Oscar numbers mean. As I said, in ResScan, there aren't any numbers shown on the vertical axes for the Flow Limitation graph.

Resmed's engineers don't try to worry about every possible flaky thing that can happen during a night of using a Resmed PAP machine. They know enough from the folks in sleep medicine who study flow rate charts for a living to know that when the shape of the inspirations starts to get distorted, that can be an early sign the airway is in danger of collapsing enough to trigger a hypopnea or an obstructive apnea.

But the actual algorithm for how the Resmed machines decide that a particular set of breaths is distorted enough to warrant being flagged as a "Flow limitation" that warrants increasing the pressure setting is a piece of proprietary information. All we know is that the algorithm for evaluating the flow rate data is based on a moving average of what the current inspirations and expirations look like, and things get flagged when stuff looks sufficiently "different" from that moving average to the machine's algorithm. But like all algorithms, the Resmed algorithm for scoring events is not perfect.

And it's worth keeping in mind that Resmed and other CPAP manufacturers specifically say that the data of most use for evaluating efficacy is trending data. In other words, is the AHI usually low enough? Is the flow limitation graph usually not too busy? Is the snore graph usually pretty empty? And it's also worth keeping in mind that Resmed and other CPAP manufacturers also say that all of this data has to be used in addition to the all important subjective way an individual patient reports feeling while using the CPAP all night long, every night.

[super7pilot]

Very true robysue1,