BMJ  2005;330 (26 March), doi:10.1136/bmj.330.7493.0-g

Editor's choice

Heading where exactly?

"The hospital care was spread over three sites, delivered by six teams and by numerous members within each team, while the information passed to her GP was patchy." So runs Craig Gannon's sobering critique of the fragmented care delivered at his hospital to an elderly woman whose death from unaddressed renal failure could have been averted (p 737). How many such stories lurk in the corridors of today's NHS, and what do they tell us?

Gannon says they tell us that, while individual clinicians are doing what they are asked, the UK's new tick box, target driven culture means they are doing no more than is required. People are working in silos and without clear ownership of patients by a lead clinician, so there is no continuity of care. If this is, as Gannon implies, a result of recent reforms, what can we expect from the next phase of the government's NHS improvement plan unveiled by Nigel Crisp last week?

Choice, we are told, will be the engine of the new patient led NHS. Patients needing elective procedures will choose from a menu of hospitals, including those in the private sector. A fixed national price for each procedure will mean that GPs and their patients make choices based on quality, not price. Payment by results will mean that hospitals get the money only if they get the business, which will encourage them to do a better job.

Won't it be wonderful if it works? Better care is after all what we all want and it's hard to argue against choice. But a report also published last week (p 691) voices concerns shared by many. Firstly, choice is not everyone's cup of tea: apparently two rather large sectors of the population (men and people over the age of 55) prefer, in general, to leave decisions to their doctor. Private operators could skim off less difficult cases, leaving the NHS to deal with anything complex. Hospitals that patients choose not to use could close, which may not be in the public's best interests. And rather than increasing equity as the government hopes, choice could unfairly advantage those better able to access and interpret information about a hospital's performance.

And what types of information can properly hold health services to account? Quantitative measures like waiting times or death rates somehow adjusted for case mix? Or some qualitative measure of continuity of care? Not at any rate patient satisfaction, argues Barbara Stocking (p 736). Instead, well tested tools—such as simply asking patients whether their health is improved after care—could give a rational rather than a preference basis for decisions. Christian and Lise Lotte Gluud call for evidence based diagnosis (p 724), but where is the evidence for NHS policies? How do we know that they won't result in more cases like the one described by Gannon?