How scrutiny of freely available data might save the NHS money
THIS week Britons were reminded yet again of the strains on the government’s finances. But another resource—data—is in abundant supply. Like governments in many other countries, Britain’s is turning more and more of its trove of information into “open data”: downloadable at no charge, arranged in neat rows and columns, to be sifted, manipulated and combined with other sources by anyone who is interested. All sorts of nuggets can be found in this informational ore—including ways of making those scarce financial resources go further. A new analysis by one group of data-miners suggests that the National Health Service could be spending a good deal less on drugs.
Part of this team, a start-up called Mastodon C, is housed at the Open Data Institute (ODI), a non-profit company set up by the government, which celebrated its official launch this week. The ODI is due to get £10m ($16m) over the next five years from the Technology Strategy Board, a government agency, to which it expects to add some private finance: this week it welcomed its first slug, $750,000 from Omidyar Network, a philanthropic investment firm. Although lots of countries are opening up all manner of official information, none has anything quite like the ODI. The institute’s purpose is to foster the exploitation of open data—not least by the sort of young tech businesses that are clustered around its office in Shoreditch, in London. Mastodon C is one of four infant firms being incubated under the ODI’s roof.
In collaboration with Mastodon C, Open Health Care UK (another start-up founded by a doctor and a programmer) and Ben Goldacre (a doctor and critic of the drugs industry) have examined a vast open data set: the prescriptions written by every family doctor in England. They looked at regional patterns in the prescription of statins, cholesterol-lowering drugs intended to reduce the risk of heart attacks and strokes (see map).
Statins are expensive. In 2011-12 the NHS in England shelled out more than £400m on them, out of a total drug budget of £12.7 billion. Some are much dearer than others: patented ones can cost 20 times more than generic versions. For this reason the NHS’s prescribing guidelines say that unless doctors have good clinical reasons for choosing an alternative, they should prescribe simvastatin, a generic product, at least in the first instance.
Yet the analysis by Mastodon C, Open Health Care UK and Dr Goldacre shows a good deal of local variation in prescriptions. Between September 2011 and May 2012, 35% of statins prescribed by doctors in Shropshire were for branded drugs, against just 8% in Hardwick, near Derby. Although doctors may have good reasons for giving some patients pricier treatments, this sort of gap is difficult to explain on solely clinical grounds.
The researchers estimate that, had every doctor prescribed cheap statins, the drugs bill would have been more than £200m lower. (They could have been more precise were the prices paid for drugs by the NHS also open data, but they are not.) This particular problem may be abating: Atorvastatin, which accounted for about two-thirds of spending on statins by the NHS in England last year, has since come off patent—and its price has fallen by 90%. Even so, the variation is fascinating—and the exercise is well worth repeating for other classes of drugs.
Carl Reynolds, the doctor at Open Health Care UK, says that the team plans to do exactly that. When spending is too high because branded drugs are prescribed instead of generics, money is being wasted that could have been used to treat other patients. A study in the British Medical Journal in 2010 reckoned that the NHS could save more than £1 billion by switching from branded drugs to generic equivalents. The new analysis backs up that claim.
The purpose, say the researchers, is not to lambast individual doctors for squandering scarce public money. But the results do pose them a pointed question: could your prescribing be less costly? “As a clinician,” says Sir Bruce Keogh, the medical director of the NHS, “it’s quite easy to think you’re doing the same as everybody else. It’s not until you see the data that you know whether you are or you aren’t.”
This article appeared in the Britain section of the print edition under the headline "Beggar thy neighbour"
