How to improve the value of health care without spending more

STAR, the Socio-Technical Allocation of Resources, brings commissioners, providers and patients together over the best available data on what each treatment costs and how much health it buys. Gwyn’s team at LSE developed it; he calls it applied common sense. In Mid-Nottinghamshire, a study of back pain care found the cheapest treatment bought a year of healthy life for £124 and the most expensive for £220,612.

The same test applies to preventing type 2 diabetes. Programmes that help people at high risk eat better and move more are good value, at about £7,500 per year of healthy life gained. Paradoxically, they will do little to change the projected epidemic, because they reach only people already picked out as at high risk: the economic review put their effect on new cases at between 0.1% and 1.6%. De Poli, Bevan, Keng and Pizzo reached the same conclusion when they simulated the programmes for England. When they examined the projections themselves, they found the models disagree: for England in 2025, the figures ran from 3.95m people with diabetes to 9.07m.

Rates of treatment vary between regions and between hospitals by more than differences in illness can explain. Some of that treatment does patients no good, so cutting it saves money and spares them the intervention. With Appleby, Raleigh, Frosini, Gao and Lyscom at the King’s Fund, Gwyn sorted the differences between places into the good, the bad and the inexplicable; when the NHS began publishing its Atlas of Variation, he and Schang, Morton and DaSilva asked whether the people who hold the money act on it. With David Goodman of Dartmouth, Gwyn co-founded the Wennberg International Collaborative, a community of scholars and policymakers that supports emerging investigators studying unwarranted variation in their own countries.