A lack of government guidance means SHAs and mental health trusts have little clarity in how to examine why service users were able to kill. Investigations drag on and findings from cases as long ago as 1996 have not been acted on. Charlotte Santry reports

Fatal attacks by mental health patients grab headlines, throwing the spotlight onto mental health trusts unused to the glare of publicity.

But public attention rarely focuses on the behind-the-scenes work by trusts dealing with the aftermath of tragic, often shocking, incidents. After someone has been tried in court and an independent inquiry has reported its conclusions about their care, interest tends to wane.

However, an HSJ investigation into murders and manslaughters by mental health patients, published exclusively on hsj.co.uk last week, highlighted the need to track cases long after the inquiry panel has disbanded.

Our investigation revealed that in 27 cases dating back to 1996, action plans drawn up by trusts in response to the independent reports are yet to be signed off by strategic health authorities. This means that some problems have still not been resolved many years after the event, with consequences both for mentally ill people and for public safety.

There is no official timescale for trusts to address failings. The Department of Health's guidance is just five pages long. The 2005 document, called Independent Investigations of Adverse Events in Mental Health Services, devotes only a few paragraphs to how SHAs should commission independent investigations. Rather vaguely, all it says is that they should do this 'as soon as possible' after a killing by a person who has used mental health services in the previous six months.

The dearth of official guidance led Ed Marsden in 2002 to set up his consultancy Verita, to support commissioners through this process.

As the former West Kent health authority mental health lead, in 1997 he had to set up an independent investigation into the Michael Stone case. Mr Stone had attacked Lin Russell and her two daughters with a hammer as they walked down a country lane in Chillenden, Kent, killing Ms Russell and six-year-old Megan and leaving nine-year-old Josie with permanent brain damage.

Mr Marsden says: 'The day that one of my colleagues came in and said Michael Stone was in the care of mental health services, I put my hands on my head. I was responsible for commissioning the report from scratch with no guidance and very little central know-how. It was a long learning curve.

'These incidents don't happen very often and generally speaking organisations don't have much experience in commissioning investigations. That's what prompted me to think public sector organisations would really benefit from support and help.'

Verita's support ranges from training for staff responsible for commissioning inquiries to carrying out investigations on behalf of SHAs. The company's success suggests a degree of insecurity about the task among NHS managers. But Mr Marsden thinks they are better equipped than in his day.

For example he believes, somewhat controversially, that the 2005 amendment permits the NHS to commission independent investigations before or during legal trials. This is significant, as the need to avoid prejudice during police inquiries is often cited as the reason for the long delays in implementing action plans.

'My sense is that quite a number of SHAs still await the outcome of the criminal process, not least because the conviction justifies the independent investigation,' he says.

This tendency to wait and see meant that, in Stone's case, the report of the independent investigation was shelved for six years pending two court appeals and a judicial review. But Mr Marsden does not regret that decision.

'Robert Francis QC, who chaired the investigation panel, thought it would be prejudicial to publish and indeed I think it probably would have been,' he says. 'I think if someone is appealing against a conviction, putting a report into the public domain would be quite improper.'

But Kent and Medway Partnership trust nursing director Peter Hasler, who is now overseeing the case, disagrees.

'The report was completed relatively quickly but held in abeyance for years because of the appeals taking place. The victims' family waited years to see the document, and by the time it came around it wasn't all that relevant.'

Organisations referred to in the report were defunct and named individuals had moved on, he says.

NHS West Midlands interprets the 2005 guidance as a green light to commission investigations before court convictions. Director of nursing Peter Blythin says: 'We want to tighten up the process and make it much more simple. We will move much more quickly.'

Five so-called 'legacy' cases that the SHA inherited from its predecessor organisations after last July's merger are still awaiting action plans, despite relating to incidents dating back several years. These and any new cases will be subject to new guidelines that Mr Blythin is preparing, advising that investigations are carried out at the earliest possible date.

A 1996 memorandum of understanding between the DoH, Association of Chief Police Officers and the Health and Safety Executive fails to clarify the issue, referring only to local investigations carried out internally by trusts.

These internal inquiries can start immediately, meaning substantial policy changes are often in place by the time the external investigation has concluded - as trusts and SHAs are quick to point out.

So why do SHAs need to spend an average of£75,000-£100,000 on commissioning independent investigations?

Mr Marsden says it is because they provide an 'objective, proper, full account' to interested parties'.

He says: 'Mental health is not just what the health sector provides, it ranges across the voluntary sector, social care, criminal justice system. Independent investigations look at the broader issues as well as the narrower issues of what the NHS has done.'

Many people would agree that an independent view is critical in ensuring public bodies are accountable. So perhaps the issue is not whether these investigations should be carried out, but how panels go about their work. Could they be streamlined so that faulty policies and practices are dealt with more quickly?

Mr Hasler thinks so: 'The people working on independent investigations are doing it on top of other jobs. They're psychiatrists, barristers, senior social services directors. This is one of the reasons it takes so long - getting regular meetings is difficult.'

The investigative process will inevitably involve scrutinising professionals, which can only be done credibly by those with similar, recent, qualifications. Any practising lawyer or health worker will have a hectic schedule so the responsibility will add a hefty workload.

Given the time and cost involved in independent investigations, it is important to assess their impact. Close examination of the outstanding cases reveals a striking similarity in many of the problems.

Information-sharing, record-keeping, the care programme approach and risk-assessment training crop up with depressing frequency. This last is surprising, given that it was the subject of a national survey in 2001 co-authored by national mental health director Professor Louis Appleby.

The report How Much Risk Training Takes Place in Mental Health Services? found that training in assessment of risk of harm to others was compulsory for only a quarter of ward nurses and community psychiatric nurses and just under half of junior psychiatrists. It states: 'It appears from this survey that the recommendations on training from the [DoH] and college reports, as well as from homicide inquiries, are not followed in many trusts.'

This was an important factor three years later when John Barrett stabbed Denis Finnegan to death as he cycled through Richmond Park.

The day before, Mr Barrett, who had previously stabbed three people and was obsessed with knives, had been given unescorted leave from Springfield Hospital. South West London and St George's mental health trust identified a need in its action plan to improve systems for assessing and managing risk. Many other trusts are following suit in response to events in their patches.

The problem was revisited in the five-year report of the national confidential inquiry into suicide and homicide by people with mental illness, published by Manchester University last December. Professor Appleby is director of the inquiry.

Key areas it identified as contributing to mental health homicides include absconding, risk management, care programme approaches and ward environment.

The fact that the same problems turn up in investigation reports time and again points up the question of whether information is being adequately shared nationally.

'There can always be more sharing and dissemination of information,' Mr Marsden says.

'One of the real tests for commissioners is to ensure that those lessons that come out of the investigation are learnt and that primary care trusts and SHAs use their performance-management responsibilities to ensure that, along with trusts, practice is changed.'

Investigation process: murder and manslaughter

  • Initial management review - A rapid, usually 72-hour, internal review should cover immediate action regarding staff, safeguarding notes or equipment as evidence, changes in policies and procedures, security, communication with relevant individuals and organisations and initial contact with carers and families.

  • Internal mental health trust investigation - This should establish a chronology of events, determine underlying causes and whether action is needed. This must be available to the independent investigation.

  • SHA independent investigations - These are commissioned by the SHA and review the patient's care and treatment, and risk management.

  • Action plan - These are drawn up by the trust in response to the recommendations of the independent investigation. They are monitored by SHAs. When all the points have been implemented, they are closed.

After an incident: top tips on conducting an inquiry

  • Be clear about why you need an inquiry or investigation.

  • Write terms of reference.

  • Retrieve and safeguard evidence.

  • Make immediate improvements.

  • Support staff and victims. Bring in specialists if necessary.

  • Appoint an inquiry chair and panellists. The credibility of the inquiry will depend on these people.

  • Manage the inquiry process. Strike up a working relationship with the chair and panellists based on agreed procedures, costs and timetable. Identify a senior manager to maintain links with the chair, ensure that the reasonable needs and expectations of the panel are met and manage potential conflicting interests of other investigators.

  • Implement recommendations. If there are recommendations you are not prepared to implement, explain why.

  • Communicate. Decide who you need to tell, what you need to say, how you are going to say it - and when.

To read hsj.co.uk's full investigation into mental health killings, click here