In the mid-1990s the established pattern for adult psychiatric services in the Western Bay of Plenty, New Zealand was an in-patient/out-patient split. Psychiatrists either worked full-time on the in-patient unit or did only out-patient work.

In-patient work was not popular and it was hard to employ any psychiatrist for any length of time in this role. The ward itself was partly to blame, being small, cramped and dangerous. Although it had sufficed when it still had the support of a regional psychiatric hospital (Tokanui), it began to fail when the old hospital closed and the in-patient unit at the city general hospital was expected to accommodate seriously ill and committed (sectioned) patients.

The community teams had become ossified and territorial, spending time and energy in arguing over criteria for acceptance. Following discharge from hospital there was a dangerously long period of up to a month before the community mental health workers were able to see patients and provide support. This was despite the evidence that the first few days after discharge from hospital can be particularly risky (Reference OyebodeOyebode, 2005).

Part of the reluctance to do in-patient work was a result of the spirit of the times. The old psychiatric hospitals were closing. Forward-looking psychiatrists liked to be in the vanguard of community care. Psychiatrists whose outlook seemed to be stuck in the old era of institutional in-patient care risked being despised. Some psychiatrists were suggesting it was even possible to manage without any in-patient beds and they escaped ridicule.

For a while we solved the problem of consultant reluctance to do in-patient work by rotating our jobs, so that all of us took it in turns to spend a year on the in-patient unit. Continuity of care suffered. Patients and their families who had got used to one consultant, and were beginning to establish some trust, had to rapidly accommodate a very different individual. The same problem was true for nursing staff, social workers and occupational therapists.

When things went wrong, and even when things were going relatively smoothly, there was a tendency for the in-patient and community teams to blame each other, and express irritation at each others' unreasonable attitudes.

Introduction of crisis and home treatment teams

In the early 1990s a crisis team had been established which responded to every kind of psychiatric crisis in the community with no age restrictions. Sometimes the team would revisit the same individual on two or three occasions but it would not provide home treatment in any extended way.

In 1997 we created a home treatment team, partly to relieve dreadful pressure on the in-patient beds and partly because it seemed a better way to manage many people. Rather than combine crisis work and home treatment in one team, we found that home treatment was more successful when done separately, without unpredictable diversions. By this means we succeeded in reducing the number of people needing admission by about half. At the time we were driven entirely by the need to find practical solutions to pressing problems.

Although it is over a decade since Burns et al (Reference Burns, Raftery and Beadsmore1993) described home-based treatment in a positive way and the refreshing question was asked, ‘Is admission to a psychiatric hospital an ethical alternative to home-based treatment?’ (Reference FalloonFalloon, 1993), the supportive evidence for home treatment remains limited. Research in this area is bedevilled by the fact that crisis and home treatment teams vary greatly from place to place, both in the way they are constituted and in their aims and objectives. In theory, it is possible to create a home treatment team that adds to the expense of running a service without adding much to the quality of care. Crisis/home treatment teams are nevertheless prescribed by British policy implementation guides (Department of Health, 2002a , b , 2003). The most recent Cochrane review (Reference Joy, Adams and RiceJoy et al, 2005) remains positive but calls for more evaluative work, and a recent London study (Reference Johnson, Nolan and PillingJohnson et al, 2005) reports favourably.

Implementation of sector model

The in-patient/community split in psychiatric services in the Western Bay of Plenty changed in 1998 under the then clinical director who reorganised services along geographical lines into four sector teams. Patients were allocated to a team according to their home address. Allocating approximately equal numbers of patients to each team proved to be difficult because some lower socio-economic areas generated more patients and more difficulties. This left long-standing issues for teams who felt they were having to do more work without additional resources.

The British practice of allocating patients to teams according to their general practitioner (GP) was not directly applicable to New Zealand, where patients pay to see their GP. Many psychiatric patients cannot afford to consult a GP and therefore spend most of their lives without seeing one.

The transition to the sector model took about a year before disgruntled feelings settled, territorial walls were knocked down and people began to see that the new model really did work, and had advantages for continuity of care. There was, initially at least, a degree of healthy competition between the sector teams to have the fewest in-patients and to view in-patient admission as a failure of community care. To my surprise the psychologists happily embraced the sector model and enjoyed working in their new teams. The fact that they had different approaches, varying from analytic to behavioural, did not seem to matter and they all gave much appreciated input.

Although the quality of the service was greatly enhanced by the provision of a new, much larger, purpose-built ward, considerable strains continued, for example, those generated by the failure in much of New Zealand to provide any secure in-patient rehabilitation facilities for people with severe long-term illness who had been assessed as too dangerous for management in the community. These people were not adequately catered for on the admission ward and were a disturbing influence on people needing brief acute admissions. The British, in contrast, not only retain a few rehabilitation beds, and high, medium and low secure regional National Health Service (NHS) units, but have recourse to a number of (expensive) private secure hospitals for people who cannot be adequately managed within the NHS.

Experience of UK system

In September 2003, after 9 years as a consultant psychiatrist in the Western Bay of Plenty, I had the opportunity to compare different services in England where I worked as a locum consultant. It was with a keen sense of irony, therefore, that I found that in two of the areas I worked, Exeter and Ipswich, energy was going into reform in the opposite direction, that is, from a sector model to an in-patient/community split.

There now seem to be a number of psychiatrists who are willing to pursue a career in acute in-patient work. They appear to find the care of severely disturbed individuals genuinely rewarding and perhaps find ward work simpler than dealing with large numbers of people in the community. As consultants move towards a more consultative role, the out-patients they see in person are increasingly complex, difficult cases. A half-day clinic can therefore be strenuous and draining.

Division of labour into purely in-patient work, rehabilitation or acute community psychiatry can have the same effect as division of labour in industry. Although specialised skills are gained, jobs can become repetitive and boring. Many consultants relish working in a variety of settings during the course of a week and seeing a wide variety of clinical situations.

For nurses on in-patient wards, dealing with just one consultant can be far easier than having to accommodate several different teams, all needing different times for ward rounds. On the other hand, nursing staff can relish the different approaches offered by more than one consultant and can sometimes find dealing with only one consultant oppressive, especially when that person is rigid in outlook. The same point of course applies for the staff of sector teams.

Conclusions

My experiences have led me to the following conclusions.

1. • Changing a system takes energy. Even simple changes in people's work practice may take longer than initially envisaged before all the knock-on effects are dealt with.

2. • Before putting a service through the stress of change, it is important to be very clear about what one is trying to accomplish. Change for essentially fashionable reasons cannot be justified.

3. • The best development of the last decade is the provision of home treatment as an alternative to hospital admission; after all, this is usually what you would want yourself.

4. • In-patient treatment has an important place in the range of services provided. In-patient units should be comfortable and well-designed, allowing everyone to be cared for with dignity, even if they require seclusion. Single rooms with en suite facilities are now the expected standard. Once again ‘ What would I want?’ is the golden rule to apply.

5. • Both sector model and in-patient/community split have their strengths and weaknesses. Both depend on good communication and trusting relationships between colleagues if they are to work well. Separation into specialised teams creates boundaries which can deepen into dangerous chasms unless energy is specifically directed into border diplomacy and efficiency.

Declaration of interest

None.