Towards managing the whole system of care and improving continuity of care

As mental health services become more complex with additional teams, the risk of discontinuity of care for patients with complex needs may rise. Clinicians may have more difficulty in analysing the whole service picture. Hence they will be less able to contribute to the smooth running of the whole

As mental health services become more complex with additional teams, the risk of discontinuity of care for patients with complex needs may rise. Clinicians may have more difficulty in analysing the whole service picture. Hence they will be less able to contribute to the smooth running of the whole system. This paper describes how a north London trust is addressing this important challenge in straightforward and practical ways.

The problem
It has been recognised in recent years that failures in care are more often due to system problems than acts or omissions of individual professionals. Such failures could become more common as the system of mental healthcare becomes more complicated. The number of teams and interfaces between teams is increasing. There is more autonomy of professions and more specialisation. All this means that patients, and especially those with complex needs, have to journey across more boundaries. Kennedy & Griffiths (2001) found that many psychiatrists, nurses and other professionals are involved in frustrating and festering disputes about who should do what. This is not only detrimental to interprofessional relationships, but also makes patients' journeys through the system more perilous. Such concerns about continuity of care were highlighted at two national conferences in the spring of 2003 (British Medical Association, 2004).
The introduction of care coordinators and the 'care programme approach' during the 1990s was a step in the right direction. However, care coordinators are dependent on a system that works and can provide each component of care at the time it is required. No care coordinator would claim to have the powers to overcome delayed admissions, out-of-area treatments, delayed discharges, and lack of supported housing. When people in specialist crisis or assertive outreach teams, rehabilitation psychotherapy or forensic services consciously or unwittingly alter their boundaries, excluding patients for whom there are no alternatives, no care coordinator, consultant or senior service manager can fix the system on their own.

Looking for good practice
There has been plenty of comment on these problems, especially in reports of serious incident inquiries, but little on the solutions. Hence, a visit to a north London mental health trust in search of good practice was so enlightening. What was modestly called their 'bed-management committee' demonstrated rapid communication followed by decisive action from frontline staff to the top of the organisation on immediate system problems, bridging all the teams and components of the mental health service -and beyond into social care and housing.
This trust covers inner-city populations with high levels of deprivation, drug and alcohol misuse, and many refugees. It is one of the more developed services in the country with well-established crisis home treatment and assertive outreach teams supporting the sector community teams. It used to spend millions on out-of-area admission, and preside over 200 acute beds that were permanently overcrowded and in crisis. The bed management committee is now attempting to manage the whole system.
It meets every Wednesday morning and is chaired by the medical director. Core membership is small, with the local authority director of one borough and the lead managers of each of the in-patient sites managed by the trust. Consultants and other senior personnel attend as they wish, or when invited to attend ad hoc depending on the particular system or service problems requiring scrutiny at the time. The meeting lasts 1 h. The first half of the meeting monitors standard performance items such as bed occupancy, leave and absent-without-leave beds. Problems that have arisen for the site managers during the week are discussed, and any out-of-area treatments. There is monitoring of patients that have been in hospital for more than 60 days, and of the effectiveness of the crisis teams in working with the community mental health teams (CMHTs) to prevent admissions and facilitate early discharges. The second half of the meeting is devoted to a chosen topic: an emerging problem or new idea to improve the system.
The trust's executive committee meeting, chaired by the chief executive, runs back-to-back with this meeting. The first item on that agenda raises key issues from the bed management committee that require chief executive decision or even trust board resolution. In this whole process there is surprisingly little paper and few data. Only simple bed and head counts carried out within the past 24-48 h are needed.

An exemplary approach
The following is an illustration from one meeting of the kind of problem-solving achieved: To put it in context, the previous 2 years in the trust had seen the successful introduction of crisis resolution home treatment teams that had virtually eliminated out-of-area treatments and reduced bed occupancies so that there was usually a vacant bed when required. However, the meeting opened with recognition of an unusually large Kennedy et al Towards managing the whole system of care opinion & debate seasonal rise in admissions that needed action to avert a crisis. Clinically unsatisfactory and very expensive out-ofarea admissions were beginning to recur. Extra 'fold-up beds' were starting to be used in office space and other unsuitable locations in wards. Discussion ranged widely across the whole system of care to identify all contributory problems and solutions.
Starting at one end of the system . . . For example, it appeared that some patients had been admitted by inexperienced senior house officers from accident and emergency departments during the night. Why had experienced personnel in crisis home treatment teams not assessed these patients whose admissions might have been avoided? It was agreed that admissions from the accident and emergency departments would be tracked to ascertain the size of the problem and the detailed reasons. There were indications that night-time on-call responses from crisis home treatment teams were sometimes not swift enough. It was suggested that the relatively small amount of funding to allow night-time cover by crisis teams with staff who are awake and on duty might be more than repaid in savings from the high cost of out-of-area treatments. It was decided to research and cost that option. Moreover, a check would be made on the extent to which the crisis home treatment teams were achieving a target agreed in previous weeks, i.e. that a quarter of their case-load would involve providing home treatment allowing accelerated discharge from hospital.
. . . pursuing problems to the other end of the system Forty per cent of the patients on the acute wards had been in hospital for more than 60 days or 100 days. This meant that only 60% of the beds were available for more acute admissions. A better system was needed to speed up moves to more independent accommodation of people in the 400 sheltered housing places provided by the local authorities. A pilot study had recently been carried out showing that this was possible. The group discussed setting a short-term target to move 20 people. This in turn would enable people to move from the 140 residential care and nursing home places, and subsequently would enable a number of people with delayed discharges to be moved from the wards. Thus a small change in the proportion of available places in sheltered housing could make a very big difference to the availability of short-stay acute beds on the wards. It was agreed that a wider understanding of this issue by the CMHTs and assertive outreach teams was required, so that more energetic collaboration with the trust's accommodation team would allow suitable patients to be moved more quickly.
. . . optimising bed use Information sought by the bed management committee demonstrated that the recent high bed occupanices were associated with a breakdown of sectorisation. In some wards only half the beds were occupied by patients from the sector that the patient came from. Patients admitted to beds in other sectors tended to stay longer because their community teams had a tendency to 'forget them'. CMHTs would be asked to review more quickly any patient placed outside their sector wards and accelerate discharge or transfer to their sector ward. An up-to-date list of out-of-sector patients would be maintained and circulated to encourage more rapid transfers to appropriate wards.
At the end of the meeting a range of specific action points were agreed. Of particular importance was the briefing of staff on an accurate up-to-date analysis of how each part of the system could contribute to averting the developing crisis. Experience had shown that most frontline staff were keen to cooperate and help solve service-wide problems, but they were limited in their ability to do so by a lack of understanding of the bigger picture. Without that knowledge, pressures from other parts of the service are perceived as unfair 'dumping of work' -engendering a lack of cooperation.
The meeting then deliberated on the single topic of whether 'fold-up beds' in in-patient wards could be an acceptable response to pressure for beds. Experience of the past 2 weeks gave rise to the fear that such arrangements could become 'normal'. A recommendation would be taken to the chief executive that use of 'fold-up beds' was high risk and fell below minimum standards.
. . . then ensuring action from top to bottom of the organisation The trust 'executive committee', chaired by the chief executive, endorsed this recommendation. It was agreed that the service managers would make the position clear to all staff concerned and personally ensure that patients in fold-up beds were better placed and these beds removed before the day was out, or as soon as possible thereafter. The chief executive would alert the trust board and the primary care trusts to this minimum standard issue and seek agreement that previous plans to make bed reductions to fund other services would be deferred.
Consultants' reactions to whole system management Consultants in the trust have been supportive of this process. The main reason for this has been success in reducing the pressure on beds and increasing usable beds by reducing the number of long-stay patients. The committee's work on facilitating closer collaboration between the crisis teams and the CMHTs, and improving sectorisation has also been welcomed. An important reason for the success of this approach is the fact that the committee has never challenged the admission of patients nor any other clinical decisions, but rather has facilitated the better use of resources by clinicians.

Conclusion
It seems self-evident that structure, process and leadership of this kind is essential for the sound operation of a complex mental healthcare system. Resources will always be tight. Inevitably there will be fluctuations in demand affecting different parts of the system, creating bottlenecks and areas of slack that, considered together, can lead to smoother running of the service.
What impressed the external observer (P.K.) was: . how ordinary and easily repeatable this model seems; . how limited are the data, paperwork and executive time required to make it work; . how odd it is that so few trusts have yet discovered the need for such an arrangement -including the trust I managed myself for 10 years.
This north London trust will continue to refine its approach from which others can learn. It seems hardly credible that any modern mental health service can function well without such real-time operational management that addresses system problems across and up and down the organisation.

Declaration of interest
None.