Over-complex IT results in diminishing returns, says Andrew Sakals
I have always disliked people who say “I told you so”, but I find I am turning into one of them when I read about the controversies and travails surrounding the NHS National Programme for IT.
I remember telling my Labour councillor that the idea of a Leeds resident falling ill while holidaying in
After hearing that the NHS system was going to cost £6bn, I told my councillor that we already had systems available to send medical records, ie. by express delivery, fax or e-mail, and that the probability of this being necessary for most people was very small.
The Ford Cortina solution
I do not have the statistics, but I believe that the number of people who claim for medical treatment on their holiday insurance is fairly low. My family have been going on holiday for over 30 years, and have never had to avail ourselves of emergency NHS treatment while outside our local area.
In my early career as a systems analyst in the 1970s and 1980s, we often designed and implemented very adequate Ford Cortina-type systems that users found easy to use and that dealt with 99% of their regular cases. The rest were treated as special cases to be handled manually or with a system work-around.
To have gone on to develop systems that handled every eventuality would not only have made the systems unnecessarily complex, but would also not have been cost-effective. The users never wanted a “Rolls-Royce” system, just one that would make their lives easier. Do these basic principles no longer apply in today’s IT development projects?
A problem I found with some IT consultancy firms was that in their eagerness to please the client they would sometimes promise the earth, even if what they promised was technically unfeasible. Another crucial factor with any development project was the level of commitment of the users; in NPfIT’s case, this would be the medical profession.
What doctors want
I suspect doctors would find it difficult to defer to non-medico IT consultants telling them what data to collect with every patient visit. Such data administration activities are unlikely to make the doctor’s life easier and would detract from the time they would like to spend actually dealing with patients. What happens when the system goes down? The only chance of the system being a success is if the impetus comes from the doctors themselves.
I am surprised to find out that MPs are only just beginning to talk about investigating the feasibility of the system, as reported in Computer Weekly, 30 May. It begs the question, how much money has already been spent on a system that may not even be feasible?
Andrew Sakals is an IT consultant who has worked on projects for Bank of America, Pacific Bell, SwissRe and Norwich Union
Have your say: Do you disagree with Andrew Sakals? If you have an opinion about any article in Computer Weekly, e-mail firstname.lastname@example.org