Choice GP

A health planner's delight?

This week's issue should please health plannersfor it shows that spending more money on health care doesn't produce better health, that simple interventions can be more effective than complex ones, and that an intermittent treatment may be just as effective as a continuous one; against that, it shows that giving feedback on prescribing to general practitioners may have no effect at all.

For people who argue that all the NHS needs is a bit more money, there is a natural experiment to test that assertion within the British Isles. The NHS in Scotland, Wales, and Northern Ireland receives more funding per head of population than it does in England, and Jennifer Dixon and colleagues have looked to see what the Scots, Welsh, and Northern Irish do with their extra money (p 522). Theirmodestconclusion is that funding the English NHS at the same level as the Scottish NHS would "make life easier for providers."

The clinical messages start with a randomised trial from Norway showing that pelvic floor exercises are more effective in managing genuine stress incontinence that either electrical stimulation or the use of vaginal conesand have no side effects (p 487).

They continue with a pragmatic trial based on the observation that, even though long term maintenance treatment is recommended, treatment for gastro-oesophageal reflux is often given by doctors (or taken by patients) in short courses. A multinational European study group has thus performed a randomised trial of intermittent treatment with omeprazole or ranitidine in patients with uncomplicated symptomatic gastro-oesophageal reflux disease (p 502). Half the patients were effectively managed with intermittent treatment, the most important factor being the response to initial treatment.    GP choice

Lots of effort goes into influencing general practitioners' prescribing, and on p 507 Dianne O'Connell and colleagues from Australia assess the effect of feedback as a single intervention. Doctors were randomised to receive or not receive graphical displays of their prescribing compared with rates for all the doctors in the trial, together with a newsletter. There were no differences in subsequent prescribing between the two groups, and the authors conclude that this type of feedbackmailed, unsolicited, centralised, and with aggregate datais useless because it does not include elements that are important for creating change, such as timeliness, local ownership, opportunities for discussion, realistic alternatives, and incentives.

Acknowledgments

To receive Editor's choice by email each week subscribe via our website: www.bmj.com/cgi/customalert