Yet another significant NHS date is looming. On 1 April the first primary care trusts will go live, many combining the responsibility of commissioning around 80 per cent of the local NHS budget with delivering a range of community services.

Their provider roles can cover areas such as diabetes, elderly care, HIV and now also mental health.

The inclusion of mental health was of course announced in the national service framework, and came as a surprise, since the preferred model of mental health provider, as stated in The New NHS: modern, dependable , was supposedly specialist mental health trusts. Specialist trusts made sense, as they would concentrate minds.

Large-scale redrawing is still taking place, and sensibly much of the reconfiguration is coterminous with local authority boundaries.

Community service components of many combined trusts will be incorporated into PCTs, and leftover mental health chunks are being merged with other chunks.

If there are any concerns, it is the large scale of these new specialist trusts, often covering populations of over a million, and therefore not always in a good position to develop sensitive partnerships with local communities.

Such partnerships are precisely what primary care groups and PCTs are about, and if we want to achieve mental health services sensitive to local needs and local people, PCTs might well offer a valid option. Areas where people distrust specialist psychiatric services may gain from the scaling down of care and the integration of mental health and primary care services. So why am I ambivalent?

It may have to do with the perplexing number of different scenarios: PCGs and PCTs which do or do not commission mental health, and PCTs which do or do not deliver mental health services. All these may even be active within the same health authority and mental health trust boundaries.

How confusing it will be.

PCTs should only be allowed to take on mental health if other PCTs in the area would agree to do the same, but is this likely or even desirable unless there is a pool of expertise in each of these PCTs?

Alternatively, one PCT could act on behalf of the others, as PCGs do now.

But considering a typical PCT already covers a population over 100,000, they would become responsible for commissioning and providing for a large population.

So we will return to a structure remarkably similar to the pre-reform districts, and will in the process have lost the reason why we started - the close connection with a relatively small local population.

Another circle I cannot square is the boundary between a PCT and what will be the new specialist service. A traditional faultline in mental health is the division between community services and hospital care. The most likely option is the transfer of community mental health services to PCTs, leaving trusts to cope with hospital care.

But it would set mental healthcare back by decades, recreating the old divisions and stigmatising hospital care yet further. Alternatively, we pass on the whole lot to primary care, and we will have recreated combined trusts with the addition of primary care. We could just have put a few GPs on their boards.

I have alluded several times to the importance of partnerships between the NHS and local authorities.

Coterminosity has been crucial for the local framework implementation teams to develop better integrated services.

Most PCTs will not be coterminous with local government boundaries, nor any other boundary. Several directors of social services have expressed their concern to me about the massive increase in bureaucracy if they have to create partnerships with PCTs and HAs for commissioning and PCTs and trusts for providing, multiplied by the number of new combinations created. A layer of management has been added, but noone is clearly in control.

All these scenarios will have to be acted out, and some effective structures may arise in a few years, able to serve mental health well . I am quite excited about the potential of bodies resembling health maintenance organisations, with powerful commissioning and local accountability at their core, offering vertical and horizontal integration.

My main concern is that in the meantime we are in a situation similar to what the late Sir Roy Griffiths observed at the end of the 1980s: mental health is everyone's distant relative, but no-one's child.

The structural changes affecting the NHS are primarily intended to address inefficiencies in acute and community services. Mental healthcare is squeezed into that framework, and in the case of PCTs, in hindsight.

It would be unrealistic to expect mental health needs to drive the full NHS reforms. The minimum expectation ought to be that a structure optimal for mental health is created, provided this does not damage other NHS and community interests.

We are at last close to that. The new freedoms allowing joint commissioning and integrated provision across health and local government, and the creation of mental health trusts, means that de facto we have specialist commissioning and providing.

If a mental health commissioning director with a ringfenced budget were added, the line of responsibility for the implementation of the framework would be entirely transparent. Florence Nightingale, carrying her lamp through the NHS, would know who was in charge of mental health, paraphrasing Sir Roy again.

I am not so sure about singleminded leadership radiating from PCTs. Why create new problems when we are approaching solutions?