http://www.theregister.co.uk/2015/02/17 ... res_scare/

http://www.zdnet.com/article/sensor-iss ... RSSbaffb68

Problems getting the sensors reliable mean that Apple are dropping the Health aspects of the Apple Watch and they are now marketing it as a companion device for the iPhone,

Despite this, they are still looking to shift 6 million units before the end of the first quarter.

Sensors are important. Nobody wants useless readings. I've had people buy wrist-mounted blood pressure monitors and the results are so inaccurate it's ridiculous. Even NICE have recommended against their use for BP monitoring. Furthermore, all studies involving blood pressure were based on mercury-based sphygmomanometers and the way the sounds are heard. Oxygen monitors use infrared light (I think) to pick up oxygen levels. Again, useless for most people unless they're unwell eg severe pneumonia, in which case they're better off seeing a clinician.

They have always marketed it as a companion to an iPhone, at least initially, and this is a report, siting a report, siting an un-named source. Let's wait and see what actually is in it when it launches because, let's be honest, none of us are sat in front of an Apple branded TV with our Apple branded netbook.

They haven't changed their marketing at all - Apple Watch Heath and Fitness.

To be fair ALL the Apple watch marketing bears the following footnote : 'Features are subject to change and may not be available in all regions or all languages. Requires iPhone 5 or later.'



Again, nobody knows what Apple are thinking they'll do with the thing. This is someone who has no intimiate knowledge of the project pulling a number out of their arse.

It kind of does depend on what you think the device is for. i saw a recent item on the BBC where somebody tested several fitness bands and found that their pedometer readings were significantly variable between each other. But they also made the point that in fact saying you have walked 7400 or 7600 steps a day is less useful than saying 'you need to walk more' i.e. the trend data is what is important and as long as the device is self-consistent, the trend data will be useful.

If people think these devices are as accurate and exacting as expensive specialist medical equipment that costs several times the price and a certain amount of training to use, they're deluding themselves. But to use a parallel example, the fact bathroom scales are basically a fairly rudimentary strain gauge hasn't stopped them being a useful tool for lots of people who are trying to lose weight for a very long time.

it is possible though that once/if people expect such devices to be used as part of a mandated medical regime then yes, that's an issue. And it may be the basis for the rumour as to Apple's health monitoring in the Apple watch - if Apple can essentially be held legally liable for the accuracy of the data to the same degree a medical technology company would, then the Apple watch probably isn't going to cut it. But neither will any of the other lifestyle health devices currently on the market.

A different picture seems to have emerged since people got a chance to look behind the WSJ headline. The current view seems to be that the Apple Watch was originally intended to have several biometric sensors that either couldn't be guaranteed to be consistent or would have required FDA approval in the US. Apple decided to drop them at the design phase and the watch that they eventually launched had the sensors that ended being practical.

so Apple haven't dropped a thing, at least not unless you believe every feature suggested as part of every prototype of any product ever is 'dropped' if it doesn't make it to the marketed version. There's no indication that the final model will vary from the one they demonstrated and talked about at launch. Although presumably the caveat still applies, and there's no guarantee it won't.

Look, there appears to be a raincloud in this drinking vessel.

There was a report, I think it was on TWiT yesterday, that a recent round-up of sensors found that some were reporting significantly wrong values. One of the health bands reported a calorie usage that was 2,700 calories less than had actually been burnt (control was a certified medical device)in a week. That is more than a normal days intake! That means that the calorie usage for 1 week is something like 20% off. That is a huge margin of error.

If the device is that inaccurate, what is the point of having it at all?

If the problem is making the sensors reliable, then I am glad Apple decided to not implement them in the first version. No information is better than false information.

This is what i was driving at. I would hate the idea of people turning up to the surgery or hospital based on a spurious sensor reading.

Here’s a question - which of those two devices are more accurate - the automatic/computerised ones, or the ones where the doctor uses a bladder to inflate the cuff and a stethoscope to listen?

Given that most hospitals and doctor's surgeries still seem to use the manual method, I would guess that it is more reliable. The other plus point is that a human can double check the results. If you only have the automatic device, you cannot double check the values, even if you suspect they are wrong.

Given the reliability of the home test blood pressure monitor and also the long-term monitor I had to wear (sometimes the long-term device would have to take 5 or 6 readings before it was happy), I wouldn't trust them.

We have both in the surgery. Each year, they're calibrated.

I tend to find automated ones either under or over read. The better ones are very close to what I get using a manual sphyg. The average ones lose accuracy rapidly. The cheap ones are way off the scale. I tend to get patients to use one to monitor at home and bring in to ensure they're reasonably accurate. I reckon about a sixth of our hypertensive patients have white coat hypertension as well.