Regional / National platform for resp/sleep tests

Discussions
, ,
1

Kia Ora - first post from me, thanks Nathan for the introduction!

We are trying to develop community testing pathways for common respiratory conditions across the lower half of the Te Wai Pounamau - there is already a Canterbury platform which I understand to be well past its best before end date, having been written by the late Josh Stanton (who died in 2016). The Canterbury platform must be at least a decade old and cannot be updated or ‘grown’.

In brief, we want spirometry and low complexity sleep testing to be available in the community from strategically placed primary care units. Getting a sleepy driver to drive from Wanaka to Dunedin and back for a test makes no sense! There needs to be QC oversight and (at least for sleep) we in the hospital unit need to be able to see the data for clinical decision making and reporting purposes. In turn obviously the referrer and ideally the patient should be able to access the results and relevant reports. Currently we have HealthOne as the place to review these but challenging to get community data and reports on to this.

I gather Canterbury are investing money in rewriting a local solution using local IT skills. There are (as best as I can tell) two potential commercially available solutions - Rezibase (an offshoot of cardiobase) and Respiro, which I understand is fairly widely used in Australia (the company are a spinoff from the Prince Alfred).

What we really need is a South Island platform (or possibly a national one) and at the moment we are putting money in to two competing solutions - the Canterbury only solution and a Southern one that might be more widely applicable. Respiro only covers respiratory testing whereas Rezibase are actively developing sleep as well as respiratory testing.

Does anyone have experience of either of these platforms? I gather Respiro is used in Palmerston North but AFAIK no Rezibase experience in Aotearoa. How can we get buy in for a South Island digital platform for the future? Who do I need to talk to? And has anyone had success in similar scenarios and how did you make it work?

Thanks in advance for your thoughts,
Ben

2 Likes
2024: The eHealth Forum Year in Review
2

Nice to see you on here @bendun!

I’ve always found Jon to be great at knowing who to talk to around the traps with this kind of thing, particularly if you’re wanting advice at a National level.

Any pointers for Ben, @jon_herries?

1 Like
3

One of the questions we are trying to explore is whether the screening element of this could be done by the consumer and their mobile phone. Feel free to reach out to me on email to have a chat.

4

I think that @bendun is talking about a more serious platform for managing the information flow from community testing that either is ‘out there’ or hopefully soon will be.

From my thinking, a true national solution (that covers both public and private) would be ideal for this as it isn’t only a South Island problem - although we all do snore a bit…

Rezibase seems to be the better option as not having sleep would clearly be a big problem. And we know that cardiobase is pretty widespread, giving a head start with integration into the clinical portals we all know and love.

Anyone from @NCPUG got any thoughts on this?

1 Like
5

One of the questions we are trying to answer is “How different is this data” or “how different are these requirements”.

Not clear to me beyond forms and a way to store waveforms, or raw blobs this would be (eg. how does storing spirometry data differ from storing ecg data). Visualising it is different in terms of scale and units over time.

6

Hey @Jon - Oura Ring uses disturbances in blood oxyden levels as a first indication of sleep disorder. Is this something to think about?

1 Like
7

Hi Ben, I can only offer advise on your immediate issue.

RespiroTM goes to mgcdiagnostics and they say they can send data through via “HL7 or other”, HealthOne can consume that. Rezibase (cardiobase) have a diagram on their website with an arrow pointing from their system to an “EMR” via Report. So that suggests that they could send data to HealthOne (or other platform).

So when you say its challenging to get the community data onto this I wonder why that is. Is it interfaces from the sender or receiver or something else? You can DM if you like.

I’ll ask my Aussie colleagues if we have integrated either of these speciality solution providers data into a HealthOne looking platform before (but it doesn’t ring bells atm). I’ll leave your question about national procurement and strategy to the likes of @jon_herries. FYI there was also a recent post somewhere about national interoperability and rules around when to use FHIR etc. If you want atomic sleep data into platforms then getting it raised on that list would be a good shout I would’ve thought but looks to me like its in its infancy from what I can see.

@dianasiew I love my Ring!

1 Like
8

Was thinking even more simply than that - a questionnaire plus audio of the room while you sleep. Might have instructions to try and sleep in the room by yourself?

A paper from late last year using an XGBoost model on audio:

Assessment of obstructive sleep apnea severity using audio-based snoring features - ScienceDirect

9

Hey @NAHSTIG team, here’s a thread of interest for respiratory & sleep physiologists. I know there’s a few of you there.

#respiratory #sleep

1 Like
10

Thanks for the responses and sorry for the slow reply - I have been in Australia on holiday.

For clarification, there are a few levels at which the information needs to be visible.

For normal people there’s the need to see the report (usually as a pdf) e.g. in HealthOne as a GP or as a hospital doctor, to check the results. Generally this would involve 1-4 pages of text and diagrams depending on complexity of the test.

For us more abnormal ones, we need the data in a reviewable format - for instance, in order to create an accurate report on spirometry we need to be able to review the tests as they were done - e.g. flow volume loops for breathing tests or heart rate/saturation data for a low complexity sleep test. Once the data has been assessed for technical acceptability, a report is generated - a process that usually requires assembling stock phrases in order to allow some uniformity of reporting. So a typical report written in the current (really quite difficult to use Breeze) software and uploaded to HCS as a .pdf might run along the lines of…

The tests are technically satisfactory. The BMI is normal. Spirometry shows moderate airflow obstruction (z score between −1.64 and −2.55). Reversibility was not demonstrated. The DLCO is severely reduced (z score greater than -2.5). This is consistent with COPD. Smoking cessation is strongly recommended. Report by Ben Brockway, Respiratory Physician on Sunday 28 Jan 2024).

As you can see this is totally something that AI could do (and frankly I would be delighted not to do given that we do thousands of such tests a year and currently no-one has reporting them in their job plan - but that’s a different matter altogether!).

Sleep reporting is a little more wooly and less protocolised, but perhaps not much. While I can make clinical decisions based on a summary of the results (and often have to), the quality of the reports is better if you can review the data generated overnight - this can be reviewed in different ‘epochs’ e.g. for some things a 10hr/page allows the overall structure of the night’s sleep while 10 minutes per page allows assessment of the fidelity of the recording and whether the desaturations are because someone has stopped breathing, or just because the oximetry trace has a lot of interference. Again, the reports are more readily generated by assembling blocks of comments as it’s way faster than me typing.

Rezibase and Respiro both offer the ability to drill into the data to allow review; the file sizes vary from small for spirometry to about 50Mb for a full in-hospital polysomnogram (we only do about 50/yr of those vs 500 home-based studies).

Thanks to all who have mentioned proprietary devices for home assessment. It’s a fascinating area and moves quite quickly. In our experience when we compare an in-hospital PSG (the gold standard) against wearable devices there’s an OK correlation but not great, and almost all the commercial products have ‘black box’ algorithms that can change with software updates and don’t play nicely with AASM definitions of what is, for instance, obstructive sleep apnoea.

Thanks again for the interest and sorry for the delay!

b

Samsung’s Sleep Apnea Feature on Galaxy Watch First of Its Kind Authorized by FDA