"Wearable Wars: Evolution in Sleep Assessment" Douglas Kirsch (02.10.2021)

February 26, 2021

ID6231

To CiteDCA Citation Guide

00:17Alright, hi everybody,
00:18I am Lauren Tobias. I would
00:20like to welcome you to our Yale
00:22Sleep Seminar this afternoon.
00:24I'll make a few brief announcements
00:26before I turn it over to Larry Epstein
00:29to introduce our speaker today first.
00:31Please do take a moment to
00:33make sure that you're muted.
00:35In order to receive CME
00:36credit for attendance,
00:37please see the chat room for instructions.
00:39There's a unique ID that you can text
00:42anytime until 3:15 PM and if you're
00:44not already registered with go see me.
00:46You will need to do that first.
00:48If you have questions
00:49during the presentation,
00:50please make root use of the chat room
00:53throughout the hour and we will invite
00:55people also to unmute themselves and
00:57ask questions allowed at the end.
00:58We will have recorded versions of these
01:01lectures available online within two
01:03weeks at the link provided in the chat.
01:05And please feel free to share
01:07it announcements for our weekly
01:08lecture series to anyone that
01:10you think may be interested,
01:11or contact Debbie Lovejoy to
01:13be added to our email list.
01:15And then before I introduce today's speaker,
01:17I just want to let you know that
01:19next week stock is going to be
01:22given by Doctor Francesca Facco,
01:23who is an associate professor of obstetrics,
01:25gynecology and reproductive Sciences
01:26at the University of Pittsburgh,
01:28and she's going to be speaking
01:30about sleep and pregnancy.
01:31So please make a note of that
01:33and join us next week as well.
01:36So with that I will turn it
01:37over to Larry Epstein.
01:39Thank you, it's my pleasure
01:41today to introduce Doctor Doug.
01:42Curses are speaker doctor.
01:44Kirsch is a good friend is and is
01:46well known to many of us with him
01:48having spent a long time in the
01:51Boston area tended medical school at
01:52the University of Massachusetts to
01:54this neurology training in Rochester
01:55and completed a sleep fellowship
01:57at the University of Michigan.
01:58He was a member of the Brigham
02:00program for many years.
02:02First sleep health centers and at
02:03Brigham and Women's Faulkner Hospital.
02:05He's currently the director of the
02:07Sleep Program for Atrium Health,
02:08the largest healthcare system
02:10in North and South Carolina.
02:11And a clinical professor at
02:13UNC School of Medicine.
02:15He is familiar to many of you.
02:17As an educator,
02:18running many Jimmy courses,
02:19and having served as the program
02:21chair for the Annual Sleep Meeting
02:23for the American Academy of Sleep
02:25Medicine and Sleep Research Society.
02:27He's also served on the Board of
02:29directors of the ASM and is a past president.
02:33He's long held an interest in emerging
02:36sleep technologies and help has
02:38helped direct the ASM's valuation
02:40and policy towards new technology.
02:42Appropriately,
02:42his topic today is wearable Wars
02:45evolution in sleep assessment.
02:48Thanks for that kind introduction Larry,
02:51and I'd like to say coming from the
02:54Boston area, I can not think of
02:56many times at which Harvard and Yale
02:59have actually done things together.
03:01This seems like a novel in different world,
03:04so maybe while Covid has brought
03:07a lot of terrible things upon us,
03:09maybe joint Harvard Yale,
03:11things may be a bonus.
03:13Covid park. So with that,
03:15let me start by telling you a
03:17little bit about this.
03:22The world of sleep obviously has gone
03:24through a lot of changes over the years.
03:27For a long time, sleep problems
03:30were really not medical problem,
03:32but then overtime they became an
03:35ultra specialized medical problem.
03:37And they became specialized.
03:38But then testable in the home.
03:41And then Lastly,
03:42your mobile phone can be asleep tester.
03:44But the reality is this isn't the timeline.
03:47This is the timeline right?
03:49That the speed of this change has been rapid.
03:55We are all comfortable with seeing this
03:57kind of image in lab polysomnographers
03:59he is and remains the gold standard
04:02for us to assess patients sleeping
04:04problem in this particular case,
04:06you're seeing a patient who has pretty
04:08significant obstructive sleep apnea.
04:10This is a two minute respiratory page
04:12in the bottom and you can see they
04:15have a prolonged apnea with a longer
04:17saturation on the tailing behind it.
04:23More and more comfortable we have
04:25gotten with home based testing,
04:27and so here's an example of a
04:29person who has obstructive sleep
04:30apnea in the home based setting.
04:33A lot less leads,
04:34but the diagnosis still pretty clear.
04:39It starts to get trickier though in the home,
04:42sometimes the subtleties of
04:44sleep apnea can be missed.
04:45It can be harder without some of
04:47the things that were comfortable
04:49with all those, EG leads.
04:51These kinds of events that may or
04:53may not be events, but we can't tell.
04:56We don't have arousals.
04:57We have no way of kind of knowing
04:59in this particular type of home
05:02based environment what's going on.
05:04And at the same time,
05:06here's a 5 minute page of a home based test.
05:09Is that a four and a half minute apnea?
05:13Probably not right.
05:14That person is.
05:16Clearly removed the nasal flow
05:18for a period of time and then
05:20put it back into place.
05:21On the other hand,
05:22somebody who's not well trained
05:24in this wouldn't recognize that
05:25a four and a half million Afghan
05:28may not be a realistic thing.
05:30Home base tests aren't perfect,
05:31but we are increasingly comfortable
05:33with some of the areas in
05:34which they have faults.
05:38So moving then into the world in which we
05:41are going to have to become increasingly
05:44comfortable with is is this sort of
05:48idea of consumer sleep technology.
05:50So consumers like technology are
05:52really non prescriptive devices
05:54that are directly marketed to
05:56consumers that make an assertion to
05:58perform sleep monitoring tracking.
06:00Or sleep related interventions
06:02and these can take different.
06:05Sir forms that are apps which
06:07obviously we're all comfortable with.
06:09This device programs that
06:10run on our mobile devices.
06:12There are wearables which
06:13are things that people wear.
06:15Obviously the consumer wears
06:17during their sleep period to both
06:19potentially assess their sleep
06:20in that particular situation.
06:22And then there are near Obols,
06:24which are devices that are in the
06:27environment that track people sleep.
06:30So let's start with what these
06:32devices often started with,
06:33which is that measured activity,
06:35and so this is a study from 2015 that says,
06:39broadly,
06:39that a whole bunch of different
06:41activity trackers that their
06:43reliability and validity of trackers
06:45for measuring step count is good.
06:47However,
06:47research in real life did
06:49not necessarily intervene,
06:50and so this is one of my
06:52favorite articles I've read,
06:53which was from the well known
06:55scientific Journal Vanity Fair,
06:56where it says that he was wearing a
06:59Fitbit and jawbone at the same time.
07:01In the Fitbit said I had 7000 steps,
07:03Jawbone had me 2000 steps,
07:05which was a bit a discrepancy,
07:07and he says it's somewhat dismaying to
07:09check my daily graph on the Jawbone app.
07:11And behold,
07:12a few skyscrapers spikes of activity
07:14between wide plateaus is suspended animation.
07:16By comparison,
07:16my Fitbit chart looks downright jazzy.
07:19And as he say,
07:21further testing is clearly indicated.
07:24And so,
07:25here were some more scientific
07:27based looks at some of these
07:29devices and the there were accuracy
07:31issues in terms of well you can
07:34get 100 steps for eating cereal.
07:35You're getting 1000 steps by shaving
07:38and brushing your teeth, right?
07:39That movement and steps are
07:41not the same thing,
07:43and so recognizing there
07:44are some challenges here.
07:49So in the early world of using
07:51wearable technologies they were using
07:53mostly micromovements and heart rate.
07:56And the accuracy with which they
07:57tracked sleep was somewhat contentious,
07:59and so why is that?
08:01Well, because not everybody
08:02sleeps in the same way.
08:04People who move a lot during
08:06sleep may be considered awake,
08:07but they may actually be sleeping.
08:09People who sit for long periods
08:11of time but aren't sleeping may
08:13be considered to be sleeping.
08:15And there are settings within
08:16devices whether you're in
08:18regular mode or sensitive mode,
08:19that will actually change the results.
08:22So that makes this whole process a
08:24little bit concerning when you are
08:26a clinician trying to look at this.
08:30So. How is the difference is right?
08:33Two, broadly speaking,
08:34we're all comfortable with the lab.
08:36We hook him up with all these wires.
08:38All these things.
08:39We look at a bunch of funny looking waves
08:42and we have a scoring mechanism that
08:45gets us into different stages of sleep.
08:47However, these wearable devices,
08:49for instance, you get some amount of motions,
08:52amount, heart rate.
08:53There's a black box algorithm that sort
08:56of spits out some amount of information.
08:59And I'll thank Kathy Goldstein
09:01for this image.
09:02I highly recommend this article
09:03that she wrote about the
09:05topic we're discussing today.
09:08So what came of this one, 2015?
09:10There was a review and you can
09:12see the devices that they kind
09:14of reviewed and more or less.
09:16I love this discussion.
09:17The review identified a critical lack
09:19of basic information about the devices.
09:215 out of 6 devices provided no supporting
09:23information on their sensor accuracy
09:25and four out of 6 devices provided
09:27no information on their output.
09:28Metric accuracy so broadly in 2015,
09:30which is not that long ago.
09:32Five years ago there was really
09:34no data about these things.
09:36But then we start moving
09:38in on gathering data,
09:40so this is data from 2012.
09:43Looking at an early Fitbit and
09:45measuring it against an active graph
09:48and there were some differences
09:50there versus Poly sonography.
09:52And that they found that the Fitbit
09:54overestimated the time participants were
09:56asleep by about 67 minutes on average.
09:58So about an hour missed.
10:00Compared to the polysomnogram.
10:04You can see it a few years later,
10:07again 2015.
10:08This is when this data start taking
10:10off and you can see differences
10:12in children and adolescents.
10:14So in normal mode there's
10:16an overestimation of sleep
10:17insensitive mode for the Fitbit,
10:19there is an underestimation of sleep
10:21and again this is compared to PSG,
10:24which is considered articles standard.
10:29Jawbone company that doesn't exist and
10:31went out of business in 2016, 2017.
10:34This was early data about
10:36that device and it said, oh,
10:39this device is pretty decent agreement
10:41with Poly, some Poly sonography.
10:43However, there were published in 2015,
10:46at which time the Jawbone Up,
10:48which was the device that they tested,
10:50was now three generations behind where they
10:53were actually the consumers were using,
10:55so the IT was great that they had
10:57tested it against polysomnography.
10:59But was the new device is better?
11:02Worse, different hard to know the
11:05speed of consumer technology changes
11:07done it much faster rate than that
11:09of the scientific technology.
11:11And so this is what you can see.
11:14Is that in 22,008 to 2015 there
11:16were at least nine tracker from
11:19Fitbit and six trackers from Jawbone
11:21in a short period of time,
11:23and that speed of change has continued.
11:28What you can see, though,
11:29is that the data around these devices
11:32is growing rapidly, so you can see
11:34where I was kind of pointing out,
11:36and that inflection .20 fourteen
11:382015 that there was not much data.
11:40So this is a graph of the publications
11:43that cover Fitbit and sleep in Pub Med,
11:45and you can see it climbing rapidly.
11:48There is an entire research library
11:50devoted to Fitbit, and so there
11:52are almost 700 publications as one.
11:54I checked this the other day.
11:57And so, recognizing that we're gathering more
12:00and more information about these devices,
12:02but the challenge is that the speed of
12:05change in the consumer world is still
12:07faster than that of the research world.
12:12So as of 2021, what is Fitbit saying
12:14about their sleep calculation?
12:16They say Fitbit estimates your
12:18sleep stages using a combination of
12:20movement and heart rate patterns.
12:22While you're sleeping,
12:23your device tracks beat to beat
12:24changes in your heart rate and
12:26his heart rate variability,
12:27which fluctuate as you transition between
12:28light sleep, deep sleep and REM sleep.
12:31And then they sort of use that data,
12:34calculate sleep.
12:34So let me show you a couple of cases.
12:36These are cases that.
12:39That were patients of mine.
12:41So this is a 42 year old woman who
12:43came to see me said needs more
12:45sleep than the average person.
12:47She's always been someone who's
12:48just been kind of more tired.
12:50If she doesn't get that 9 to
12:5310 hours of sleep.
12:54And so she got in a Fitbit fairly early,
12:58self tracks or sleep and she sees this.
13:01Now again, this is early Fitbit data,
13:04but pointing out the fact
13:05that her sleep is fragmented,
13:07that she's considered in restless time
13:09for about 220 out of this 200 minutes
13:12out of this under 7 hours of recording time.
13:16Well,
13:16she comes with a sleep doctor.
13:18It turns out she snores and has
13:20restless sleep and she's not a
13:22particularly overweight person,
13:23but she gets a home sleep apnea test done.
13:27Finds that she has moderate sleep
13:29apnea and she was placed on CPAP
13:31and is sleeping much better.
13:33Case 254 year old presents to the
13:35clinic with Insomnia state she
13:37doesn't sleep well during the night,
13:39even with medication.
13:40She's using ambient somewhere between
13:4110 and 20 milligrams nightly.
13:43She often will say that she just
13:46doesn't sleep at all at night now.
13:49And she too is tracking her sleep
13:51with a Fitbit. And this is her.
13:53Her Fitbit data,
13:54which shows almost no restless time.
13:56Now it's not a particularly long
13:58period of time.
13:59It's about five hours of recording,
14:01but it shows she barely moved and
14:03suggests he's in fact asleep now.
14:05Is she sleeping? She not asleep?
14:07The argument being is very hard to
14:10stay that still for five or six hours,
14:12and so the suggestion is she's
14:14probably asleep,
14:15which is common in patients who
14:17have insomnia.
14:17But she was unwilling to admit
14:19this could represent sleep.
14:21That there was something wrong with her.
14:282015 timeframe Chris Winter, who's a
14:30known sleep doc down the Virginia area,
14:32decides to do something for the Huffington
14:35Post and sort of straps a whole bunch
14:39of different devices to himself.
14:41And you can see he kind of
14:43lines them all up together.
14:44Kind of post the data as they look at
14:47sort of each of the devices in sequence,
14:49comparing them all against
14:51polysomnogram at the top.
14:52And you can see the differences
14:54in what that data looks like.
14:56That different different
14:57pieces of data lined up.
14:59You can see the basis Health Tracker
15:01was one that looked a little
15:02bit more like what we would have
15:04expected compared the polysomnogram,
15:06but each device has its own data.
15:09That is what I would say not quite right.
15:12And So what he says here is with one
15:14only one night of data collected,
15:16I'm reluctant to declare winners and losers.
15:18But more purely sleep mounting perspective,
15:20the basis seem to distinguish self
15:23from her perspective and accuracy.
15:25And so does that mean anything?
15:27Now it was a one night study of a bunch of
15:30things put together back five years ago,
15:32but the point being is that there's an
15:34enormous level of variance in how these
15:36different devices track and look at data.
15:38And they are all for the most
15:41part done in a black blocksworld.
15:44So Fitbit now has over that 5-6 ten years
15:47gathered an enormous amount of data,
15:49and they can push out data about some
15:52of the things that they have, right?
15:54So this is average results based
15:57on millions of nights.
15:59So why do you like the data or not
16:01in terms of how they interpreted the
16:03fact is they have so many users and
16:06so many nights compared to in some
16:08ways we have in our laboratories that
16:10regardless of how right or wrong it is,
16:12there is an enormous amount
16:13of information there,
16:14and it's potentially helpful,
16:15and so you can kind of see that the
16:18average Fitbit user is sleeping
16:20for better or worse what they say,
16:226 hours and 30 minutes a night.
16:24This is their average bedtime and wake time.
16:26Their average light sleep,
16:27deep sleep, and REM sleep.
16:31You can see the power of large data blocks,
16:34right? So this is useless data,
16:36But it's interesting, right?
16:37So this is Jawbone published at this point,
16:40when there was an earthquake in the Bay Area,
16:43and you can kind of see how what happens to
16:46each of the users in their different area.
16:48The earthquake was kind of based
16:50out of that Napa Sonoma area.
16:52You can see the biggest spike in wakefulness
16:55there and the further away you got from
16:58from this epicenter of the earthquake
17:00could see fewer and fewer people woke up.
17:03And so this is not surprising, right?
17:05Earthquake happens.
17:06Some number of people are actually going
17:09to wake up from it and again shows
17:11you the fact that all this data is is
17:14effectively somewhat useful potentially.
17:15But right now it's not necessary.
17:17Being used in that way.
17:20So Fitbit continues to evolve overtime.
17:22You saw kind of some of the basic data
17:24forms that I've shown you in clinic today.
17:26You're seeing things that look much
17:28more like this in the upper right
17:30corner where there is REM sleep
17:32and light sleep and deep sleep.
17:34For better or worse,
17:35in patients will come in and have these
17:37discussions about you know how right or not,
17:39right?
17:39These things are,
17:40and I will tell them that it's hard
17:42to know what to make of the data from
17:44the Fitbit because it hasn't been
17:46well compared to Poly sonography,
17:47but it's worth saying that it's
17:49important to recognize,
17:50you know when the points of start
17:51and end arc as if nothing else.
17:53You can get a sense better than sleep diary,
17:56perhaps of when there's sort of
17:57trying to get to bed and when
17:59they're trying to get up,
18:00and that is a valuable piece of information,
18:03particularly people who are
18:04tracking their sleep regularly.
18:08Now what's new in your Fitbit is
18:11oxygen measurement and this has
18:13been known to be coming for awhile.
18:15And it's not 100% clear how
18:17this fits in with sleep apnea.
18:19Yet, as presently Academy I went
18:21to meet with the Fitbit folks
18:23and I will tell you I was. Um?
18:25Surprised and amazed at how far
18:27ahead where they thought they were.
18:29They were that they had already
18:31been planning for.
18:32So how to think about sleep apnea and
18:34how to think about oxygen measurement?
18:36Then there really their quest to get
18:38their users to be as healthy as possible.
18:42And so this sort of oxygen measurement
18:43was the next step in their process,
18:45and you can see that they have what
18:48they are considering oxygen variation.
18:50So small variations is consistency.
18:53Big variations is inconsistency.
18:54And this is the kind of chart
18:57that people are getting.
18:59Now when you look at what
19:01a Fitbit actually says,
19:02it says Nope,
19:03your Fitbit device is not intended
19:04for medical purposes and consult your
19:06health care professional about any
19:08questions or health issues you may have.
19:10So as of yet they have not crossed the
19:12line into being a diagnostic device.
19:14They are still in any assessment mode.
19:18And so they answer some questions
19:20about oxygen testing, but they say,
19:22broadly speaking as a practical,
19:24met practical way of thinking about this.
19:28You get oxygen to reports but
19:30but how does it get reported?
19:33What does that variability mean? How much?
19:36How much variability is abnormal?
19:38And what does that mean in terms
19:40of how likely you are to actually
19:42need to test someone who comes in
19:45with a variable oxygen report?
19:46And then what happens as you think
19:48about this in the longer term, right?
19:50So night after night they're
19:51having these reports?
19:52What if those variable report those
19:53variability reports are actually variable?
19:55That some nights are better
19:56and some nights are worse?
19:57And what do you make of that
19:59right that the challenge?
20:00Here is that there's increasing
20:02amounts of data,
20:03but not very easy ways of taking that
20:04data into a useful reporting structure.
20:07And that was part of what we
20:09actually talked to them about,
20:10is it would be helpful to have
20:13some a way of assessing that.
20:15And affectively,
20:16how does this variability compare?
20:18In fact, to sleep apnea, right?
20:19They know,
20:20probably on the back end a
20:22little bit about what they,
20:24what their oxygen devices showing they
20:26don't necessarily at this point know
20:28whether it results in sleep apnea or not,
20:30but at some point I suspect
20:32that will be made more clear.
20:34I suspect research is going on
20:36to that level at this point.
20:38But these devices are currently in
20:40what I would consider a pre assessment
20:42mode right there, they're not.
20:43They don't want to be the
20:45diagnostic device as of yet,
20:47but they want to raise the question
20:49they want people to think about
20:50getting the test done that is going
20:52to be more conclusive,
20:54and so they've wanted.
20:55At least they went when we met with them.
20:58As the ASM said, hey,
20:59we want to work with you guys.
21:01We don't want to replace you guys.
21:07I showed you a bunch of data from 2015.
21:09Let me show you a little
21:10bit of the more recent data.
21:12So this is a meta analysis done
21:16in 2019 and looking at sort
21:19of Fitbits of various models.
21:23And more or less what they said
21:24is that the Fitbit model showed
21:26higher sensitivity and specificity
21:27and detecting sleep than the
21:29quote non sleep staging model.
21:30So the more recent ones are the ones
21:33that will show you the rehmann light
21:35sleep and deep sleep as opposed to the
21:37ones that I showed earlier that are
21:39only soaring sort of sleep and restless time.
21:43And so there was what they said in this
21:45article is promising performance definition,
21:48wake from sleep and their
21:50convenient and economical.
21:51But there are limited specificity
21:53and not a substitute for PSG.
21:56Which is of course not terribly surprising.
21:59And so here's something that just
22:01came out late last year looking at
22:03Fitbit Charge Two and Fitbit Alta HR,
22:06which are again fairly recent models
22:08recognizing they're not the most recent.
22:10But again,
22:11they showed acceptable sensitivity
22:12but poor specificity,
22:13and so they are not accurate enough
22:16from a clinical perspective to
22:18replace what we do in the laboratory.
22:21I'm not sure that they want to replace
22:23what we do in the laboratory as of yet,
22:25but I will say I suspect some of those
22:27things may be coming down the line.
22:30So as one of my favorite movies,
22:33they quote 60% of the time.
22:35It works every time, right?
22:37That's that's kind of where
22:38Fitbit is at currently.
22:40But I think they continue to
22:43refine and improve and it will
22:45be better than this overtime.
22:48So the world continues to evolve, right?
22:50They? It's not just Fitbit, right?
22:52Fitbit is part of this?
22:54And why is Fitbit part of this?
22:56Will Fitbit isn't just Fitbit anymore?
22:58Fitbit is now Google and so how do you?
23:01Why does Google by Fitbit?
23:03Well, I think you have a sense right?
23:06Here's millions of people with millions
23:08upon millions of nights of data.
23:10And what does Google do really?
23:12Well they manage data and so
23:14clearly you can see where these
23:16interconnections are going to continue.
23:19But Google Plus Fitbit is going in
23:22going to create some interesting
23:24data streams I would imagine.
23:26But Google and Fitbit are not the
23:28only people in the marketplace right?
23:31There are.
23:31Garman there's the band,
23:33the subscription model for many
23:34the athletes are using these days.
23:36Then of course the Apple Watch which
23:39is actually stayed on the sideline of
23:41those sleep world for the most part,
23:43but is clearly increasing their
23:46interest in that area based
23:48on what we can see so.
23:50These devices.
23:59You're muted, Doug.
24:02Somehow you got muted. That's
24:04interesting. I don't know when I got muted,
24:06I didn't do anything,
24:07but somebody muted me.
24:09OK, they didn't like what I was saying.
24:11Maybe it's Google. Sure would.
24:15Some point after Rupan so the Apple Watch.
24:19I will say they had stayed on
24:21the sidelines for the most part.
24:23On the sleep world, for the most part,
24:26but they're slowly entering
24:27this marketplace as well.
24:29They also have oxygen monitoring
24:31capabilities on the newer Apple Watch
24:33is so recognizing that they're going
24:35to enter this this space as well that
24:37the players are beginning to change,
24:39but the outcomes aren't changing
24:41and the technology.
24:43Of the devices is going to change
24:45faster than what we can keep up with.
24:48OK, let's see now that I'm.
24:50OK, so let me then shift a little
24:52bit and erables so these are
24:54devices like the res Med S plus.
24:56Now this device was an older device,
24:58it was fancy and expensive backrow
25:00days and then it got less expensive
25:02because people weren't using it.
25:03'cause it seems sort of bulky
25:05and complicated to have a device
25:06that's near your bed as opposed
25:08to something you could wear.
25:09But this is actually the data
25:11that I had from this device.
25:13They actually loan me one to try
25:15out at a time and you can kind of
25:17see you would get a score and you
25:19get what your mind would charge.
25:21Your body would charge and I
25:23told him that this was.
25:25A fairly a bunch of who high and
25:28there really wasn't a good reason
25:30to give me a score of some kind,
25:32but they felt it was important and
25:34now you can see that most people
25:36in some way or getting a score what
25:39that score means is hard to assess,
25:41but it is something that is.
25:44Element that patients are following
25:46this and you can kind of see
25:48the data you would get off.
25:49And S plus again a hypnogram similar to that.
25:52And this is an amazing way something
25:54that's not even touching you,
25:56right?
25:56It's able to stage your sleep based
25:58on what it can actually assess by
26:00being near you and also provide
26:02information on light and temperature.
26:05And you can kind of see where
26:06you get your score from,
26:08but it's some combination of,
26:09well, how much were you disrupted
26:10and how much you know how?
26:12Like how much light was in there,
26:13and you know what is actually
26:15not light light sleep.
26:16How much sleep did you have and
26:17how much deep sleep did you have,
26:19etc.
26:22But it's beyond that right that the
26:24beds can be trackers to sleep, IQ,
26:26sleep number beds have data that kicked out,
26:28and so this is information from one
26:30of my colleagues is asleep doctor.
26:32She was willing to share some of
26:34the data for her again looking at
26:36her breath rate, her heart rate.
26:38You know what kind of restful sleep she had.
26:41I will say at least she was getting.
26:44You know she was in bed for a good
26:4710 hours almost, which is great.
26:49You know some of that time.
26:51Clearly she's not sleeping,
26:53but but in the end she was
26:56practicing what she was preaching.
26:58But you can track it for your kids too.
27:00For those people who really want to
27:02know what's going on with their kids.
27:04Beyond the miracles,
27:06there are obviously the apps,
27:08and so in terms of the apps.
27:13There are obviously sleep apps galore.
27:17Mobile health apps in general.
27:19You know they can promote Wellness
27:21relatively inexpensively and maybe help
27:22with management of chronic diseases.
27:24There are bewildering number of
27:25apps that are available and it's
27:28difficult to see which ones are the
27:29safest or most effective and at
27:31least according to the JAMA article.
27:33It would be nice to have some sort of
27:36unbiased review and certification process,
27:38but that has yet to occur.
27:42So what kind of apps are
27:43out there regarding sleep?
27:45I thought I would cover a couple of them.
27:47This is one that was built in combination
27:49with the VA and Stanford which is CBT.
27:51I coach I still in fact use this in
27:53my clinic for some of my patients,
27:55'cause it's free and it has some things
27:58that are useful in terms of tools.
28:00But you can see it has things like
28:03a progressive muscle relaxation
28:05or or wind down exercise.
28:07It has tools for sleep Diaries
28:10and ability to chart some things.
28:12It's a useful tool and and best of all,
28:16it's free.
28:18There are circadian applications,
28:19like for instance in train,
28:21which you can gather a whole bunch
28:23of information about in terms of
28:25people who are traveling, right?
28:27This was a particularly useful thing
28:29if you're going to Tokyo in a world
28:31in which we are in covid may be less
28:34useful 'cause we're not traveling very much,
28:36but the idea is that you can use
28:38some of the information we know
28:40scientifically and put it out in
28:42a way to make patients ability to
28:44travel somewhat easier, right?
28:46When should you get bright light?
28:47When should you get dark,
28:49and how can you prep?
28:51Your trip to make your trip is
28:53as helpful as possible and make
28:55your sleep as good as possible.
28:57And this is one of the classic apps.
29:00The Sleep cycle app.
29:01This is one that you could stick under
29:03the sort of sheet and it would kind
29:05of give you a report of your sleep
29:08depth at some basic level and you
29:09can see this is some what we would
29:12see is a fairly normal looking cycle,
29:14right?
29:14That somebody is in deep sleep
29:15and then in lighter sleep and deep
29:17sleep in longer sleep in cycles of
29:19somewhere between 90 and 120 minutes.
29:24Where this starts to get more interesting
29:26is in the fact that there have been a
29:30number of applications used for snoring.
29:32And there was an article published
29:34in 2016 looking at several of these
29:37apps overtime and effectively showed
29:39that there was excellent positive
29:41predictive value for snoring detection
29:43in the in the populations used.
29:47So here's an example.
29:48The smart alarm that I just showed you
29:50also has an ability to track snoring,
29:51and you could actually listen to those
29:53snores that pop pop up during the night.
29:55All those little circles or
29:57periods where you could listen.
29:59Another app called quits.
30:00Morning where you can see it was in this
30:03particular case compared to Poly sonography,
30:05and so you can see kind of the how
30:07they synced based on time the arrows.
30:10Anne and.
30:13Showed effectively in this particular app,
30:16again, the positive predictive value
30:18of these three patients at versus PSG.
30:21Was upwards of somewhere between 93 and 96%,
30:23which is really pretty good.
30:25And why is that?
30:26Well, because snoring is a pretty easy
30:28thing to track for the most part.
30:29It may be difficult if there's
30:31more than one person there, right?
30:32So remember,
30:33if you have a bed partner,
30:35you may be picking up on snoring of
30:37two people, not just one person,
30:38but broadly speaking,
30:39if you're sleeping alone,
30:40it should have a pretty good chance
30:43of picking up snoring and giving
30:45you a sense of how bad it is.
30:47Snore Lab is another app that
30:49has been used and I'll show you
30:51why I highlighted this one.
30:53This is a patient of mine.
30:56Who was an MIT engineer?
30:57And of course you know had all
30:59sorts of charting here, but he had,
31:01you know,
31:02tracked his snoring when he used
31:03nothing in the top left and then with
31:06his old kind of mouthpiece that he
31:08had built for him on the top right
31:10and then with CPAP on the bottom.
31:11And he did a bunch of calculations and
31:14you can even see where he marks off
31:16that he took where he C Pap was on and wear.
31:19His see.
31:19Pap was often kind of noises that were made.
31:23And so, broadly, you can see how these
31:25apps can be used to assess progress, right?
31:28We always talk about how, for instance,
31:30and use of dental devices.
31:31You know it's all kind of a little
31:34bit it hit or miss, right?
31:36You know, is the snoring better?
31:37Is it not?
31:38Better?
31:39Was the bed partner say,
31:40well,
31:40you know,
31:41would using an app in some basic
31:43level be useful in terms of tracking
31:45whether a mouthpiece is or is
31:47not effective and how effective
31:49is for controlling aspiring?
31:50Well, sure,
31:51something like this could be done.
31:53And and it's not an unreasonable
31:55approach for some of these things.
31:57It doesn't obviate the need for follow up.
32:00Testing at some point to ensure
32:01that sleep apnea is controlled,
32:03but if you're trying to titrate somebody,
32:05this is not an unreasonable
32:06way to consider it.
32:09So moving beyond snoring,
32:11then can you assess sleep
32:12apnea via phones and apps?
32:14And so this was sort of the original
32:16study that I could find using a
32:19phone and there actually strapped
32:21the phone to somebody's chest,
32:22which wasn't obviously going
32:24to be a long term solution,
32:26but more or less with this group showed
32:28in 2014 was that in fact they could
32:31show with reasonable sensitivity and
32:33specificity that the the RDI was decent
32:35for picking up and hi greater than 15.
32:39But it moves beyond this, right?
32:41So the University of Washington there
32:43was an app which I'll show you here.
32:46Where they actually had a
32:48again an app where they didn't.
32:51In fact, having the phone touching
32:53anybody and so they in a contactless way,
32:56was able to identify respiratory events
32:59when compared that versus Poly sonography.
33:02And so this technology ended
33:04up in the durable.
33:06The Sleep score Max.
33:09And so you know you can see the evolution
33:11of the S plus into this sleep score,
33:14where now not just to you
33:16assessing sleep quality which was
33:18effectively with the S plus did.
33:20But now you are able to assess
33:22to some level you know.
33:24The grieving someone might have.
33:28Now they have done a lot of work
33:30with this particular device,
33:31however,
33:32I haven't seen a lot of the
33:35data percolate out of this.
33:36And then there's apps like this,
33:38which is sleep check which hasn't
33:41yet hit the American market.
33:43But is it is in Europe and this is
33:46a device that was able to identify
33:50OSA with a sensitivity of 85%.
33:53For mild OSA,
33:5483% for moderately and 83% for
33:57severe OSA compared to full but
34:00unattended Poly sonography?
34:02And in the patients home,
34:03and so you can see,
34:05you know this is the wave of
34:06where this is moving now,
34:08whether they're going to be truly
34:10diagnostic or pre pre assessments
34:11before people are actually
34:12willing to go and spend the money
34:14to get formally diagnosed.
34:15But this is the stuff that is
34:17coming right that we are going to
34:19move from a world in which they
34:21are just assessing sleep to
34:23assessing sleep and breathing.
34:24And I think we have to be prepared for.
34:29So you gotta prepare yourself for this world.
34:31The world is going to change.
34:33Continue to change in front of us.
34:35We need to be ready for that. So.
34:39Is all this technology good or is it bad?
34:42I think in the end. It's neither.
34:48That adaptation and evolution of
34:50us of the practitioners of Sleep
34:54Medicine is going to be a key
34:57aspect of what we need to do.
34:59My belief is that there are.
35:04Numerous patients out there
35:07that are undiagnosed.
35:09Who have a sleep problem and this
35:11may be way this technology to get
35:14them to take that next step to come
35:18to a Sleep Medicine practitioner.
35:20The problem is,
35:21there's not a lot of clinical teams right.
35:24The number of sleep practitioners is
35:26growing at a fairly slow rate and retirement.
35:30Of sleep doctors and clinicians
35:32is a also stable rate,
35:35which means that we're not going to
35:39be increasing our team size with.
35:42Board certified sleep doctors.
35:44Very quickly,
35:45the world of Aips nurse practitioners
35:47NPS is going to grow rapidly
35:49and nurses and artizen are PSG.
35:51Tees are all going to be increasing
35:54as we look at the at the sleep team,
35:57but still there are not going to
36:00be a lot of growth in the world
36:03doctors and so we need a way of
36:06assessing and managing these patients.
36:08And where are these devices going to go?
36:10Is it going to be identification
36:12or is it and then send the to the
36:14clinical team for a formal diagnosis
36:15or they're actually going to try
36:17and diagnose them at home?
36:20And without the care of a local doctor right,
36:24you can see this with some things
36:27like the hair.
36:30The world of hair growth or erectile
36:32dysfunction, where there's all done
36:34via these sort of apps and monitoring
36:36right that they monitor you?
36:38They connected with the doctor
36:40for a quick consultation and and
36:42then you get your treatment right.
36:44The medicine is delivered to your door.
36:46Well the same could be true. Firstly,
36:50back again and it and it may come to that.
36:52We're not there yet.
36:54Is technology going to hurt the sleep lads?
36:57Well, maybe maybe not.
36:58Depends on how many more patients
36:59come in versus how many patients
37:01do you lose because they're
37:03being diagnosed somewhere else?
37:04And I think that that's hard to predict,
37:06but in the short term,
37:08I think more patients come in
37:10then were before because they're
37:11looking at their sleep.
37:12Please,
37:13I think that in some cases there may
37:15be too many patients right that wait
37:17times will continue to increase if
37:19too many patients are identified and
37:21you've got to be able to manage that,
37:23that comes to that Internet
37:26and clinical team question.
37:27And what about patients who have
37:29incorrect diagnosis right there,
37:30identified on technology that is black box?
37:32Or that you're not sure about,
37:34you know, is there a problem there?
37:36Is there something we have to worry about?
37:38What if patients are told that
37:40they don't have sleep apnea,
37:41when in fact not only do they have
37:43sleep apnea, but it's bad sleep apnea?
37:45How comfortable can we be with
37:47some of the things that are going
37:50on behind the scenes?
37:51And then Lastly treatment paradigm, right?
37:53So it's not just about diagnosis and
37:55then this is kind of where I'm going to.
37:58I'm going to close the talk a little bit,
38:00but it's recognition that the diagnosis
38:02of sleep apnea or diagnosis of a sleep
38:05problem is not the only thing we do.
38:07It's about the long term management
38:09of the patient,
38:10and I think that that's really
38:11crucial to recognize.
38:15So what are the ASM juice the
38:18ASM brought together a consumer
38:20sleep technology task force.
38:23They came out with the position
38:26statement in 2018 and largely
38:30said that these applications.
38:33Purport to measure an any,
38:35maybe even improve sleep and that we,
38:37as you probably all are aware
38:38we encountered this consumer
38:40sleep technology in practice.
38:41And though there's not
38:42always validation data,
38:43you gotta understand a little bit about
38:46what these devices are and what they do.
38:49You need to sort of recognize that
38:51many of them are not validated.
38:53Many of them don't have FDA clearance,
38:56and you really shouldn't use them at
38:58this point for diagnosis and treatment
39:00of sleep disorders at the at the time,
39:02but this may change in coming
39:04years would be my guess,
39:06but I think if it talking about
39:08that data when somebody says hey,
39:10do you want to see my Fitbit data?
39:13I almost never will say no.
39:15I think it's recognition that hey,
39:17they're bringing another data point.
39:18How useful is it? Hard to know.
39:21On the other hand, saying no to that.
39:26Doesn't invalidate the patients
39:27feeling about that data,
39:28and so it's worth opening that data
39:31looking through their screen and talking
39:33to them for a minute or two about that data.
39:37And so you can kind of see
39:38where the symptoms come down,
39:39which is to say, hey,
39:41the more we get validation,
39:42the more we get raw data and
39:44understanding algorithms the better.
39:45We're going to be comfortable
39:46with how these things go.
39:48Yes,
39:49I'm also created a asleep technology
39:51section on their website and part of
39:54what this task force did besides coming
39:57up with the position statement that you.
40:00See is they also built and have
40:03continued to build out a database.
40:06And so in the sleep technology,
40:07if you were an ASM member,
40:09you can log in and look at it.
40:11For instance,
40:12you can look up the outer ring
40:15and so this is a ring that is.
40:17Purports to monitor things.
40:18It's obviously a wearable
40:20'cause it's on your finger.
40:21It's small and it's easy to
40:23sort of have on through the
40:25night for at least some people,
40:27and you can kind of see.
40:30All the sort of tags of data
40:32that are compared here.
40:33You can sort of see that there was
40:35compared to PSG that you can't look at
40:37the raw data and it is not FDA cleared,
40:40but the idea here is that there
40:42are summaries of all these data.
40:43So if you have a question or
40:45if a patient brings in a device
40:47you're not familiar with,
40:48you can quickly look up to see if
40:50it's in the ASM sort of database.
40:52This is not a.
40:55There is no.
40:56The SM is trying not to give any credence
40:59to anyone device more than another.
41:01It's just trying to gather the data
41:04in a way that makes it easy for
41:06you to track and so for instance,
41:09here is the data around the
41:11aura ring from 2019
41:13and you can again see how
41:15PSG versus the ring kind of.
41:17They're fairly similar.
41:18They're not exactly the same,
41:20but you can see again that there is
41:22fairly good correlation with sleep,
41:24but a lot of source scatters, right?
41:27So the scatter plot is certainly higher.
41:29I think for the ordering that it is for PSG,
41:33and I think that's part of the challenge.
41:35With this is.
41:36Yeah, it looks good in some cases,
41:39but it's still not meeting up to
41:41the gold standard necessarily.
41:45And so I thought,
41:46I closed here just sort of saying,
41:49you know, this data is still coming.
41:51This is an article that has not yet been
41:54formally published and has been accepted
41:56for print and sleep for this year.
41:59And you can see again now, four different
42:02wearables and three non wearables.
42:04Where they were testing them
42:07versus Actigraphy and PSG and you
42:10can kind of see where there is.
42:12Again some good sensitivity,
42:14but problems with specificity and the
42:16sleep stage comparisons are mixed,
42:19and the worst the sleep is,
42:21the worst the devices do,
42:23and I think it's worth recognizing.
42:27That this is the problem, right?
42:29That our population is not a normal sleep.
42:32When you see a lot of the validation
42:34data is often comparing normal PSG.
42:37It's me patients who are normal
42:39between PSG and the wearable or
42:41niarbyl that that's under study,
42:43but we recognize we don't see
42:45normal people in our clinic.
42:47We see people who are abnormal
42:50and that's really the issue.
42:53So how am I going to close?
42:55I'm going to close by giving
42:57you sort of the tag line to the
42:59editorial I wrote in JC SM.
43:01Just a short period of time ago,
43:04but to say to seize this opportunity
43:05Miss Re Orient are thinking expanding
43:07beyond sleep apnea detection to
43:09fully embrace the holistic importance
43:11of sleep for disease prevention
43:13and management productivity and
43:14satisfaction and safety and well being.
43:16And health system that demands
43:18greater efficiency,
43:18we need to redefine the problem instead
43:21of asking how we can get more people.
43:23Into the Sleep laboratory we determine
43:26how to manage millions of people
43:28with undiagnosed sleep apnea through
43:30collaborative care models with.
43:32Primary care doctors,
43:33other specialists and our own sleep team.
43:37And finally we need to redirect our
43:39field by leveraging the disruptive
43:40technology like the consumer
43:42technology we're talking about today,
43:44to help us improve patient access to
43:46sleep care and improve their health,
43:48and by refraining Sleep Medicine,
43:50we can provide greater value payers,
43:52patients and employers.
43:55So.
43:55In summary, obviously the evaluation of
43:57sleep is evolved significantly overtime.
44:00It's not unique to Sleep Medicine,
44:02pathology, radiology,
44:02other fields are going
44:04through this change as well.
44:05It's just that the consumer side.
44:07This is been particularly notable
44:09in sleep just because so many people
44:11are interested in your sleep,
44:13and that's great.
44:14It means the word has gotten
44:16out that sleep is important.
44:18On the other hand,
44:19we have to deal with all this consumer
44:22technology as patients bring it in.
44:25Number two that the scientific data around
44:28consumer sleep technology is growing rapidly,
44:30though consumer technology is still not gold,
44:33standard right is validated
44:34against PSG in normal sleepers,
44:36really that same population would care about,
44:39probably not,
44:40and so it's really validating against PSG,
44:42and patients who have disease,
44:44that,
44:45to me would seem to be what eventually
44:48we need to get to and then three.
44:51We,
44:51as clinicians need to understand,
44:53adapt and evolve to this incoming technology.
44:56Because it's only going to get to be a
44:59bigger piece of what we deal with, right?
45:01I didn't even get into things like my air,
45:04right where patients are looking
45:05at their own CPAP data and giving
45:07getting some feedback from it.
45:09Is that disruptive technology?
45:10No, it's actually great.
45:11Technology is helping our patients
45:13be more successful and it's adding
45:15to things that we already do,
45:16but this is the world in which
45:18we are going to see continued and
45:20constant growth and staying on top
45:22of it's going to be difficult.
45:24I know the ASM is working to
45:26try and keep people.
45:28Educated as much as they can about
45:30devices as they come out and apps and
45:32all the things that are out there.
45:33But even for them,
45:35it's going to be difficult,
45:36and I think it's important to recognize
45:39if we just shut this all down.
45:41Say, you know I'm not dealing
45:43with any of that.
45:44I think that that's going to lead
45:46you down the path where you're
45:48eventually going to be left behind.
45:50That these folks who do technology
45:52work are very smart and they are
45:54moving much more rapidly than I think
45:56we feel sometimes comfortable with.
45:58But eventually we'll get to a
46:00place where I think we will meet
46:02in a both comfort and use.
46:04Perspective,
46:04so with that I wanted to thank
46:07everybody for their time and attention.
46:09I'm happy to take questions
46:10at this point and if anybody
46:12has questions after the fact,
46:14my emails on the bottom of the slide.
46:16So thank you.
46:19Thank you John drive.
46:21That really was very timely appreciated.
46:23I'm gonna ask people if they have questions.
46:26Just make some sort of notation
46:29in the chat box and will call you.
46:32I'll start you off with a multipart
46:35question given your knowledge of
46:37and working with the organization.
46:39Are there any current
46:41collaborations between the Academy,
46:42an industry and what that looks like?
46:45What are the opportunities for the
46:48people here to collaborate with?
46:50Industry and is the Academy providing any
46:53structure slash funding for any of them.
46:57So that's a that's a really wide question,
47:00Larry, but but thank you for making it.
47:02I'm obviously less well connected that I
47:04once was as I've rotated off the board.
47:06But I will say my belief is that The
47:09Academy Is still working towards.
47:11A partnership in which there is clarity on
47:14how how the data the technology can be used.
47:17Right so I think the challenge that
47:19the Academy has these technology
47:21companies like Fitbit or Google
47:23or really large and we asleep.
47:25Doctors are not a very large group.
47:27They are interested in working with us
47:30but sometimes on their terms more than
47:32our terms and I think that that is a
47:35challenge that we will continue to face.
47:38I think it is true also of things like AI,
47:41sleep staging.
47:42Right that that there is,
47:44the Academy has some data on sleep scoring
47:46that AI companies would probably like
47:48to have to use this comparison points.
47:51So I think that there are partnerships
47:54that are being worked on and I
47:56think that there is interest from
47:58the research perspective.
48:00I think if you look at the foundation
48:02they're always looking for technology they
48:04don't want people to be trying to promote
48:07their own technology with foundation funds.
48:10But I think that they are
48:12looking for research that is.
48:14Interesting within that
48:15particular technology world,
48:16so I think the Academy recognizes
48:19that the challenges are out there
48:21and I think they're really open
48:23to ways trying to work together.
48:26But at the same time want to protect the
48:29sleep clinician as much as possible.
48:32Recognizing that you know we don't
48:34want to be replaced by the robots.
48:39And then.
48:42We have other questions. Yeah, can I
48:45ask Doug a question or make a comment
48:48so there are millions of people
48:51that are using oximeters.
48:54At home, and it turns out that the ox,
48:58the standards in order to
49:00sell an oximeter do not take
49:02into account different pigment types.
49:05And so there's a huge as a matter
49:09of fact that couple of senators
49:11Elizabeth Warren and Cory Booker
49:13have asked the FDA to look into
49:16this issue because there has not been any
49:19validation of these devices on people
49:21with different skin colors. So this
49:24is, you know there are millions of these
49:27out there an we just need to
49:29keep that in mind that a lot of these devices
49:32have really not been validated.
49:34I think that's an excellent point.
49:36I mean, I think they certainly oximeters
49:38have blown up in the setting of kovid, right?
49:40That there's more and more of them
49:42being sold, which interesting,
49:43of course, is that patients will also
49:45use that in the clinical sense to say,
49:47hey, I don't have sleep apnea,
49:49'cause every time I check my oximeter,
49:51it's normal, so I couldn't have sleep apnea,
49:53to which I say yes, but you're awake.
49:57And to your point around validation,
49:58I think the issue is,
49:59you know there's been a.
50:00Real push to get these things out
50:03because people wanted to have them.
50:05And yet at the same time knowing how much
50:08variability there may be and whether
50:10skin type or pigmentation or whatever, right?
50:13That variance among population
50:14really does matter,
50:15and that's true for the oximeters,
50:17but it's also true for all these wearables.
50:20Yeah, now I think that the wearables
50:22obviously they have tried to gather,
50:24you know, for instance,
50:26Fitbit has so many users that they
50:28probably have looked at some variances.
50:30Around skin color if they have
50:33access to that.
50:34But again,
50:34who knows because that data is not always
50:37being published in a scientific way.
50:40It's in their back rooms
50:41being discussed amongst them,
50:43which is a challenge.
50:51River question from Karen Johnson.
50:53You wanna unmute and ask your question?
50:57So I just want your thoughts on sort of,
50:59you know when we think about
51:01public health issues like daylight
51:02savings time or school start times,
51:04what can we get from you know Fitbit
51:07and Google and where do you think
51:09it'll sort of help answer some of
51:11these sort of unanswerable questions
51:12about the effects on sleep and
51:14potentially sleep disparities and such.
51:16Well, I think that's a great question.
51:18I think the real issue is will we
51:20have access to that data, right?
51:23I think.
51:24The number of people who have a wearable
51:27device is growing continues to grow.
51:30As these devices get more
51:32and more affordable.
51:34The companies will have more
51:35and more data points right?
51:37And so just like I showed
51:38you with an earthquake,
51:40you know it's not just that right,
51:42it's what is your sleep look like
51:44the day after some big event, right?
51:46You know the inauguration, you know,
51:48for whether you are for or against the
51:50current President you know was that you know?
51:53Was the sleep better or worse, right?
51:55Is it on a school day,
51:57all 17 year olds?
51:58How much sleep are they short
52:00and how much does it matter?
52:01And how much can we involve
52:03ourselves in that right?
52:04I think.
52:05This data is not as of yet publicly
52:09available, and if it became so,
52:11I think it would be a goldmine of research,
52:15but as it currently stands,
52:17it's all behind firewalls and I think.
52:21The more that the Academy and
52:23these technology companies are
52:24able to work together, the better.
52:26I think we chance we have of
52:28being able to use that data for
52:30public health initiatives.
52:32I think it would be lovely.
52:33I just don't know that we can
52:36plan on having that access.
52:44I was wondering if I could ask a question.
52:47This is under his truck over from Yell,
52:49really nice talk.
52:50Thanks for talking about a topic
52:52that is very actual for patients
52:53and and a little bit distant for us.
52:56So I like your point of acknowledging
52:58that you know we should be looking
53:00at these devices and looking at the
53:02data that patients are presenting.
53:04And I'm faced with that everyday.
53:05I've I've seen patients bring in every
53:07single app report that you have shown
53:09in others and so I don't curiosity.
53:12You know what?
53:13Which one of these apps?
53:14And do you actually use and how
53:16do you use them?
53:18So I mean, I've used nor lab for, you know,
53:21titration off the back on the back and
53:23maybe oral appliance stuff as well.
53:25I think I've found a little bit more
53:27difficult to find devices that measure
53:29oxygen reliably and also devices that
53:31measure position reliably or not.
53:33Devices, apps,
53:33I should say the measure position reliably,
53:35so was wondering what what do you do
53:38with information that comes in from
53:39the sleep cycle or it comes in from
53:42another particular like circle for example?
53:44Right,
53:44so I think I think that's an
53:47excellent question and I think I
53:49think we're all faced with this.
53:50Is is the challenge of how do we use this
53:53data and what do we use in our own clinics?
53:57So what I'll tell you is,
53:58at least in, at least in my clinic,
54:01I don't tend to use much more than
54:03sort of generics, which is to say,
54:06hey,
54:06if you download an app such as quit snoring,
54:09or snore lab,
54:09it might tell us how bad your snoring
54:12is with that custom mouthpiece or
54:14without the custom mouthpiece.
54:15And I will say the number of times
54:17I get that feedback back from the
54:19patient with that recommendation is small,
54:21right?
54:21But sometimes they just never come
54:23back 'cause they're tracking their own data.
54:24Sometimes they come back and
54:26they never did it.
54:28Similarly, I think one of the things that we.
54:32We may have use for is the
54:35patients and I have seen several
54:37of these recently who are really.
54:40Not well scheduled and we don't
54:42have a great sense of what that
54:44schedule for sleeping looks like,
54:45so I've had several these recently where
54:47I've made a recommendation to say look.
54:49You know I'd love to get an actigraph on you,
54:52but the wait time for our actigraph is
54:55long and ones out 'cause it's broken.
54:57And so why don't you buy yourself
54:59a device of some kind and actually
55:01look at schedule?
55:02I don't even care what the sleep
55:04quality looks like,
55:05it can be the cheapest non
55:07sleep staging Fitbit.
55:07But I just need to know when you're
55:10going to bed and getting up every day.
55:12And to me,
55:13the most valuable information
55:14I can get from these devices is
55:16really that is how well scheduled
55:18are these 'cause the people
55:19for whom they are bringing
55:20these devices?
55:21A lot of the time and want to talk about it?
55:24Or the people who aren't sleeping well.
55:27And a lot of the time that understanding
55:30that schedule to me becomes so crucial
55:32and suspected is for you as well.
55:35And you know, even then I can
55:37scan through that data and take a
55:39couple of screenshots and upload
55:41it to their medical file, but.
55:43I have no good way of summarizing
55:46that data in a one page snapshot of.
55:48Well, here's their bedtime and wait
55:50time average and their variances.
55:51And wouldn't that be really nice and useful?
55:54And that's kind of what we told Fitbit
55:56when we went to meet with them.
55:58We have yet to see that.
56:03Thank you. I think another opportunity
56:06is for respironics and res Med to put
56:09in some smaller emitters in the masks
56:10so we have a sense of what's happening
56:12with the patients and rash leaping in,
56:14and I think that there is a
56:16wealth of options out there.
56:18There is no question and I think I
56:19will say I think technology companies
56:21some of them are more willing
56:23to listen to other than others.
56:25Some of them are just trying to provide
56:26the best consumer experience they have.
56:28Some of them are just kicking out scores
56:31that have no relevance to anything.
56:33The ones that are at least willing
56:36to collaborate a little bit and
56:38sit down in some way as much as
56:40a big company can sit down,
56:42but you know there was a sleep scientist
56:44who spoke at a from Fitbit who spoke
56:47at an Academy American Accounting
56:48medicine conference like to me,
56:50those are the small steps and what
56:53they can release is small and they
56:55can talk about the algorithm and they
56:57have NDA's and all these things,
56:59but it shows at least that they're
57:01willing to take step step to meet.
57:04Somewhere in the middle we may never
57:05be thrilled with having an algorithm
57:07determine what we see about somebody sleep.
57:09At the same time, I'm using a watch
57:11path that does the exact same thing.
57:13How comfortable might with that?
57:15OK, 'cause there's a lot of data behind it,
57:17but.
57:19Probably the Fitbit has more data,
57:20we just don't know what it looks like, right?
57:22So the more that there is openness
57:25about the data, the better.
57:27I think we care about,
57:28but Fitbit as a general rule,
57:30doesn't they want best health but their
57:33company they want to sell things right?
57:35And so how?
57:36What is that validation do for them?
57:38Maybe not a lot.
57:41Because they're already selling
57:42millions of the devices,
57:43right?
57:46On that. On that note,
57:48I think we're getting close to the end here.
57:51Clearly this is something that
57:53we will be talking about a lot
57:56for the foreseeable future,
57:58so thank you for coming on.
58:00Starting our conversation on it and.
58:04Andrea, don't know there's anything
58:05you want to do to wrap up.
58:08I know I just wanted to thank
58:10everybody who participates in
58:11these the joint conferences.
58:12I think. So far it's been a great
58:14success and we're looking forward
58:15to more and more of these talks.
58:17And thank you Doug,
58:18for making the time for us and just
58:20for everybody else for next week
58:22will have another session of the
58:24yields the conference at 2:00 PM,
58:25so we'll see you then.
58:28Say very much everybody.
58:29Take care everybody.