The budget of Britain’s Health Technology Assessment programme has grown from £13m in 2006 to £88m in 2010, and it has conducted a swathe of trials on new technologies, published dozens of papers, and supported a study that won the BMJ paper of the year. But could it do even better? This was the question addressed at its biannual meeting yesterday, which I chaired.

Liam Donaldson, England’s chief medical officer, kicked of the meeting, and gave a “traffic light analysis” of how Britain is doing with the kind of research that is variously labelled health technology assessment, public health research, or health services research. It was green for studies of the distribution, causes, natural history, treatment, and prevention of disease, amber for studies of process and outcome and of patients’ views and experiences, and red for research into models of care.

Then he revisited a list of the “hard nuts” to crack in public health research that he’d compiled years ago. Britain has made progress with increasing the number of world class researchers in these subjects — but is still behind the US. The research was largely disparate and descriptive-and still is, although better. Good models of care remain hard to spread, and, he said (choosing his words carefully), NHS leaders and managers are still not as evidence literate as they might be. Big system changes are still introduced without adequate attention being paid to evidence, but piloting big change is, he said, “tricky.” And the “hardest nut of all” is influencing mainstream clinical activity.

When I came to pick out the good and the bad of the day, I saw lots that was good. One of the most pleasing things was the increase in publicly funded trials, something that many have wanted for a long time—not least to counter the bias of industry funded trials. There were 30 publicly funded trials in 2004 but will be about 100 by 2011, 80 of them funded by the health technology assessment programme.

The last issue of the BMJ that I edited contained several articles on the nightmare of trying to get ethical approval for multicentre studies, and there has been a huge improvement in this problem—although difficulties persist. And sadly a new bureaucratic barrier to research has been developed: ironically, research governance seems to be one of the main barriers to research.

The programme has done an excellent job of including “service users” or patients in the planning, prioritising, designing, and carrying out of research studies, and their voices were prominent in the meeting.

The highlight of the meeting for me was the session on “failures,” where Marion Campbell gave a marvellous presentation on her long drawn out efforts to do a placebo controlled trial of knee arthroscopy for osteoarthritis. Placebo controlled trials of surgery seem to be a step too far for Britain, but the whole fascinating story of trying to do the trial  is written up in the journal of the programme.

When it came to suggesting improvements for the programme I worried that their conceptions of both technology and evidence may be too narrow. Most of the trials are about drugs and devices not models of care, but this is perhaps because other parts of the National Institute for Health Research are responsible for such evaluations. But there seems to be room for more different forms of evidence, including more observational studies.

Although the programme has done well at including patients, it seems to have scope to improve its interactions with managers and commissioners, people who are most able to improve the overall system.

And with the dramatic cuts that are coming in the NHS the programme will have to do as well as it can at proving value for money, which has to mean improvements for patients and in health not just publication sin journals. Similarly there is probably more work to be done on dissemination, making sure that the results of studies reach the people who are most able to implement the results and make change happen.

But being an old dog, I can remember a time when the NHS had almost no research capacity. The National Institute for Health Research is a huge improvement.

Competing interest: Richard Smith chaired the meeting but wasn’t paid—and he had no expenses because he cycled to and from the Royal College of Physicians. He did get a free lunch and several cups of coffee, which may have influenced him as his price is low.