The methods for best caring for people with severe mental illness are in the dock, but the evidence on assertive outreach and personality disorder is sadly lacking, reports Laura Donnelly

Mental health services should be built on 'what works', says the government. 'Who knows?' was the shrugged response of delegates at the annual conference of the team for the assessment of psychiatric services (TAPS).

'History has taught us - belief is no substitute for evidence,' warned Dr Robert Kendell, president of the Royal College of Psychiatrists as he drilled home the lack of proof surrounding personality disorder.

The government's review of the Mental Health Act is due to publish its findings within weeks - fleshing out plans to extend future mental health law to cover the controversial diagnosis.

But Dr Kendell said determining the best environment in which to detain people with personality disorders would need a 'properly designed randomised controlled study' - and that would take at least 10 years.

He conceded that the government's 'primary motivation to protect the public' meant that the issue of what could be done to therapeutically benefit someone with a personality disorder was 'in political terms very secondary'.

But he appeared to doubt the likely impact of legal changes on public safety.

And he cited work submitted by the Royal College of Psychiatrists to the Home Office mental health unit which suggested that 'to prevent five serious crimes you would probably have to lock up 50 to 100 people for several years on end'.

The strategy's emphasis on safety came under sharp focus again from Dr Richard Ford, head of evidence evaluation at the Sainsbury Centre for Mental Health and senior policy adviser to the Department of Health.

Dr Ford outlined a decade's research on assertive outreach - a key element of the mental health strategy. 'There is absolutely no evidence as to what impact assertive outreach has on safety. I personally remain a little bit sceptical on that point.'

And he said research remained 'rough and ready' on which individuals were appropriate for such programmes - and how many it could extend to.

Figures on the number of people defined as 'having a severe mental illness and difficult to engage' varied between 14 and 200 for every 100,000 in the adult population.

While broadly welcoming of the government's policy focus on assertive outreach, Dr Ford flagged up its limitations. He also warned: 'I think there is the possibility that assertive outreach has been oversold as a fix.'

Dr Ford outlined a 'systems approach' under which needs identification of service users came well ahead of the development of service structures.

The need to develop responsive services would mean that no single service model would fit all, he said.

But Dr Ford admitted that the current 'explosion' of the policy, with half of all health authorities currently setting up assertive outreach schemes, or having them already in place, could threaten that approach.

'There are risks,' he told HSJ. But he said guidelines on the setting of teams had highlighted the necessity for needs assessment 'which could be done fairly quickly and should be ongoing anyway'.

Earlier, National Schizophrenia Fellowship chief executive Cliff Prior outlined the 'almost impossibly difficult process' of developing plans for a national service framework - due to be published shortly - which he had worked on.

As chair of the framework's external reference group on long-term care, Mr Prior was one of 40 members battling with 30 Department of Health staff to produce final proposals, which went to ministers in February.

But Mr Prior admitted that the group's proposals had 'rather less' focus on evaluating treatment than had been anticipated - 'with rather more than might have come of the National Institute of Clinical Excellence than expected'.

The debate on research flared up once more when David Taylor, chief pharmacist at the Bethlem and Maudsley trust, took on the topic of novel antipsychotic drugs.

Though fiercely critical of much of the research about the industry, Mr Taylor played up the benefits of new atypical drugs in addressing problems such as 'revolving-door patients' less likely to comply with medication.

But he criticised 'silly prescribing' by psychiatrists unaware of the differences between similarly named drugs and the 'co-prescribing' of new with old drugs - 'giving patients extra side-effects and really annoying your health authority'.

Condemning the 'rosy picture' Mr Taylor had painted of the new drugs, Dr Kendell highlighted weaknesses in research surrounding the area. He insisted that older-style drugs - with average monthly costs as low as£5 per month compared to£215 a month for some atypicals - should always be prescribed first time.

But Mr Taylor was backed by consultant psychiatrist Dr Paul Rowlands of Chesterfield and North Derbyshire Royal Hospital trust.

'I think sometimes we only have one bite at the cherry,' he said, pointing to the higher engagement rates of patients prescribed atypical drugs as a first response.

Chair Dr John Carrier of the London School of Economics told delegates that when Mr Taylor had finished, 'I thought the jury was in' on the side of the new drugs. At the end of the debate he admitted: 'The jury is still out' - which rather summed up the day.