As one of the core principles driving MediPi is to enable low cost solutions, the pilot demonstrator for MediPi has been configured based on commercially available physiological measuring devices. As such, our selection of hardware components during development, has been driven more by cost and less by medical device categorisation.
The devices have not been classified as approved medical-grade devices by the Medicines and Healthcare products Regulatory Agency (MHRA) although all carry CE markings. We understand that a measurement device can be accurate but not precise, precise but not accurate, neither, or both, and this raises the question as to the positioning of such devices in this pilot.
Across the apps sector blood pressure, weight, pulse, activity etc. are all widely measured by the public using unclassified devices as part of their proactive health monitoring.
We are looking to work with MHRA for guidance but in the meantime wondered what others’ views are regarding these devices and in what context and with what restrictions they could be used.
Have you had any experience of using commercially available devices in your solutions? Do you have any tips or suggestions you can share that would support us continuing along a ‘low cost’ path until such time that wider production-scale considerations become key?
Hi Richard, we have integrated blood pressure monitors into iOS and Android devices. I have a desk full of blood pressure monitors we purchased, and we had very mixed results.
On iOS you have the benefit of apple health app, which makes import of readings into an app you would develop easier.
the downside is apple health is iPhone only, so you would be shutting off older demographic.
Also not all BP monitors are created equally from a usability standpoint. We tested one which was a Medisana one, which had CE marking, clinical validation - but was very difficult to work with it’s own app. We generally found traditional BP monitors which had their own screen and settings were more difficult to work with apps, as there could be date conflicts etc.
On Android, we had to work with API’s from commercial BP monitors. The best integration we got going here was with Withings. Main downsides here were cost of the monitor, and users we tested with found it onerous creating two user accounts, and then linking the two.
This has then led us to looking at direct BP monitor integration via Bluetooth 4. We have a good prototype working here, but with a monitor which doesn’t have clinical validation (but has CE marking). We are still in this phase where we are looking for a monitor which can have a direct integration, but be a slave of the app (i.e. no time & date setting on monitor).
Hope this helps
Thanks Eamonn,
That’s valuable information.
WRT the devices:
The BP and scale devices we are currently using are Beurer USB devices and are “Pseudo wireless” (measurements are taken whilst the device is unattached and the results are downloaded when plugged into the MediPi Patient unit). The Contec USB enabled finger oximeter streams data and therefore needs to be plugged in.
As MediPi is a open source Java software project we are looking at all platforms (although IOS less so because of it’s restrictions).
I have been looking at Continua approved devices (bluetooth) which seem to have an open API - have you had any experience of these?
The BP cuff is approved by the British Hypertension Society:
http://www.aandd.jp/products/medical/bluetooth/ua_767pbt_c.html
The scales have no clinical validation:
http://www.aandd.jp/products/medical/bluetooth/uc351pbt_ci.html
WRT the data produces by the devices:
I’m trying to really answer the question of how we can work with the limitations of the devices we use. What can be done with the data produced? It would be great to have clinically approved devices that provide empirically correct information, but these are sometimes many times the cost and we want to explore what can be done with low cost devices.There are instances where highly calibrated and accurate devices are required and that must be considered base cost in these areas, but if we are measuring deltas over a period of time, an absolute calibration can be mitigated - if the device has consistent accuracy but not precision.
If a patient were to present to a GP with a list of readings they had taken with their bathroom scales and COTS BP monitor - the clinician wouldn’t dismiss the data - it would be viewed in the context of how it was taken.
Have you any thought on this?
In our applications we’re looking at engaging with patients, and trying to engage and change behaviour. Our initial focus has been on Kidney Transplant. With demographics, and interviews we have carried out we decided using people’s existing smartphones was best route - which means we need to find a solution which works with iOS and Android.
hand’t looked at Continua - will definitely take a look, and at the BP.
We’ve also looked at Weighing scales - for our usability testing we felt wifi was the route to go (and not bluetooth).
On accuracy / provenance of data. Nephrologists we’re working with in Transplant are trying to get patients to do this regularly, they’re interested in trending and getting patients to take corrective action. In my opinion, this needs to be done with lower cost devices rather than clinical grade
I agree, with the pressure on costs we must define what is possible with these devices.
Additionally, if we can it provides an exciting opportunity to use a great deal of information from the proliferation of consumer health devices (wearables like fitbit etc).
I assume (with your wifi enabled weighing scales especially) that you’re interfacing with the device cloud storage?
we’re tying it in with our own cloud storage. we also integrate with Apple Health / Google Fit for bringing in activity data - but trying to make in clinically relevant.
So in our transplant use case, typical pathway would have patients on dialysis and then transplanted - so its about motivating to get more physically active post transplant. As (most) people carry smartphone around with them - there is no need to purchase a fitbit or similar device, keeping cost of the solution down.
Hello, I’m looking for an example of a former alcoholic, seen in outpatients for potential liver failure and using telehealth in support of his long term care plan.
Has anyone written this up as a use case please?
Regards
Paula
Tel: 07709250037