2011; The Arab Spring, a Tsunami in Japan. Riots in London, following peaceful protests over the police shooting of Mark Duggan and Apple’s founder Steve Jobs died… and…
…the year IBM’s Watson Computer, developed as a natural language answering system, appeared on the US quiz show, Jeopardy and won…
The implications were profound. Science, mathematics, space travel and healthcare.
What if the machine’s learning could answer questions, diagnose, predict, rationalise resources, drug development… if you can imagine it, it could happen.
Except… it didn’t.
I’m pretty hard nosed when it comes to technology.
The dreadful state of NHS data capture, analysis and all the rest is bad enough but when you hear stories (as I do, regularly) that it takes more time for a doctor to log-in to a ward system than it does to attend to the patient… all is lost.
I know my scepticism is right.
I take the view that the front-line of health care has, consistently, been let down by so-called IT experts, who’ve frittered away millions and still can’t explain why an A&E doctor in Bournemouth can’t connect, live, to a patient’s primary care record in Birmingham.
For heaven’s sake… I can take money from a cash machine, at three in the morning, in Timbuktu but I can’t decide who I share my health records with, at lunchtime at the pharmacy.
That said… I became a fan of Watson.
The idea that they could gather medical learning on a global basis, distil best practice and answer questions in plain English was… well, mind boggling.
It’s not difficult to do, if…
… you have big enough storage and the computing power to access it, quickly.
Speech recognition? Well, it probably sits on your coffee table and is called Alexa. Putting it all together and creating an application that has use and value, is the very smart thing.
Even I was persuaded to expect great things. No, wonderful things. No amazing, fantastic things!
Actually, Watson never really worked and has recently been broken into component parts. The data sold to a private equity firm for a $billion.
A financial disaster for IBM.
Just as there is no connectivity worthy of the name in the NHS, there’s no global connectivity in healthcare. No way to collect an holistic overview of disease across the world.
A heart attack in HongKong, Holland or New Haven… is just a heart attack but the why’s and wherefore’s, the diet and life styles that may have contributed to it, the treatment and outcomes… such vaulable information to learn from.
What works fastest, safest and is the most economic. The Holy Grail questions in healthcare.
Watson couldn’t do it. Like I doubt any of the new ICBs can do it, from one side of town to the other.
There much to gain. Microsoft, Google, Amazon… all recognise the importance of health informatics but none of them have a really good handle on it.
Knowledge is power but in the right hands it is democratising. A way to leverage poorer nations with threadbare health systems. Accelerate their training and knowledge base.
Machines that can tell a doctor what the diagnosis is, suggest optimum treatment. Global knowledge, sliced, diced and served up like a Sushi chef.
IBM bet-the-farm on Watson. They spent $5bn buying-up health data companies in pursuit of patterns and clues to a healthier future.
There were problems with the complexity of patient files. Problems with the socio-demographic impact on disease, climate and determinants of health. Problems that optimum treatment recommendations were not universally available.
In many cases Watson told clinicians what they already knew.
The tech industry is full of swagger and its disciples are audacious. The catastrophe that is Watson is a case study.
NHSE is currently subsuming ‘NHS digital’. I hope it is in the pursuit of the sort of common sense that realises;
• Interoperability is beyond the NHS but mutualised access is good enough.
• Access to health records, Martini style, any-time, any-place, any-where is a mission-critical-safety issue.
• It’s a national health service, that means; buying everyone the same kit is not a procurement sin, it’s an operational crime not to.
• Simplicity, such as blue-toothing observations from the bedside to the patient record saves hours of nursing time and because of accuracy probably lives
• Make it a priority to ease the burden on the front line, by eliminating process inefficiencies. For instance…
• 60’s login is probably responsible for more doctors leaving the NHS than 90’s wages.
It’s all very elementary, my dear Watson.
News and Comment from Roy Lilley
Contact Roy – please use this e-address roy.lilley@nhsmanagers.net
Reproduced at thetrainingnet.com by kind permission of Roy Lilley.
