Sounds like things aren’t going well for the Oracle team at Cerner?
Reasonable market share does not mean under the hood it’s modern or robust from data interoperability or operation approaches. There’s always fundamentally a problem with how vendors engage with sustainable software procurement. The license for procurement does not cover enhancements and there will always be enhancements, so how does procurement align with local changes. If Cerner decides not to enhance when workflows change in the business, it doesn’t serve both options well, but it’s similar to CAPEX vs OPEX equation on how a platform can operate.
None of these big systems is ‘up to date’ under the hood.
Cerner has a massively denormalized database with between 5k and 10k tables because of their long-term policy of acquiring and bolting on; they can’t fix this.
It’s major competitor Epic works on MUMPS, which was invented pre-Codd and has no concept of normalization at all! As I understand it, they churn every evening, turning the b-trees into something that looks like SQL—with about 16k tables.
Go figure!
Dr Jo
2 Likes
I saw that running in Ireland when I worked there - MUMPs that is - they were using it for vaccinations in Wicklow I think… this was in 2016.
Birmingham’s home grown EHR is also MUMPS based. They really struggle to find developers who can work on it due to it being VERY legacy / niche.
The R&D investment by these companies is significant. It’s a constant race to stay contemporary. EPIC reinvests 50% back into R&D.
From having spent a fair chunk of my career implementing these systems, they are technically as good as the big ERM systems such as SAP. Functionality and feature rich, highly scalable, cloud based, with open API integration. Arguably as strong as banks in terms of robustness, with high availability and reliability to match.
I’ve also not encountered problems in being able to adapt these systems to match clinical workflow. Probably the biggest issue in my experience is the sheer amount of configuration that is required to meet the needs of the different medical disciplines, e.g. ED, Theatres, Outpatients, Community, Renal, Matenity, Oncology, Allied Health etc are all very different, additionally the workflow needed for a large hospital is different to the workflow needed for a small hospital, (resulting in many workflows with hundreds of business rules and many design decisions).
In my experience the technology used is not the limiting factor. As with most tech projects the amount of effort that goes into the planning, design, build, test, transition, and ongoing support are generally the limiting factors, (which is easy to underestimate when you are replacing the core systems used across many disciplines with a single ieMR).
1 Like
I’m not taking sides here - just saying that there is usually more than one side to this type of story…
Thanks for posting Peter. Having helped design and implement an enterprise grade integrated EMR into multiple hospitals, it is really disappointing when you see lazy uninformed journalism like this. It’s a real disservice to patients and staff, (which that reporter is sure to be at some time).
Another similar example I heard recently was a senior politician saying he would not consider spending public money on enterprise grade EMRs as he had a terrible experience with them when he was working as a doctor. The important bit of context was his experience using an EMR was 20 years ago. It would be like saying I don’t like operating systems because windows 3.1 was lacking in functionality and had a rubbish GUI.
The only good part from the looming heath care spiral, (recent report for Australian Parliament says 46% of the workforce will need to be working in health care by 2050), will necessitate leveraging technology as opposed to just throwing more resources at the front line.