HRM Pro plus and AFib

You could look at something like the https://uk.fourthfrontier.com/

Frontier X2 HRM. It is expensive, however, and although frequently discounted, still expensive. However, it might give you valuable data. I do have one myself, along with HRM-PRO, Polar H10 and various Wahoo straps.

I bought it when I experienced some concerning heart behaviour last year and Black Friday pricing persuaded me to purchase. I think that as a medical device it probably functions quite well, but as a sports HRM I'm not so convinced. No problem with accuracy, but Bluetooth only (multiple channels) and HRV data is only transmitted to their app, not your watch. I rarely use it now, preferring my HRM-PRO (with running dynamics) for running and Polar H10 for cycling, both of which support ANT+ and HRV. 

I am really dissappointed that Garmin hasn't dug into this more beyond posting a disclaimer buried in their web site. Filtering out noise and identifying the basic sinus rhythm and then abnormalities from the baseline sinus rhythm using simple algorithms or machine learning is not difficult. The "code" itself is pretty simple to write and can easily run on a garmin watch or edge type unit with virtually no stress to the CPU. Heart rhythm abnormalities are pretty normal. In my cycling group I know of 5 people with similar issues and doing a nice easy ride and seeing your HRM say you have a rate of 195 is a little frightening.. and then the watch says 125

I think the problem relates to how straps measure heart rates 

Most straps  (including Garmin and Polar) record a two lead ecg and measre the R-R interval in milliseconds  An instantaneous heart rate is derived for each beat  (effectively 60000/RR interval in milliseconds)

Because the RR interval in AF varies,  your instantaneous  heart rate may be 180 for one beat, 90 for the next and 120 for the following with an average of 130.

A better measure in this case would be the number of R waves over a fixed time.  A standard ECG measures a rhythm strip over 10 seconds

your Kardia measures over 30 seconds and I believe most watches measure peripheral pulses over a short time; certainly, my Apple Watch needs a few  seconds to settle when I measure my heart rate at rest.  In AF, particularly at high heart rates, not all QRS complexes result in a demonstrable pulse, hence clinically in fast AF it is important measure the pulse peripherally  as well as listen to the chest for a heart rate.  This may account for the difference in rates between your watch and strap

I’ve  just checked my watch and unlike cadence or power there isn’t a standard field for 3s or 10s averaged pulse. It may   be worth seeing if there’s something on connect, IQ

Diltiazem is a rate limiting calcium blocker and has a similar effect on heart rate as a beta blocker  this may cause ECG changes hence Garmin’s warning  

Interpretation of AF by a device  requires regulatory approval with health authorities in the local country or economic bloc. This is why the Apple Watch ECG feature was released in different regions at different times.  Initial approval was granted by the FDA in  the States with EU regulation, Asia  and Australasia occurring  much later  It requires more than just adding an option to the Watch in software  and licensing comes with development and economic costs

Dee 

Interpretation of AF by a device  requires regulatory approval with health authorities in the local country or economic bloc. This is why the Apple Watch ECG feature was released in different regions at different times.  Initial approval was granted by the FDA in  the States with EU regulation, Asia  and Australasia occurring  much later  It requires more than just adding an option to the Watch in software  and licensing comes with development and economic costs

Dee thanks for the comprehenisive and accurate response.. I believe Garmin can figure out the AF issue in the same way Kardia soes as they are both two lead devices but Garmin would have to build some form of not very complex algorithms to stabilize or interpret the core rythm. Kardia does and its not like garmin doesn't have a huge data set to work from if they are using some form of ML. The other issue as you identifiied is the Apple watch is I believe ISO 13485 or registered as some form of Class X medical device whereas the Garmin is not and I know from experience that process is hideously expensove based on jurisdictional fees and documentation..  I think ultimately my solution will be to buy some form of high end Garmin watch that will determine HR from wrist pulse versus the strap and use the strap as a visual indicator as to level of cardiac activity whike riding.. I suspect I will get cardioversion in the next 60 days and it will be interesting to see what the read outs are after that