Thursday, April 05, 2007

Another two trips this week (26th March). To Copenhagen on Tuesday for the European Advisory Committee on Health Research. As always, lots to discuss as we struggle with how to strengthen the evidence base for the work of WHO, where it is so important to take account of the context in which one is operating. Thursday it was back to Brussels (I’m getting to know Eurostar far too well). This time it was for a consultation on the 2008 WHO conference on health systems, which will be held in Estonia. It will be 12 years since the WHO European Regional Office discussed this topic, in Ljubljana in 1996. We were struggling with the question of how to measure health system performance and I was giving a presentation on our work on avoidable mortality. As always, Peter Smith (University of York) gave a superb overview, highlighting the challenges involved. His focus was at the level of performance of organisations delivering health care and he provided more warnings (as if they were needed) of the problems associated with public disclosure of performance data. Of course everyone is in favour of openness. The idea is that surgeons or hospitals that are getting poor outcomes will improve their practice if they are named and shamed. Unfortunately, people don’t always behave in the way that you would like them to. They stop operating on sick patients, they change the way they record data, and the net result, at least in one major study, is that patients who do not have severe illness obtain no benefit whereas those who do, do worse, as they miss out on operations that, although risky, might have brought them benefit.
Earlier in the week there was an excellent example of the problems in assessing health system performance. It is increasingly clear that it is almost impossible in many parts of England to register anew with a dentist, because of problems with their new contract. Which Magazine undertok a well-conducted survey in which their researchers phoned up dentists to try to register with them. A health minister commented that ‘The results were “deeply flawed” because they focused on the possibility of dentists taking on new patients, without looking at how many patients they were already treating', conveniently ignoring the fact that prospective patients couldn't care less how many patients were already being treated if they could not join them. Still, thanks to papers released reluctantly by the British Government under the Freedom of Information Act we now know what ministers mean when they describe something as "deeply flawed". This was a term used extensively by ministers to rubbish the paper in the Lancet reporting that the best estimate of lives lost since the invasion was 655,000. As the BBC and others have revealed, what one adviser told them was "The study design is robust and employs methods that are regarded as close to "best practice" in this area” while another reported that the methods were "tried and tested".
I can only hope that my research meets the high standard needed for a British minister to consider it "deeply flawed" too.

No comments: