Some years ago I was in the hospital for six weeks with a very ill child ( all is well ) and they were monitoring vitals every hour and later every four hours. It is a type of torture to be awoken to do this all the time. During that season I wanted to build a solution that would collect vitals and allow the patient to rest without disturbance and see it cheap enough to become ubiquitous. But when I looked into what has to be done to get medical devices certified I backed away.
Hospitals aren't optimizing for average health of patients. They're trying to get the best minimum performance, not the best overall performance.
Injuring most patients by depriving them of sleep for no good reason is felt to be justified by the chance of catching an extremely urgent problem in time to do something about it once or twice over the lifetime of the hospital.
An example of this with a lot of popular misunderstood awareness is the idea that you shouldn't sleep if you have a concussion. Preventing the patient from sleeping is detrimental to their health; it inhibits healing.
The reason hospitals do it is that their way of diagnosing whether the concussion patient has just experienced a catastrophic complication is to talk to them and see whether they are capable of producing normal language. This diagnostic mode is not available if the patient is asleep. So the patient isn't allowed to sleep.
It depends. Early detection of worsening conditions are absolutely vital for the safety of the patient. Vital signs monitoring is the simplest way to detect them. For example, respiratory rate increasing to over 20 per minute is a bad sign. This stuff is literally part of the criteria for diagnosing sepsis and septic shock. They will absolutely be checked if such conditions are suspected or expected.
I think realizing that respiratory rate has risen to over 20 does not require waking the patient though? It’ll generally be quite obvious if someone is breathing that fast, even when asleep.
But yeah, there’s probably different things that work the same way.
It's harder than it seems. Increased respiratory frequency can be subtle, it won't be full dyspnea every time. The signs can be missed if not explicitly looked for. Observing respirations requires a well lit room for the examination.
> But yeah, there’s probably different things that work the same way.
Sepsis diagnosis also includes assessment of mental state. Mental confusion is a symptom of sepsis. Glasgow Coma Scale is a component of Sequential Organ Failure Assessment scores: it includes talking to the patient and evaluating the coherence of verbal responses.
There are rate sensors you can put on a chest like an ekg lead. Though being covered in wires is also pretty awful. Or you can use a pen light and watch for rise and fall of chest. Also, Glasgow Coma Scale absolutely can be administered to someone who isn’t awake but this is really just me nit picking, your point about needing to be intrusive or waking people up is correct overall. Certain things are bothersome, some are painful, most of it will wage people up.
> Glasgow Coma Scale absolutely can be administered to someone who isn’t awake
Of course, it is a coma scale after all. Actually applying it to sleeping patients though? Will get you false positives at best: even healthy patients will fail to respond verbally (-4), blink their eyes (-1) and obey commands (-1) if they're asleep. That's a maximum GCS of 9 even if you stop there. If you don't stop there, you might just get punched in the face when they are rudely awakened to the sight of you testing their pain response.
I don't think being able to monitor my blood pressure at any random time as useful. As my doctor explained to me, the studies that determine what is the normal BP range obtain the measurements after the patient has been sitting in a relaxed environment for around 15 minutes. This is something I need to schedule once or twice a week to check that my hypertension is under control.
If I were to start taking measurements at random times, such as while walking or watching an action movie, the results would be misleading as they were not obtained under the same conditions as the gold standard measurements.
I would also question how accurate these new devices are. I used to use a wrist cuff BP monitor but one day I took it into the doctor's office and did a side by side comparison with my doctor's BP monitor. My wrist cuff BP monitor was reading about 20mmHg too low. I thought I had my hypertension under control but it wasn't. I've since replaced it with an arm cuff BP monitor. If I have to go to the effort of sitting and relaxing for 15 minutes to obtain a measurement, attaching an arm cuff is not much of an inconvenience.
I think exactly the opposite. It's because you should take a measurement after 15 minutes of relaxation that it's useful to have a wearable monitor. Those conditions may happen naturally throughout the day, and can be detected by your watch automatically (via continuous heart rate monitoring plus accelerometer readings), so blood pressure readings can be taken serendipitously, which is way more convenient.
Also I think in theory true continuous blood pressure monitoring data ought to be a lot more informative than just resting blood pressure. We may not yet have the research to really draw conclusions yet, but it should become possible to do that research relatively quickly if a whole lot of continuous blood pressure data becomes available.
I agree that having a continuous measurement is more powerful, but one confounding factor is the location of the pressure measurement. If the measurement location goes up or down relative to your heart then that produces an error or offset in the measurement due to the weight of your blood. If you correct for this (and perhaps accelerations etc) with other measurements, then you can have a blood pressure measurement that is much less sensitive to physical motion. If you can estimate the pressures in the heart then you can make good estimates about many other nominal pressures elsewhere in the body.
Yes. Currently wrist blood pressure monitors must be level with the heart in order to be accurate. Can an unlevel measurement be corrected somehow? Would make the device much easier for patients to use. A lot of people don't know they need to hold it at the same level as the heart.
24 hour blood pressure monitoring is the gold standard. Samples blood pressure every 15 minutes regardless of what the patient is doing. Even during sleep. Wearing the equipment is pretty annoying so it's not done routinely. Better, more comfortable wearables with higher sampling rates would be great.
I am using a Aktiia wristband which measures the blood pressure every so often per hour automatically and does so without me noticing it. My cardiologist recommended it to me. I was sceptic and first (and still am, to a certain degree), but it gives me bread useful insight on my blood pressure during day and Escoffier while sleeping. Aktiia's customer service is really terrible (only via email, two weeks response time, inflexible decisions), but the product is good and it's measurements hold up when compared to "usual" measurements.
Both times I've had it done, it was every 15 minutes during the day and every 30 minutes at night. Approximately, anyway. I tend to go to bed later than the settings allowed, so the last couple hours I'm awake I get tested less.
And yes, more comfortable wearables would be grand. Blood pressure cuffs hurt: Last time I had the 24 hour monitoring, I had bruising and blisters from the cuff.
The gold standard is an arterial line, which unlike a blood pressure cuff, measures blood pressure directly and continuously. You will usually only see them used in intensive care or during surgery.
Your doctor is providing a correct explanation for the care you're receiving as this is likely a sufficiently good approach for you in your circumstances.
However diagnosis, especially in early stages, is enhanced by more frequent testing:
"To determine the most reliable way to measure blood pressure, researchers followed 11,135 people from Europe, East Asia, and Latin America for 14 years... The researchers found that the 24-hour and nighttime blood pressure measurements provided a more accurate estimate of one’s risk for heart and vascular disease compared to the in-office readings." (1)
It is a fact that your blood pressure changes while you go about your day and while those values aren't useful for comparing a diagnosis against the standardised test, there is still value in understanding how your blood pressure changes, the duration and amount of that change and risk factors for bringing about those changes.
Furthermore similar to measuring the beating of the heart, continuous blood pressure monitoring can be an useful as indicator of other kinds of medical conditions, this would go beyond what is pragmatically possible with a 24 hour testing cuff.
If you're a person who needs to have their blood pressure checked - then you are a person who will benefit from continuous measurement.
My only qualm to telegaph: continuous indirect BP measurement is hard, and past approaches have had a limited scope. This is one of those technologies where it is wise to have an amount of skepticism.
never have I ever been in a relaxed environment sitting for 15 minutes when I've had my bp taken in a medical situation. oh hey sir follow me. go over here, sit down. cool you just stood in the waiting room for 30 minutes and walked 50 yards and took off your hoodie / long sleeve shirt. now lets immediately take your bp. it's fine if it reads high we'll just spend 30 minutes talking to you about it and ignoring that you stepped on a nail 40 minutes ago.
Hypertension is not the only case blood pressure can be useful though.
On accuracy on why take measurements continuously, I'm wondering how you see pulse measurint. Do you feel it should also be done at rest only, and constant numbers bring nothing to the table ?
I would love to track my hypertension with an always-on device. The cuff measurements are very point in time. If the device was measuring almost constantly, then I could see which of my behaviors positively and negatively impacted my blood pressure.
I'm sure it varies a lot during the day. Have you had your coffee and how much? Have you taken your BP medication yet? Have you exercised? Did you just have a really annoying phone call?
Some of those real-time results might be interesting. Maybe you could cut back on the coffee a little. But there's also a natural variance because life over the course of a day or weeks that many of us would just as soon ignore and instead prefer to monitor accurately over comparatively long time intervals (week+) with at least an attempt to measure under comparable conditions. (And even then I find a fair bit of variance.)
That's a general issue I have with "quantified self" info. It's either something I know. (I didn't sleep well last night for whatever reason or I was at a desk or in a plane seat all day.) Or it's something I mostly can't or won't do much about.
I love all these new noninvasive QS gadgets, but for me the problem will always remain that to find meaningful/useful correlations, you always need to input a lot of data manually (in your example:"I just drank a coffee".)
Until that's solved, and I don't see how it could be, they will remain mostly useless toys to me (who am utterly lacking the discipline to enter data reliably enough for statistical analysis).
I just started logging my coffee by putting an NFC chip by the coffee sources. I press my phone to it, select type and volume. I was shocked at how easy it was to set up.
Section 5.2 is covers the direct measurements, that don't require you exist in some region of a distribution of the population. They don't offer any methods for continuous (the continuous kind) measurements that would be comfortable, which is unfortunate.
Section 5.1 covers "calibrated" methods, where you have to calibrate the device using a direct measurement, or fit within some distribution of the population, for the model to work, which will probably be a great annoyance for doctors.
I suspect it will be a while before we get continuous and accurate measurements.
The biggest change with this tech will be that data brokers will have a continuous record of people's blood pressure to correlate with all the data they're getting from cell phones and smart devices.
The way tech companies are today, no matter how groundbreaking and excellent the monitoring technology, the UI will completely fail on the accessibility needs of the biggest audience: adults 50+
I'm guessing you've had many of the same experiences I have of trying to help elderly parents. The tech industry is way too focused on young people and cool design.
There are many things in life that can disable a person for a time. With proper care, healing and rehab, abilities can be regained. In the mean time, many experiences that were formerly unexamined bring valuable insights. As Cindy Li observed, "We’re All Just Temporarily Abled" <https://blog.jim-nielsen.com/2023/temporarily-abled/>
It's not about tech saavy. It's about the slow decay that is life. You could be Bob Metcalf, Doug Englebart, and Marty Cooper all wrapped up into one, but none of that will suffice to overcome the challenges of the aging body.
You've lived fortunate lives if you've never experienced mobility or perception issues sufficient to interfere with daily life, even temporarily. Not a single broken leg, black eye, torn ligament, not even a sprained ankle. Never experienced the joy and trails of pregnancy.
Sure I've been injured, but that's not due to an "aging body." I got injuries in my 20s, but I don't attribute that to aging. Most people wouldn't put pregnancy in that category either.
Besides, a sprained ankle or whatever doesn't much affect your ability to use tech.
You could be in your 50s and have used computer tech since your youth in the 80s and still hate the current UI holy war against scrollbars and contrast.
For me, the constant stream of other health data shattered my willful ignorance.
For as much as I would worry, the data kept insisting I was stubbornly healthy and that when I did feel poorly, it was almost always attributable to something specific I’d done recently. It made it very hard to maintain the anxieties I’d spent decades culturing, so I eventually just gave in and accepted that I was pretty darn fine and had some specific bad habits that had been surreptitiously fueling the anxieties.
So while more data might make things worse for you, that experience is not universal.
So while more data might make things worse for you, that experience is not universal.
I'm surprised that the type of health data captured passively from a watch is enough to assuage the fears of any true hypochondriac. While they do capture some basic health metrics, it seems that there are many illnesses/disorders that would never be detected by the watch until they become an acute illness.
I think I feel the same way. I would absolutely love to be able to look at a graph of my blood pressure over time, but it would probably be terrible if I could see it realtime on my wrist.
Which is pretty easy to do, at least at home. I got a newer relatively auto BP measurement device that's not even network-enabled. And it still takes just about a minute for the days I use it. Yes, it's something you need to remember but there are a bunch of little daily things you remember--often because you want to.
It sounds cool but I don't think it would change much practically. I have hypertension and a BP monitor is cheap and easy to buy and use. We have plenty of data on how a measurement taken at rest correlates with CVD risk.
For the 45% of people with hypertension without a BP monitor, they probably aren't going to buy something 5-10x more expensive to do the same thing either.
Not to mention that it doesn't change any of the lifestyle or medical interventions we have to reduce BP.
I've just found the whole biohacking crowd to be generally misinformed and lacking foresight on this stuff. It's very simple - you can measure your BP in the morning as soon as you wake up for one week. If its high, you can get medicines or change lifestyle to reduce it. If you're interested in how BP changes with different activities, you could measure that quite easily too, or just look it up. Physical activity increases BP, weightlifting increases it a lot, and none of these temporary increases mean much for overall health. I don't see how getting some more data or a nice graph in Apple Health is going to change anything.
Yeah but if uhh if I could add it to a device that I already wear and it monitors passively that’s a game changer. I’d effectively be able to have it checked more often and while I’m asleep so as someone else said I can see what might be influencing any changes over time. It’s doesn’t have to replace what works if you don’t want to but adding it to tech we already use is a good thing in the long run.
It sounds cool but I don't think it would change much practically. I have hypertension and a BP monitor is cheap and easy to buy and use. We have plenty of data on how a measurement taken at rest correlates with CVD risk.
I think the big advantage of watch-based health monitors is that they can find problems in people that don't already know they have the problem.
You already know you have hypertension so you monitor your blood pressure, many people don't know they have it.
The same is true of non-invasive glucose monitoring if that becomes viable to add to smart watches -- you may still need to wear a continuous glucose monitor if you have diabetes to monitor your blood sugar, but the watch may be a good early warning that you have a problem.
Cool to read this article! 7 years ago I completed my MEng thesis on cuffless blood presurement using an ultrasound transducer measuring the response of injectable microbubbles. I don't believe it went anywhere but it was a very enjoyable research project that made me want to do a PhD.
Blood pressure is such a hokey metric at the best of times. It’s highly dependent on a dozen different external conditions, most of which your average GP doesn’t even try to control for, and yet it’s treated as some kind of oracle of overall health.
It would be nice if Samsung found a way to enable blood pressure readings on their smartwatches in the US. It's been more than two years and they still don't have FDA approval. Maybe they can just enable it with some legal disclaimer that it is not FDA approved and should only be used for guidance and not diagnosis? It's strange that they sell the watches with this feature but it is not enabled, leaving themselves open to a false advertising lawsuit. I've read there is a hack to enable it, but it might make future updates difficult.
it hard to compare it to something, but it seems to be correct. you also must to recalibrate it once a month with cuff monitor that they provide (automatic process, a couple of clicks in their program)
The article seems to skip the direct and continuous blood pressure measurement used by intensive care units. They will typically measure blood pressure via an invasive arterial cannulation. This enables them to get accurate measurements in real time. But as an invasive procedure it’s not without risks. At least it allowed me to sleep without nurses waking me up all the time for measurements via arm cuff.
