Thoughts on the nature of sleep studies

[NapsZ]

I was going through a summary of a titration sleep study I had performed on me last month, and it occurred to me that the term "sleep study" is something of a misnomer. To me, the term "study" brings to mind a careful and deliberate extended period of observation and experimentation to arrive at an accurate result.

But these studies remind me more of a final exam. It's a measurement over a brief time, and if you're well-prepared and fortunate, your condition shows up in the right way at the right moment, you "pass", and the results are used to determine a treatment that will hold for some years to come. I certainly remember lying there staring up at the ceiling trying to get to sleep. "What if I can't sleep?" "Will I get enough REM sleep?" "Will taking a sleeping pill mess up the results?" "Will I need another study?" "IT'S COSTING ME $200 AN HOUR TO BE HERE!!" And that makes it harder still to fall asleep. At least if you're taking your SATs or GREs or civil service exams you can drink coffee and be fully awake and motivated -- but try telling your subconscious to get with the program! not easy!

The measurements made are quite accurate, but are they precise? Precision goes as the inverse square root of the number of data points. Let's say you end up with 3 intervals of REM sleep lasting 10 minutes long each. How many different titrated pressures can they try? 3 pressures with N=1 (not at all precise)? or 1 pressure with N=3 (still not great, and no titration)? What if the condition depends on the time course of the applied pressure? When doing an experiment, parameters usually need to be varied more slowly than the time course of events being averaged over. But so many variables are being changed quickly, and simultaneously! pressures rising at the same time sleep stages are shifting. Tossing and turning. Mask changes and seal adjustments. The sleep tech told me it was really an art getting a perfect titration in a few hours, and I believed him.

On the other hand, it looks like many forum users have discovered the value of "slow titrations" performed over weeks using data-capable machines. Not as accurate (only 1 or 2 sensors at best), but far more precise -- averaged over more days, mask adjustments, sleep positions, sleep stages ..

But this has to be done on the sly, for liability reasons, for insurance reasons, because there are incentives built in at every level to keep the procedures the way they currently are. I am sure Resmed could easily release their software for purchase by end users. That's not happening any time soon though because their primary customers, the doctors and DME suppliers, wouldn't want to be pushed out of the loop. And to be fair, it only takes a few patients to mess up their settings and suffer adverse medical consequences and file a lawsuit to .. yeah.

Can't there be an in-between solution though? My doctor gave me a machine with fixed pressure settings and no data, and told me to check back with him in 6 weeks, where they could adjust further based on whether i felt "better" or "worse." I asked the RT, "can I adjust the pressure?" She was like, "No no no no. Absolutely not." But I was thinking, "what if something goes wrong in those 6 weeks?" I called up my aunt. She herself is a CPAP user, and a medical doctor to boot. She was like, "oh I don't know what to tell you. I just upgraded to the latest S9 auto machine, but I still feel tired all day. I've got to go see my doctor friend so she can look at my machine for me." I ended up forwarding her a link to Resscan 3.13 so she could look at her own sleep data. Even doctors aren't doing this for themselves!

Anyway this was long-winded. But this led me to my most recent morning "in the shower" thought. Why can't there be a better way to do sleep studies and sleep titrations? Like why couldn't someone take an autoset machine, and place by its side a hardware board with some extra sensors added:
-- a way to determine sleep stages. For REM vs. NREM maybe a pair of electrodes (like the Zeo) + accelerometer? or even something lower tech, like a two fluorescent dots drawn on the face with a highlighter (one on the eyelid, one on the cheek as a reference), watched by a webcam. And to distinguish asleep vs awake a switch you hold, that you release involuntarily when your muscles relax?
-- a chest strap to distinguish centrals from obstructives.
-- an oximeter.
all integrated together and recorded continuously, with a time-stamp so it could be synchronized to the machine. That wouldn't be too expensive to build right? and even cheaper to build once, and rent ..

Wouldn't this get most of the essential data? and if someone could use it over the course of a week or a month, they could look at different sleeping positions, or sleeping pills, or number of pillows beneath one's head; while varying pressures, and masks, and all the other variables one can't easily examine for $1000/night. And really be able to do a sleep study ...

[rested gal]

The sleep tech told me it was really an art getting a perfect titration in a few hours, and I believed him.

I believe that, too. As a side note, I've always thought..."give me a good, experienced sleep tech and forget the sleep doctor."

My doctor gave me a machine with fixed pressure settings and no data, and told me to check back with him in 6 weeks, where they could adjust further based on whether i felt "better" or "worse." I asked the RT, "can I adjust the pressure?" She was like, "No no no no. Absolutely not."

heheh. The RT would be horrified to read what many of us in the forum do, wouldn't she?

Going by "better" or "worse" (and probably with some input from a bed partner... "you're still snoring" or "you're not snoring any more") seemed to do a pretty good job for the people in this study -- adjusting the pressure themselves at home. And without benefit of any machine data, either:

Link to a study that concluded, "yes."
"Can Patients with Obstructive Sleep Apnea Titrate Their Own Continuous Positive Airway Pressure?"
http://www.tnlc.com/Lara/laura/osa/CanP ... wnCPAP.pdf

[Carl LaFong]

I believe they need a minimum of two hours unassisted sleep to establish a baseline and plan of action. In my split study they made four pressure adjustments and tried two masks to get me "optimal".

[Jay Aitchsee]

Napz, I thought the same thing so I set about building my own sleep lab. See my signature line. I've got an autoset, Zeo, Pulse Oximeter, and an Infrared time lapse video. I bought all of this out of pocket, the autoset was $350 (used), the zeo, was $150, the pulse-ox about $50 and the infrared set-up home built for about $50. So, there you go, about $600 for a pretty good home sleep lab. Maybe not as accurate as the PSG but my sleep doctor sure likes it and it produces repeatable results so trends can be identified.
Still working on synching them all together, but what works fairly well is snipping the results from each machine and pasting them into a draw program where the time-lines can be adjusted to match. See "My Zeo a Report" viewtopic/t68643/viewtopic.php?f=1&t=68 ... 46#p636846
Jay

[ameriken]

As a side note, I've always thought..."give me a good, experienced sleep tech and forget the sleep doctor."

Bingo! In my first study/titration with the VA, I failed the CPAP machine, meaning it did nothing for my apneas. They had me go back for a second study and I had the same RT. He looked at the Dr request and based on the first study, the Dr requested a straight-up BiPAP. Then he read my prior test and said he was going to change the Dr request. He said he can change it if he had a justified reason to do so, and so he told me that it was justified since I had a 'potpourri' of breathing issues in addition to OSA, so he was going to change the order to a BiPAP ASV. Lo and behold, that worked, thank God for good RT. I sometimes wish I could meet with the RT rather than the doctor, I get better answers from them.

However speaking of the nature of sleep studies, the sleep lab had 2 rooms for 2 patients. The other was a no show and the RT said it was a good thing because he really needed to focus more attention on my breathing issues and I guess make adjustments as my events occured. I don't know exactly why or what that really meant, but on the other hand I wonder if that meant I would not have had a good titration if the other person showed up. Also makes me wonder why the Doctor wasn't the one to order the ASV.

[NapsZ]

Hi Jay, yeah, that's exactly the sort of system I was thinking of. *clicks on link* It looks like your Zeo is pretty accurate, to the point that your events correlate well to REM sleep. I was a bit scared off by reports that Zeo wasn't too accurate after reading some of the reviews on Amazon. But that could just be a vocal minority. I'll give the system another look (my sister forwarded me a link!)

And to everyone else, yeah, I agree, sleep techs can do a great job getting the most out of the brief time available during a sleep study. It can make a big difference. After a long night where my sleep tech worked hard to get me successfully titrated, I brought him some chocolates as a thank you

[moresleep]

I'm a big believer in home titration--by the patient and machine, if necessary, although I suppose in an ideal world, we would have RTs who made housecalls. For both of my sleep studies, I really couldn't sleep much with all those electrodes and in that environment. At most I dozed fitfully from time to time. Under those circumstances, the techs would have needed to be clarivoyant to get the titration right.

As that study on the efficacy of self-titration shows, with even minimal instruction, Cpap users can do a perfectly fine job finding their own optimum pressures, even without the benefit of Apap machines and software. The additional technology is icing on the cake.

While the medical establishment pretends otherwise, there is no real evidence that self-titration is at all dangerous for most people. I haven't heard of even one person dying from misadjusting their Cpap machine. Apparently, in Australia a prescription is not even required for a Cpap purchase.

[Jay Aitchsee]

NapsZ, I was sceptical of the Zeo, too. I have no doubt it reports sometimes in error simply because it uses an algorithm with some basic assumptions. One of which probably is that you're in bed. Another thing that may cause it to appear to be reporting in error is that the graphic display is in 5 min intervals. Although, the data reported is in 30 sec intervals. You can see the 30 sec intervals by downloading to your own spread sheet (easy). If you want to see the raw data, it's available as well. I find the Zeo to be quite satisfactory for objective assessment of a single night's sleep, but I think its strength is in being able to gather that information over time, to establish trends, and develop the "precision" you mentioned in your post.
If you want to try your hand at putting together an infrared video system, you'll find the info in this thread: viewtopic.php?f=1&t=65101&p=607701#p607701 . You'll need a laptop, an inexpensive webcam, and a few parts from Radio Shack. viewtopic.php?f=1&t=65101&st=0&sk=t&sd= ... 15#p609795

Jay

[rested gal]

As that study on the efficacy of self-titration shows, with even minimal instruction, Cpap users can do a perfectly fine job finding their own optimum pressures, even without the benefit of Apap machines and software. The additional technology is icing on the cake.

I really should mention that the people in the study who successfully self-titrated at home most likely had plain vanilla Obstructive Sleep Apnea -- which is probably the most common sleep disordered breathing.

A person who has primary central apnea, or complex sleep apnea (application of cpap causing persistent centrals in great numbers) would not be a good candidate for home self-titration. They'd need a good smart sleep tech, like ameriken was lucky to have. Then... a good sleep doctor could be the icing on the cake.

[NapsZ]

Thanks Jay. Good to know you can get more detailed data out of the Zeo. I read a little about the algorithms they use, it looks like they use a neural network to map the data coming out of that single electrode into a sleep state. So I can imagine it would work great for say, 80% of the populations, but then fail to work for anyone whose brain patterns are too different than their training set. I guess the only way to find out is to get one! well ideally while also hooked up to a real ECG for comparison purposes. But I don't think you could get away with bringing a Zeo into a sleep study setting ...

You posts on creating an infrared webcam are neat! I had no idea it was that simple to do. Also amazing that the heat from a single 9 W christmas tree bulb is enough to illuminate the entire bed. I guess it really does highlight how much energy these incandescent bulbs put out as heat vs visible light! On an off topic note, I recall some enterprising business person reselling 100 W incandescent lamps as "heat globes", to get around regulations banning them in more and more locales ..

[NapsZ]

Yeah that's a good point, rested_gal. I could imagine a scenario where a person is on CPAP, doesn't feel like it's doing it's job, bumps up the pressure, still feels worse, increases the pressure again, and so on .. not realizing the problem is actually being made worse due to centrals. Maybe better algorithms by the machines would alleviate that, but yeah, in many cases there is no substitute for a smart human keeping an eye on things!

[chunkyfrog]

Sadly, we cannot put much stock in online reviews--they could be from anyone--including soon-to-be-unemployed sleep techs.