AnUser
#161
Easy fix: wear apple watch on one wrist, aktiia on the other.
I got my Aktiia now, BP was 118/75 when setting it up.
Seems very good. Easy app to setup, it’s all in English despite I set region in Germany. I don’t know if Americans have access to the app in their app store. One of the cables to charge doesn’t seem to work but I used the one that came with the cuff. Noooo way I am using a wrist based one with inflation, that would be sooo annoying.
Mailboxde.com worked successfully, shipping was about 20 euros, I had to email them about the package though since and ID thing got probably covered during transport. They fixed it within a few hours. Apparently it’s mostly used for shipping PC parts from germany and has worked for many.
2 Likes
I wear a fitbit on one arm and Aktiiia on the other. I think Aktiia is managing to measure something linked to blood pressure, but it seems to be quite different to the Omron cuff that I use. I am playing around with cations and alcohol at the moment both of which affect blood pressure (the alcohol is more fun). Omron follows this in a reliable manner, Aktiia not so much. I still keep it charged and will recalibrate next time. I am impressed by Aktiia as a company, but what they are trying to do is difficult.
2 Likes
Thanks, that is a very useful reference.
1 Like
J0hn
#164
Just my humble opinion, but the way I look at is, Aktiia are a ‘one-trick-pony’, so they have got to be good at doing that one trick, or they are screwed!
1 Like
That’s true, but it is a difficult thing to do. I am not sure it is actually possible within a reasonable margin for error.
Tomnook
#166
Aktiia on wrist for the past three years - the ‘enforced’ monthly cuff calibrations have always been similar to my former Omron device when compared and I’d say the regular daily measurements are correct in my case - give or take - it certainly detects when I’m stressed and my BP increases.
Apple watch on the other wrist 24/7 and Oura ring 24/7 (less and less impressed by that though - and increasingly impressed by the watch - they rarely tie in with each other) CGM to Apple watch 24/7 as well.
Can’t recommend Aktiia highly enough
1 Like
AnUser
#167
High Blood Pressure is one of the leading causes of death, and right now Aktiia is the only player in the market with a non-inflation based 24/7 BP monitor as far as I know.
It seems to be measuring the physical diameter of the arteries during each heartbeat, so the limit might be how accurate the sensor is:
Aktiia’s Optical Blood Pressure Monitoring (OBPM) algorithms leverage optical sensors worn on the wrist and provide blood pressure values using photoplethysmography (PPG) to analyze the changing diameter of the arteries occurring at each heartbeat
I don’t understand what the accuracy is for the device however.
What does the ± mean? is it standard deviation?
If the mean or average is only 0.45 off for systolic and 0.38 for diastolic, that seems pretty accurate.
I don’t know. I like the product. I just don’t think it is accurate enough. That is not a criticism of the company. My BP is all over the place from mixing up binge drinking and weird proportions of cations. Omrom seems to follow this reasonably well. Aktiaa does not.
2 Likes
AnUser
#169
Here’s data on accuracy:
Correlation is 0.83 and 0.92 for sitting.
It passes ISO81060-2 whatever that means. The 0.46 is the mean for systolic. 0.39 for diastolic.
2 Likes
I have just measured my BP with Omron at 119/72-65, the last average (of one measurement) from aktiia is 127/83-54
Tomnook
#171
ISO 81060-2:2013 specifies the requirements and methods for the clinical investigation of me equipment used for the intermittent non-invasive automated estimation of the arterial blood pressure by utilizing a cuff.
Means the readings are equivalent to a regular cuff device - in my case this is correct.
2 Likes
AnUser
#172
If it’s regularly inaccurate in the same way, that might mean it is accurate but just uncalibrated in some way.
If the inaccuracy can be predicted that is. I am just guessing I am not good at maths.
Your result is about the limit at its inaccuracy, 1 SD.
1 Like
Bicep
#173
Their own material says it’s within about 10. So if you get a bp of 120, it could be 110 or 130, that’s as close as they can come. John’s result confirms this. It may be close enough. How close do you need it to be?
1 Like
scta123
#174
Actually within about 7.
What John is claiming is inaccurate it is impossible to compare. If the app displays a measurement, even if it it was just the average of one measurement it was taken within the two hour frame. Blood pressure fluctuates a lot, just standing up, sitting down, crossing legs change the cuff reading…
I was a bit OCD about testing the accuracy and used the green light as a guide to start the cuff measurement on the automatic cuff. I did it about 10 times, it is really difficult to predict when Aktiia measures otherwise I would do more measurements. My results with automatic cuff reading and time stamped Altiia reading were all within ± 4 and this is about the same accuracy I get with two automatic cuffs one on right and the other on left arm. My blood pressure measured manually with manual mercury sphygmomanometer does not fluctuate from arm to arm. The results measured with automatic cuff usually come within this same range as optical reading vs automatic cuff.
I must say that if we are talking accuracy I have no complaints.
2 Likes
I am pleased if that is the level of accuracy that you find. I think the do a good job of implementation of a difficult task. If it works for you, but not for me I am happy.
1 Like
scta123
#176
Yes, I am happy about it. But I do not claim it is always as accurate. I bet there are measurements that are way more deviated, especially when measuring in less than optimal conditions. But I see Aktiia more as device that measures trends over longer periods not single measurements.
1 Like
Bicep
#177
OK, all positions average systolic error is low by 1.11 plus or minus 9.85. Diastolic all positions average error is low by 1.32 plus or minus 7.56. So we’re kind of both right.
But sitting with wrist at heart level is up 4.6 plus or minus 7.75 and sitting with wrist at lap level is down 3.02 plus or minus 6.1? See, if it’s low by 3.02, then the error is 6.1 then this one (which looks not so bad) really is off by nearly 10.
If you want to be within 10 get something else. jmho, lol
By the way, thanks to all you guys that tested and reported on this. This is very good information.
2 Likes
AnUser
#178
It is not super accurate, but it is the best 24h monitor that doesn’t inflate. Worst case and where 95% of the measurements are within ± 15 it seems. Around 68% ± 7
It will be interesting to see if they can improve the device without new hardware from more data or if they need better hardware, and if it is even possible.
I will continue to happily use this as I think doing a BP measurement 1 time a day or every now and then is not so useful. The multiple measurements might mean it is more accurate over a 24 h period than one single accurate measurement under the same period. Meaning it shows a more accurate picture of total blood pressure over 24 h/1 week/1 month, etc, and total exposure to higher than optimal blood pressure.
2 Likes
scta123
#179
But still it complies to the same standard for accuracy as automatic cuffs.
But totally agree with why it is a pretty useful device. And just fingers crossed the technology will get improved with time. In the mean time I will have to get used to wearing two devices.
AnUser
#180
What is that standard?
This is the only way I am going to measure my BP, I am lazy, so little more noisy information but with more datapoints will be very useful to me. It was quite demoralizing to take a BP measurement but then realize you didn’t have any data for the other 23 hours a day or while asleep.
I didn’t think this was possible so it is a very welcoming surprise to me.
1 Like
