Lunch was an apple. Also, chats with Ken (who I’d like to thank for looking out for me today), Ian and Ewan. Then back to the dungeon for a talk about EHRs from Stephen Kay. I enjoyed his assertion that “world domination” is the natural goal of EHRs. Basically, he seemed to be saying that there’s little evidence of genuine benefit for doctors or patients in using EHRs, but they are nonetheless inevitable. Stephen is involved with Eurorec, an organisation which is attempting to promoting high-quality data collection in EHRs and offering a certification process to recognise that quality. There was one line I intend to steal and use as my own: “you can’t put people in silos without damaging them, and the same goes for data”.

The second afternoon speaker was Scott Cunningham talking about the My Diabetes, My Way website. My impression was that this is a well-designed project exemplifying what you can do when all the complexities of the backend stuff have been taken care of. It was interesting to note that Scott was the second speaker to mention the need to “validate external content” as part of the service. Another interesting technique is that when the patient is initially setup to use the website, the first batch of data is validated by a human being – in this case a trained nurse – to make sure it’s up to scratch. I think that’s an admirable precaution, but it starts to look like quite a high-maintenance system. Authentication is provided via the Scottish Government’s new “Citizen’s Account”, an initiative I hadn’t heard about before.

The next two presentations were basically sales pitches for “portal” technology. Once again I felt that key pre-requisites were being glossed over. All the various systems seem to propose using matching algorithms (for patients) or lookup tables (for other entities such as lab tests and drugs) to create links between the different data stores. This is high-maintenance approach (just look at NHS Scotland’s experiences keeping SCI Store working). If probability-matching is the only way to do this, then OK, I guess we’ll have to do it. But we should at least acknowledge that’s what we’re doing and go in with eyes open.

Another issue that I feel has been glossed over is authentication. It’s all very well to say that there will be a single-sign-on point of entry. But how are you going to make sure that the person is who they say they are? Sharing of login identities is a fact of life in NHS clinical environments, and rightly so since the time spent switching identities is not acceptable. SSO is not going to solve that problem, so audit information quality will continue to be compromised.

After the last speaker, I met up with a couple of fellow Twitter users – @markhawker @psweetman – and had a lengthy, highly-informative chat over dinner. (Didn’t attend the official dinner: sadly my NHS-employee maximum expense of £15 per night wouldn’t have covered the £25-£35 cost of the official Channnings meal…)

Forgot to mention in the previous post that Nigel Fitzsimmons kindly put me at ease by chatting to me during the registration session first thing, while I tried to cool down by the open window. Thanks for that, Nigel.