Sunday, September 07, 2014

The View From The Coalface Regarding The EPAS in South Australia. Not Good News!

I had an unsolicited e-mail asking about the future in Health IT in OZ and commenting on the clinical user experience of EPAS in South Australia a few days ago.
Here is an edited version of what was written re EPAS - with the identity of the writer obscured totally.
To quote:
I came across your blog recently after googling e-health in Australia as I am becoming increasingly frustrated by the incompetencies of the government here.
I have recently had the pleasure of being forced into using SA's outsourced, over-budgeted, under-cooked mess of a program called EPAS.
How is it that a government with a budget of >$420M for a single ICT project was unable to, firstly, build the project in-house (and creating hundreds of IT and healthcare jobs in the meanwhile) and, secondly, end up buying such a mess of a program? EPAS is convoluted, it's difficult to use, it's buggy and crashes and it's an absolute user interface nightmare. It makes me cry thinking that this is what we are meant to use for the next 10-20 years. How is this even remotely possible?
I have a clinical background with some knowledge of IT. Not once did we see an ad looking for people who have a background in medicine + IT in order to help them with this project. Not once did they put it out to the public/relevant parties regarding whether this is a viable system.
As it stands, EPAS takes everything we have and makes it significantly more difficult (increasing times to do simple things like write a progress note from simply opening the notes and writing something down to 10+ clicks and multiple sources of possible errors). It appears that we have been taken for a ride and someone has made an awful lot of money off us. 
---- End quote:
Now while I am not sure anyone should be attempting to develop a major hospital EHR from the ground up (as I think that would be a very courageous decision) it is clear that, given the apparent success we have seen here, that some very poor selection and implementation decisions have been made - as the very least in the way clinicians have been treated.
SA Health would really seem to have got into a lot of trouble with all this:
You can read the view from the SA Government here:
The comments following this blog fit very nicely with what is written above - so sounds like we have a disaster in the making.
Sad about that. I hope there are some serious efforts being made to recover!


Anonymous said...

Many of these systems should be able to be built for a fraction of the cost but you need to find a senior software engineer with experience, allow them to build a team and deliver over many years with a minimum of political interference.

There are people around capable of doing it, but an environment with minimum political interference is the problem. They would probably appoint a non software person as the head and produce glossy brochures and over promise while paying little attention to the technical aspects. The ability to give smart people the resources to do their work and stay out of their way is whats missing. Dilbert tells you this. If they actually read "Peopleware", a management book written by the creator of dilbert rather than using generic managers it could work.

Alas, instead we have NEHTA and The Queensland health Payroll debacle and we are off in the weeds, far from the track. Our only hope is a cliff.

Anonymous said...

Many of these systems should be able to be built for a fraction of the cost ........

That is easy to say BUT:
[a] an R&D environment is required
[b] with intelligent leadership
[c] adequately funded
[d] inoculated against bureaucratic committees
[e] plenty of lead time [years]
[f] a greenfields site of significant size 400 beds
[g] time to architect, design and develop before the hospital opens

Questions: Who would be prepared to undertake such a high risk venture? Which hospital, presumably under development; would be prepared to be the guinea pig? What would the business plan reflect in terms of future business growth? Which State Government would embrace such a proposal?

K said...

More importantly, failure must be an option, else the budget blows out by a factor of about 20 (and failure is more likely). Hence why vendors, and why vendors charge - because failure is an option for them, and one that they get to experience too enough for the investers

Bernard Robertson-Dunn said...

Anon said: "Many of these systems should be able to be built for a fraction of the cost..."

IMHO, it's not the building of the system that's the problem, it's knowing what system to build.

And knowing what system to build is all about information; not software, not technology, not products.

Anonymous said...

What Bernard says is absolutely correct however the observations and questions raised by September 08, 2014 10:37 AM still need to be answered.

Anonymous said...

Orion has put its hand up.

PulseIT - Orion plans to take on the big guys in health data revolution; Written by Kate McDonald on 05 September 2014.

Orion Health has outlined its plans to overtake the global health IT giants.

Anonymous said...

Given the billions spent on eHealth by the various levels of government it would be sensible to allocate $20M over 5 years to actually see if it could work. A state government could easily do that, but its to sensible an idea to actually happen. It would be reasonable backup plan to what they are doing now which is why its unlikely to ever happen.

Anonymous said...

Dear September 09, 2014 2:34 PM - $20M you say over 5 years .... you must be joking --- the bureaucrats and their political masters think they have done a terrific job after spending over $1 Billion on NEHTA --- so your $20M pittance would get laughed out of court would it not?

Anonymous said...

Less is more - quite often, in my experience. At $4m a year you could assemble a crack team of about 10-15 people and deliver a good system. A team bigger than that is a problem.

Unfortunately the consulting company con artists have these "leaders" right where they want them and are milking the system big time.

If its such a pittance they should try it for a laugh. I know many hugely successful systems were built for far less than that - with the right team. The point is you need an experienced software team with pre-existing standards and healthcare knowledge and pay them appropriately. What you don't need is executives.

Anonymous said...

But if they did assemble that 'crack' team could the hospitals and bureaucrats wait for the 'new' product to be developed, then tested. On no people say, it would take too long, we must get something now, best to get something off-the-shelf that's tried and proven to work and avoid the risk of a new system being developed which fails. That's the message they all believe, that's the snake oil salesmen in the big consulting firms and large software vendors hard at work selling the same message. How do you intend to counteract that compelling spiel? And while you're at it which hospital do you plan developing the system for with your 'crack' team of unproven experts?

Anonymous said...

The place to develop it would be a place with clinicians like the one who wrote the email, so they could be involved from the start. It would have to be developed in an agile manner. I know the bureaucrats will never let it happen, but that just points out how far off the track we actually are.

K said...

If you think it's a good idea, put your own money on the line, and fund it yourself. Then, you'll have a better application, and you can sell it at a fraction of the cost of the established systems, and everyone will be happy, except for the incumbents.

If you don't want to put your own money up, stop asking for our tax money to be spent on something you won't spend yourself.

More generally, many enterprises used to have their custom built systems, but these eventually folded because adapting off the shelf systems was a better option - this is pretty much in all industries. If you people making these comments somehow think that health is magically different, why do you think that? What economics suggests that solutions that don't work elsewhere will magically work in health (because I've seen the opposite: what works elsewhere magically doesn't work in health, cause nothing will)

Anonymous said...

I already have put my own money up, but that doesn't stop bureaucrats wasting my tax money. The concept of an inhouse development team might be out of fashion, but are we seeing the benefits of outsourcing at a government level, I don't think so. Sometimes you have to go back to the future.

Patto said...

I think what is interesting in health is how much the system works against entrepreneurship and innovation. And don't get me wrong - there are very good reasons why people are risk averse in health - the fact that faulty software can actually kill people is a big one.

But when you have tenders go out for hospital systems - with requirements that the software has to have been installed in production in multiple other hospitals within Australia - don't be surprised that you don't get innovative new entrants into the market.

If there was a juicy enough carrot awaiting, I'm sure a third party company (start-up?) would actually plonk down millions of their own cash to develop hospital systems - but I'm not sure any of them would have confidence that any state government would ever take a punt on using the system they build. We have start ups that burn millions of dollars working on lame versions of Facebook 2.0 - what is it about health (software) where we don't seem to have any innovation at all?

(charm had a good approach which was to focus on a niche (oncology prescribing) and then expand out from their to general EMM..)

K said...

Patto, indeed - now we come to the real issue, and why trying to get the system to take it's own risks is silly. I do think that the growth of what we loosely call the internet platform upsets the economics, and that there's a possibility for a new way to develop apps based on open source.

But the existing open source applications have such a long way to go, and we have to figure out how the eco-system could fund the painful boring stuff of doing integration for Australian usage - a community good that would cost any one player too much to do.

The bigger problem is that health itself is about the disrupted completely by externally driven change, and the system will fight back against this as hard as it can, all the way to the top. Resistance to innovative ways to build non-innovative applications is a sideshow.

Anonymous said...

Surely this will all soon become a thing of the past now we have clinically validated data, some test stub and scenario's.

Looking forward to the snake oil sales convention, maybe that can fill the IT14 gap on the eHealth calendar