Prison, a virtual town of 85 000 people, can be an unhealthy place, prone to depression, suicide, hepatitis C and the heart and lung diseases that are caused by smoking. The best estimates show that around 7% of sentenced male prisoners have psychosis and the figure rises to 10% for males on remand and 14% for women prisoners. ^{Reference Singleton, Meltzer, Gatward, Coid and Deasy1} Two-thirds have taken drugs in the year before imprisonment and in a third this has included heroin. A third were drinking heavily in the month before they entered prison. ^{Reference Stewart2}

Lessons from community care reforms

In the past few years important improvements have been made to the healthcare that prisoners are offered. Prison in-reach teams are available to treat severe mental illness in most prisons. Sixty-five thousand people per year are now treated in prison for drug misuse under the Integrated Drug Treatment System. Around 1000 per year are transferred from prison to secure mental health units. Yet the link between poor mental health, substance misuse and reoffending remains strong.

Ten years ago, in reforming community care, we faced a similar set of circumstances. Public concern was high and, although research evidence was patchy, there was reasonable agreement about the nature of the problem and what needed to be done, based on examples of good local practice. We therefore set out to achieve three aims.

1. 1 Early intervention, before the young person, becoming severely mentally ill, embarks on a career of admissions and readmissions.

2. 2 Alternatives to hospital for those who can safely be cared for in another setting.

3. 3 Intensive support on hospital discharge for people with complex needs who are otherwise stuck in the revolving door.

Now, in offender health, the agreement and the good practice are there again and the aims are analogous.

1. 1 Early intervention before the person with mental ill health goes too far into the criminal justice system.

2. 2 Alternatives to prison for people with mental health problems who do not need to be there.

3. 3 Intensive multi-agency packages for people, such as those leaving prison, who are otherwise likely to relapse and reoffend.

Central to the mental health reforms of the past decade have been the specialist teams - early intervention, crisis resolution and assertive outreach - capable of plugging the conspicuous gaps in what we previously provided. Now, the obvious gap is in services working across the divide between the health and criminal justice systems, tackling mental illness, alcohol and drug misuse and homelessness. For those of us who have wanted to make community care work, to prevent the social consequences of severe mental illness, to ensure that people with mental illness are given the standard of care of others in the National Health Service (NHS), offender health is the next frontier.

Six steps to reform

In changing something as large and complex as the NHS and its partner agencies, you need six things. First, you need a strategic plan backed by consensus. One of the remarkable features of recent policy on offender health is the breadth of support that it has attracted across the professions, the voluntary sector and, it is increasingly clear, the political spectrum.

Second, you need leadership, starting in central government and continuing into front-line services where clinical leaders are the drivers of change. People say there is too much power in Whitehall, but for those of us who work there it always feels as if the crucial power lies with clinicians who have the energy and influence to make a policy work (or not).

Third, you need the clinical models, based on evidence - people have to know what to change to. These include models of liaison and diversion from courts and police cells, models of treatment and health promotion in prisons, and models of community support for released prisoners. The National Institute for Health and Clinical Excellence can help get this right by examining the research findings, but given the scarcity of clinical trials in offender health, data from well-evaluated individual services will be vital.

Fourth, you need money, but in days like these it is unrealistic to expect new money. So in addition to the right clinical model, we have to find a plausible economic model. How will diversion save money, not only on prison places, but on secure beds and the costs of courts? And more generally, even if upfront investment produces efficiencies and better care, can it go a step further and lead to actual savings?

Fifth, you need workforce reform - a skilled workforce is in the end what determines patient experience. The clinical care of offenders cannot be the exclusive responsibility of forensic specialists, although their leadership is vital. Offender health needs to be a mainstream issue if it is not to be marginal, and the key skills of managing substance misuse and risk are needed not only in forensic but in all mental healthcare services.

Sixth, you need information - to monitor, measure and provide feedback on progress and benefits. One failing of the National Service Framework era has been the lack of true outcome measurement - we still cannot show that our patients get better. In offender health, the aim has to be more than clinical recovery. It is recovery with a purpose - employment, stable housing, and an end to the damaging sequence of illness and crime.

With all six, you have a chance - not a guarantee - that care will improve. It is then vital to line up our aims with the broader priorities of the government or the NHS. Nothing happens in central government just because it is a good idea - there is no shortage of good ideas. It has to be a useful idea.

A key role for clinicians

So, when the government becomes concerned about the number of people in prison, it is up to us to show that better diversion services can keep some of them out. If the government says there are too many short-term sentences, it is up to us to show that we can offer - or at least contribute to - community alternatives that will address the alcohol or drug misuse that is common in these recidivist offenders. And when the government intends to revolutionise the rehabilitation of offenders, it is up to us to argue that treating mental illness can play its part.

There is an opportunity, too, for us to help government departments work out the best route through some difficult social issues. Where does the balance lie between punishment and treatment for mentally disordered offenders? In particular, what is likely to be the most effective approach to drug misuse? What does the evidence tell us on the use of maintenance methadone?

What more should we as a society be doing for armed forces personnel who are trying to settle back into civilian life? What are the factors that lead some veterans towards imprisonment and how important are poor mental health and alcohol misuse? What should we do to intervene?

How should we ensure the public accountability of publicly funded services? And how, at the same time, can we build those services on professional leadership and evidence? How do we prevent an area like offender health falling victim to public indifference?

Pressure on public sector finances is going to be greater than ever in the next few years and the case for putting precious resources into offender health will have to be skilfully made. But, as in mental health, presenting the argument for better care in a way that stresses value for money is not new to us, although we do need to be better at it. The benefits of improved health in the criminal justice system are broad and profound, on drugs, on crime, on health inequalities - and on the lives of people with mental illness who, through no fault of their own, fall foul of the law.