back to article Surprise! That blood-pressure app doesn't measure blood pressure

Quantified self types not only fill Twitter feeds with reports from every walking, running, breathing and bonking app around – but the spewed data isn't always particularly accurate. In a letter published in the Journal of the American Medical Association (JAMA), a bunch of US medical academics decided to test AuraLife's …

  1. Crazy Operations Guy

    Companies like this are pretty close to getting sued into oblivion, and quite rightly. There are quite a few impressionable people out there that would believe an app like this and skip their necessary physicals, which leads to quite a few very preventable deaths per year.

    1. Nick Ryan Silver badge

      This app sounds about as trustworthy as those apps that claim to measure blood sugar levels.

    2. Fungus Bob
      Trollface

      quite a few very preventable deaths per year

      Sorry, C.O.G., death is not preventable. One may postpone it for a while, but ultimately the death rate is exactly one per person.

      1. I. Aproveofitspendingonspecificprojects

        Re: quite a few very preventable deaths per year

        Not a zero sum then or are all heart patients neutered?

        I couldn't do the maths, is that 148000 copies at $4.99 or: less than <?> 148000 ambulance chasers multiplied by a relative or next of kin time-line?

        Whatever the case I am sure that an apple a day will keep the lawyers at busy

    3. joed

      I'd bet that affordable medical services would help more than the app. In lieu of these (at least Stateside), apps/web search has to suffice for many. At least as 1st diagnostic step (and more advanced medical procedures will probably be out of reach anyway). The same applies to other fields, just much less price competition in medical (technically it's all union, no outsourcing etc). Regulations do not allow for "shade tree doctors" - damn you don't, damn you do.

    4. DocJames
      Stop

      "necessary physicals"

      Hmm. Assuming you are talking about the general population rather than those with chronic health issues, you should be aware that there is no evidence supporting routine physical examination (and history, let alone any investigations).

      Even the US is trying to move away from this, although people are fixated on "this is what has always been done; it must therefore continue each year forever" ritual, rather than jumping at the chance to stop being a patient and get on with their lives (what most actually unwell people would do, I'd guess). Fee for service environments drive such poor quality (and high cost) healthcare.

  2. Anonymous Coward
    Anonymous Coward

    Blood pressure readings do not seem particularly consistent even using a clinically approved meter.

    There was an article a few years ago that said that only the invasive method was accurate. Researchers found that normal clinical cuff measurements can be wrong outside the normal range. A new non-invasive method based on detecting the K2 change has been proposed as being as accurate as invasive methods.

    http://www.ncbi.nlm.nih.gov/pubmed/3370770

    I take my blood pressure when I wake up in the morning as a daily record. The range of values often vary over several measurements within a few minutes. That means that only trends over a long period are good indicators. The spot readings in the doctor's surgery are always elevated - even if I don't feel stressed. They have a tick box for "white coat syndrome".

    http://www.nhs.uk/news/2011/03March/Pages/white-coat-effect-high-blood-pressure.aspx

    1. Medixstiff

      The spot readings in the doctor's surgery are always elevated - even if I don't feel stressed. They have a tick box for "white coat syndrome".

      I had the same problem with the doctors surgery readings being different. I was using a BPM my friend lent to me, however where that would show a BP of around 144, the doctors one would be in the 160's.

      Luckily mum got one through work, that was the same manufacturer as the doctors one and that was basically spot on with the doctors one.

      When I read the manual for the new unit, they specify it's accuracy to within 2%.

      As such I was put on medication and that made a big difference.

      This all happened because I was checked into hospital with a BP of 236 after having a migraine all night and the GP being worried about my BP being in the 180's and telling me to go to hospital to get it checked out.

      This was after years of doctors and the red cross blood bank people asking me if I had a higher than usual BP but not actually ever recommending I get it checked out.

      1. Chris Miller

        I have the same experience. A commercial 'cuff' machine, similar to those in the surgery costs about £30. There's no reason why a less absolutely accurate device shouldn't be useful in spotting trends, which could lead you to getting a 'proper' check.

      2. Anonymous Coward
        Anonymous Coward

        "[..] that was the same manufacturer as the doctors one and that was basically spot on with the doctors one."

        That thought had occurred to me - so I took my BP meter with me to my check-up clinic. Its reading agreed with the nurse's meter. I even bought a second identical model and occasionally do a simultaneous calibration check of left and right arms. So far both meters give effectively the same reading - allowing for the natural slight difference between left and right arm readings.

      3. VinceH

        "This all happened because I was checked into hospital with a BP of 236 after having a migraine all night and the GP being worried about my BP being in the 180's and telling me to go to hospital to get it checked out."

        Similar story. (Having not had reason to visit the doctor in many years) I woke up one morning with double vision - and a visit to the doctors later (by which time the double vision had more or less cleared) I was sent to hospital, where I spent the next few days, because my BP was at 240/140.

        I've been on pills ever since to keep it steady, and I check it regularly using a Pro Logic PL100. It may not be 100% accurate, but I wouldn't even consider using a smart phone app in its place.

    2. Sleep deprived
      Happy

      - They have a tick box for "white coat syndrome".

      Last time I had my blood pressure checked, that doctor would have deserved a "deep green eyes wrist-on-skirt syndrome" tick box...

    3. I. Aproveofitspendingonspecificprojects

      > They have a tick box for "white coat syndrome".

      It should have relative waiting times adjustment, I am always late for mine hence my need for Amlopidine.

  3. Anonymous Coward
    Anonymous Coward

    Is there an app which tests for gullibility?

    1. Pompous Git Silver badge

      Is there an app which tests for gullibility?

      http://hoaxes.org/test.html

      1. DanDanDan

        Hoaxes.org

        Was sure this was going to 404!

    2. Big-nosed Pengie

      This one seems to qualify.

  4. sysconfig

    What's wrong with the January screenshot?

    It says ESTIMATE your blood pressure using your phone... That's a pretty fair and clear hint that it won't be quite as accurate as proper medical gear.

    1. Pompous Git Silver badge

      Re: What's wrong with the January screenshot?

      I have a reasonable quality, medically approved sphygmometer. I took it with me when I visited my GP to see how its results compared with his instrument. The results were not identical. The best indication that you aren't going to get as accurate a result is the cost. $4.99 versus well north of $100.

      BP results are always "estimates" as there are factors such as the "white coat effect" that gives a higher result when it's taken by a doctor. Pretty nurse also have a noticeable effect as well, but don't make the mistake of telling them. One such lifted my surgical gown and gave the end of my dick a belt with a medical instrument to bring my BP back down. It worked!

      1. Anonymous Coward
        Anonymous Coward

        Re: What's wrong with the January screenshot?

        sphygmomanometer?

    2. IvoryT

      Re: What's wrong with the January screenshot?

      Just feel your radial pulse. You can estimate your blood pressure from the volume and nature of it. Easily as accurate as the App but useless to guide treatment of course.

  5. Pompous Git Silver badge

    OTOH "necessary physicals" aren't necessarily going to arrive at a correct diagnosis either. About a decade ago I was diagnosed, correctly, as having tachycardia and hypertension. I've been on a beta blocker and an antihypertensive ever since. Despite consulting a heart specialist, it wasn't until December last year that I was diagnosed as also suffering heart failure on admission to hospital. I couldn't walk more than a hundred metres without running out of breath. Now that I am on the correct medications/dosage, I can walk briskly for an hour without running out of breath.

    For ten years I was told that my breathing difficulties were caused by chronic asthma. Last Tuesday, the heart specialist in charge of my case when I was hospitalised told me that it's extremely rare for people in my age group to suddenly begin suffering asthma and the symptoms are almost invariably associated with heart failure. The joys of advanced youth...

    1. Duffy Moon

      Did you not have an ECG? That should have revealed the heart failure.

  6. SimonL

    The way I read their 'attack' on the tests basically told me they were admitting that their app was just some novelty rubbish that should not be taken seriously or for that matter has any real-life use!

    This business of it does not display outside 'whatever' range is worrying.

    From what the tests discovered, if your blood pressure is particularly high, rather than the machine saying "It's too high, outside of my range, go and see a doctor" it just makes up a reading!!! :o

  7. Whitter
    Boffin

    You get what you pay for.

    Most machines one can buy in Boot and the like say something like "not suitable for hypotensives; not suitable for hypertensives..." and the like. Making them about as useful as a sticker saying "120/80".

    That the measurement itself is difficult and not very repeatable (the "gold standard" by your doctor suffers similarly), and is almost always never used in the correct way, is another set of different problems!

    1. Anonymous Coward
      Anonymous Coward

      Re: You get what you pay for.

      Haven't seen any "K2 algorithm" BP meters yet. In research trials that method proved as good as invasive measurements.

  8. Elmer Phud

    " between 102 and 158 systolic and 65 to 99 diastolic"

    Well, that 's pretty useless.

    A little while ago, 99 diastolic was low for me.

    1. Sooty

      yep, I'm well over on both.

      The main worry for people who generally want to test their blood pressure is if it's very high, and this doesn't actually work in that situation.

  9. Geepjul

    It's not a simple as it looks

    As a doctor I am used to the apparently simple, but surprisingly complex diagnosis and management of hypertension.

    When I started 2 decades ago 3 high BP readings were enough and it was 'on to the pills you go'. Now we realise this is not an accurate approach so we either do a 24 hour BP recording or a longer series of blood pressures ( often taken by the patient at home). This should filter out pure white coat hypertension, although it should be noted that people with white coat syndrome still carry a higher risk than the general population.

    I am happy for patients to take their own blood pressures using good quality machines (I would not recomment the wrist worn types) but care needs to be taken about how you use them, including cuff position and size. I also check them against my pro machine which is calibrated regularly.

    This article highlights the point that not all health apps are equally reliable and some are not much better than flipping a coin when it comes to potentially important decisions about your health. Whilst regulation can be a blunt tool, I do feel that this is a situation where a testing and approval process is required before they are allowed onto App stores

    1. Anonymous Coward
      Anonymous Coward

      Re: It's not a simple as it looks

      "I am happy for patients to take their own blood pressures using good quality machines [...]"

      My doctor is also happy to use my spreadsheet of the last 20 days' BP readings taken as soon as I get up. The nurse does take three readings - but knows they are always elevated. We checked the calibration of my meter against hers. The 24hr ABPM device was used initially and I was declared OK. Then the Diabetes Type 2 diagnosis was made and the BP thresholds for medication were accordingly lowered.

      1. Whitter

        Re: It's not a simple as it looks

        As your doctor has had a look I'm sure all is well, but be aware that readings taken after waking are expected to be higher than normal: google "morning surge index" for more info (the index is a means to establish if a given surge is more/less than expected).

  10. Grease Monkey Silver badge

    So the vendor openly admits that they are selling a blood pressure monitoring app that will only work for people who don't have anything to worry abut in that area? Which when read in the context of the study also tells you that the app will tell people with high blood pressure that they also have nothing to worry about.

    So in that case why bother with all that coding? A text file containing the message "Your blood pressure is fine" would have done the job every bit as well.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like