Tuesday 10 September 2013

NHS IT: time to try another way

Many years ago, right back in the mid-to-late 1980s, I was editor of a wonderful magazine called Town & Country Planning.  My successor has done an excellent job ever since, but I have always missed it.

I remember, in those days, how many reports used to cross my desk about the estimated value of private sector housing dilapidations - the total amount that was required to repair the UK's housing stock.

One of the peculiar side-effects of the Lawson-inspired house price inflation which hit us about the time I moved on has been that we don't get so many of those reports any more.  A combination of equity in the home and a nearby DIY store has kind of tackled the problem.

Here is the point I'm making.  In the end, the problem wasn't solved by major investment by central government, as the authors of the reports tended to assume.  It wasn't solved by a new government agency sending out approved designs.  It was solved by tens of millions of ordinary homeowners, and especially young ones, going down to the shop and buying paint - and spending their weekends using it.

I simplify, of course.  But sometimes the top-down solution isn't the best one.  Often it isn't possible or  affordable.

I have been thinking about this and how it relates to the perennial problem of NHS patient information, and how you tackle the parallel problem of lost notes.

We can be reasonably sure, after £12 billion went down the drain last time, that the top-down approach does not work, though that will not stop them trying again.

The influential NHS blogger Roy Lilley tackled the issue again a few days ago, and I was distressed to hear that a new attempt is being made. Unfortunately, the new announcement makes horribly similar noises to the old ones.

But Lilley also pointed the way forward towards the equivalent bottom up, DIY solution.  St Helens and Knowsley Foundation Trust have a scheme called 'bring your own device 2 work', which allows all the staff's variety of systems to talk to each other - but not allow information out beyond them.

Last year, I met a pioneer of a similar bottom up solution that I believe will soon be widely adopted.  Patients Know Best is a social enterprise started by a tech-savvy doctor called Mohammad Al-Ubaydli, and chaired by the distinguished former BSJ editor Dr Richard Smith.

It requires us to turn our idea of patient information on its head.  Instead of the NHS owning the information about you, and getting itself into a terrible mess dealing with other agencies, and constantly asking permission to share information, PKB suggests that you should own the information yourself.

It requires a simple piece of software, and it means that you can give or withdraw access to your own information to whichever professionals in whichever public services you need.

Great Ormond Street adopted it for their patients some time ago, and the PKB approach is now spreading through GP surgeries in Kent.  There is no reason either why it should be limited to health.

It requires no vast investment from IT consultancies.  It doesn't require taxpayers to lose another £12 billion organising one inflexible centralised system.

It also fosters the kind of equal relationship between patient and professional that the NHS badly needs.  It also, and here is the real reason for writing this, is the kind of entrepreneurial approach to government described in a fascinating American article about 'agile public leadership' (thank you, Ted, for sending it).

It is characterised by US chief technology officer Todd Park as “think big, start small, scale fast”.  That is precisely the PKB model, rather than the lumbering top-down brontosaurus approach to IT that Whitehall still favours far too often.

1 comment:

Matthew Buck said...

Thank you for the excellent blog and posts including the one above. I attach below fyi a link to a recent overview of the NHS and the tendency in what remains of the national media reporting to record only problems not potential answers to the challenge of change.