AUScribe is another Australian scribe, funded by the SA government effectively without any proper look into what the alternatives are and what they might cost the state in terms of autonomy or not in patient data.
We got quite a response to last week’s story on a University of Adelaide-led team’s expansion of work on AUScribe. You can read mainly the comments from the academic community after the original story here.
As a rule I don’t read through comments, other than to check for defamation risk, and I never comment back. People are entitled to their opinion.
But in this case I was taken aback quite a bit by how defensive the academics were being and for me importantly, how they were twisting the story in all sorts of ways to justify what they were doing. So I wrote a comment back in the comments stream along these lines.
“That’s a statement I’m assuming you’ve made without any research into the commercial scribes. They are locked down seriously by the need for comprehensive safety and governance. They are in the real world. Not the academic world.
“You do the academic world no favours by posting such a poorly researched view. You make it look isolated and out of touch, which much of it isn’t.
“This post is defensive and not taking into account that these scribes are being used by millions of doctors who love what they are doing for their workflow and day.
“We obviously need good peer reviewed research into the commercial scribes. Private and profit has a lot of issues. But if you were well-read you’d know I give the commercial scribes a hard time too.
“But this? The commercial sector in Australia does an amazing job for the public sector. And this spend of money is simply wasteful in context. Groups like Heidi and Lyrebird have their faults. But you are not pointing out any of them here.
“In fact all the stuff you are saying they do badly here they do very well.”
Cue more outrage from the academic community and more arguments which were either irrelevant or, and here’s the bad bit, effectively misinformation (although probably not consciously delivered).
I’m not claiming any malice on the part of the academics who wrote on the messages or to me directly. It’s worse. It’s ignorance and a high-minded attitude.
The following are a few of the points that were argued by the supporters of the funding and the work:
- The commercial scribes were developed for GPs and specialists. AUScribe is specifically for SA hospitals and patients;
- The commercial scribes spend most of their money on marketing and not on development;
- The article assumes that AUScribe feature set mirrors that of commercial scribes;
- The commercial scribes are “locking in of customers’ systems and data to further their own economic success” at the expense of large government customers;
- The project is research into scribes not a commercial project which is vital for governance and can’t be in the hands of commercial groups;
- In the SA Health system, you can’t put any patient data on “a commercial cloud”.
To be clear, all of these arguments are either false or misleading about what is actually going on.
I’ll deal with them in order:
- Australia’s commercial scribes did start life as a GP tool and specialist tool. But very quickly they have moved into hospitals. In particular, Lyrebird and Heidi have so far, the lion’s share of scribe use across our hospital systems, both public and private. And they have gained a lot of fans in the hospital sector. Lyrebird has won a contract in England to be the scribe of choice in four of London’s major hospitals, and Heidi has won a contract to be a scribe of choice in 15 NHS trusts, which includes hospitals.
Both already have integration pathways to the major global EMRs. There’s a long way to go in hospitals and they may not make it because the major EMR vendors are building their own native products to keep them out. But it is entirely misleading or just misinformed to suggest these scribes are not suited to hospitals like AUScribe might be.
- On the above point, the arguments being made for AUScribe to be moulded for hospitals specifically, ignore the fact that SA already has a major EMR implementation across the state – Sunshine from Altera – which launched its native AI scribe nine months ago. Why is SA Health spending money on something which has no integration pathway to its EMR platform, a platform that already has an integrated scribe?
- The suggestion that the commercial scribes would be spending less money on development than AUScribe, which we know is about to spend about $1.9 million of government money on development, because they spend so much on marketing, is misleading at best. But at its worst it’s a desperate clutch for some straws. So far Heidi has been funded to the tune of about $165 million, and we know it’s in another funding round now, which if they can get it away before the massive US AI Ponzi scheme bubble bursts, will be larger than all that has preceded it. AUScribe might be spending more percentage wise, but even at 100% they are going to fall short of development investment by Heidi by, at a guess, of course, $50 million plus.
Related
- To suggest AUScribe will have a feature set unique to anything the commercial scribes could build around their core product in no time if they were working with SA Health is yet again, misleading at best. First of all, scribes ain’t rocket science. No matter what the academics like to think about how smart they are compared to other entrepreneurs – those who are risking their own money to build their product, not the government’s – they can’t match the feature sets that are already built and can be built to suit a particular jurisdiction. They are already doing this in Australia and around the world for various hospital groups.
- The argument that the commercial scribes are locking up customers and data to the detriment of the SA government is ludicrous. The SA government and every state government rely almost entirely on commercial vendors who are working with them as partners in patient data and engagement. Altera’s Sunshine, the core EMR for SA, is a commercial product that deals entirely in core patient data. Do the academics think we should maybe not be using that? NSW, Victoria and Queensland use mainly either Epic or Oracle (Cerner) as their EMRs. And all state governments use all sorts of local and overseas vendors in combination to make their hospitals work for patients. Alcidion, a great Australian AI company – not scribes but data management – is everywhere in Australian hospitals.
- The project is research, not commercial, and not competing with commercial scribes is another case of clutching at straws while misleading the market. In the original post on AUScribe’s funding, one of the researchers suggests that if they can get AUScribe working for SA, it could be used in other states. Or is it not competition if you’re government funded?
- The argument around sovereignty of data is as bad as the rest of the arguments made for AUScribe. At first it was claimed that the commercial scribes couldn’t solve the sovereign data problem and when it was pointed out that this isn’t true, the argument has been made a few times that SA is special in that, and we quote, “SA Health policy does not permit patient data to leave SA Health, for privacy reasons” … which is, yet again, misleading at best. Public hospital patient data can be placed on a commercial cloud service in South Australia, provided it meets strict data governance, state record frameworks, and national privacy laws, which the commercial scribes can all meet. Do these people think that NSW Health using AWS for their SDPR project is okay? Or is that too commercial for them?
There’s a lot more to tear into here in the context of clutching at straws arguments being made, all the way to “big tech vendors are exploiting the stock market (yep, true) so we shouldn’t be backing the likes of Heidi or Lyrebird”. Earth to SA Health and co: Heidi and Lyrebird are not big US tech exploiting the stock market, they are local vendors fighting desperately and pretty much above their weight, in the fast-evolving international scribe market.
One piece of feedback I got on this article this week was that it was a bit rough to have
SA Health as the caveman making a square wheel on the side of the road.
Was it though?
Also, it’s a cartoon, everyone … get a life.
The same arguments made last week stand on this project: we are using government money with no rhyme or reason other than “not invented in my backyard” and to suit some academics’ desire to be involved and get published.
Academics can and should be involved. Deeply. But not in this crazy wasteful manner.
Commercial scribes definitely need oversight, research and criticism. But AUScribe isn’t that. AUScribe is another Australian scribe, funded by the SA government effectively without any proper look into what the alternatives are and what they might cost the state in terms of autonomy or not in patient data.
Now that’s what you call poor governance.