Survey results

Implementation of the CPA

1. By 1 July 1997, twenty-one areas had implemented the CPA (see annex, column 1). Table 1 shows the year of implementation, as follows:

2. The remaining 4 areas intended to do so by July 1998. Only Shetland gave no commitment to implement the CPA by a specific date.

Number of people receiving the CPA

3. Two of the 21 areas which had implemented the CPA did not have anyone on it at 1 July 1997. The other 19 areas reported a total of 1,096 people on the CPA (column 2 in the annex gives a detailed breakdown of the figures) which included 228 people with dementia. More than half (126) of the people with dementia were in Edinburgh with 70 others on the CPA in Dumfries and Galloway. The numbers of people with dementia on the CPA elsewhere were therefore very small.

4. The number of people per 100,000 catchment population receiving the CPA at 1 July 1997 ranged from 0.8 to 96. Table 2 shows the wide range of CPA use across the 26 areas.

5. We also asked how many people (per 100,000 population) authorities would expect to receive the CPA, given local criteria and what is known about the characteristics of the catchment population. Five replies suggested between 0-49 people per 100,000 but 10 areas estimated twice this figure, between 50 and 100 people (see annex, column 3). We asked respondents to explain the basis for their estimates. These varied from a national estimate to objective local estimates based on the current number of people on the CPA, prevalence figures and adjustment for rural areas. As can be seen in Table 2, only 3 areas have achieved these estimated figures. This would suggest either that the bases used were unreliable or that the CPA is not as fully operational as it should be. Covering letters from some respondents suggested that the CPA was being introduced incrementally (eg piloted in one geographical area then implemented throughout the area).

6. Respondents reported 1,344 people having received the CPA since its introduction (annex, column 4).

Agreements and policies

7. Circular SWSG16/96 suggested a joint management group should take responsibility for planning, coordinating and implementing the Approach. All 26 areas reported that they had a multi-agency group responsible for planning and co-ordinating implementation of the CPA (see annex, column 5).

8. Membership of these groups was as follows:

9. Circular SWSG16/96 emphasised the need for a joint approach in implementing (as distinct from planning and coordinating) the CPA and that the roles and responsibilities of each agency needed to be formally set out. Seventeen responses (see annex, column 6) indicated that a joint approach to implementation had been adopted. Nine areas did not have a formal agreement. Twenty-two areas had a clearly defined policy for the CPA jointly agreed between health, social work and housing and all other relevant groups. Two areas did not have such a policy, one did not reply (Inverclyde) and one area (Shetland) had a policy in draft (see annex, column 7). In twenty areas the policy included people with dementia. This was not the case in Lanarkshire or Orkney. Four areas felt it was not applicable (see annex, column 8).

10. The 1996 circular stated that "the joint management group should ensure that arrangements for the CPA are fully compatible with those for assessment and care management and locally agreed protocols for hospital admission and discharge". As identified in column 9 in the annex, 17 of the 26 areas had compatible arrangements. Eight areas did not. (Shetland did not reply.) The compatible arrangements in the 17 areas varied. In some the CPA was designed to fit with existing care management arrangements and protocols for hospital admission and discharge. Some areas had ensured that documentation was compatible, while others relied on procedures such as attendance at meetings of care managers and CPA key workers to ensure compatibility. Two areas noted that procedures had been developed through multi-agency training initiatives. One area said "CPA guidelines and protocols have been drawn up to mirror those of care management" (Angus).

11. The 1996 circular also suggested that joint multi-agency assessment tools should be developed where they were not already in use. Thirteen areas said they had common formats between health and social work for assessment, and 15  said they had common formats for individual care plans.

12. Eleven areas provided information about the percentage of people on the CPA who had separate key workers and care managers. These ranged from 0-50% in 6 areas and from 50-100% in 5 others. This raises the question of whether there is duplication or overlap in the professional help being offered.

Eligibility criteria

13. For the Approach to be effective, clear eligibility criteria need to be developed and agreed between partner agencies. All 26 areas said they had such criteria (see annex, column 10). In twenty-two areas, the criteria included people with dementia.

14. Social work departments and health boards had agreed the criteria in all the responding areas, and Trusts had also agreed in all but 2 of them. In 14 areas GPs had agreed the criteria. Other stakeholders who had been consulted and agreed the criteria included voluntary organisations (13 areas), people using the Approach (8 areas), housing (8 areas), carers (4 areas) and police (in 3 areas). The criteria were available to all those who carried out assessments in 25 areas, to CPA key workers in 25 areas and to people who used the services and/or their carers in 22 out of 25 areas (Highland did not reply to any of these questions).

Operational guidance

15. All 26 areas had devised operational guidance for the CPA (see annex, column 11). However, only 8 included standards of care expected in this guidance (for details see annex, column 12). Eighteen respondents had a procedure for resolving disputes between professionals about placing someone on the CPA (annex, column 13). The remaining eight areas did not have a procedure. The most common way of resolving disputes was via the CPA steering/management implementation group (12 areas).

16. Ten areas offered or delivered care management if anyone refused the CPA. Four areas used risk assessment to help them to decide what approach to take. Three areas had a multi-agency group which would review each refusal, and 3 others would offer the person concerned what they described as "assertive outreach". Five areas indicated that people would be placed on the CPA without consent if the risk they posed to themselves or others was thought to warrant it. Some areas used a number of strategies.

17. It is important that people who fail to attend appointments or who do not respond to home visits, are followed up. We asked what policies were in place to ensure this happened. In nine areas a review group would meet to consider the course of action and 7 areas were clear that key workers had the responsibility for initiating a response. In some cases key workers were responsible but they were expected to convene review groups or report to a consultant psychiatrist first. Two areas had no policy (Lanarkshire and Orkney), but were developing one. The criteria for considering someone "lost to follow-up" also varied. For example, in some areas one non-attendance at an appointment might be enough to trigger "lost to follow-up" action, while in another area 3 non-attendances might be needed. However, several areas stressed that these decisions would be made individually, taking account of the assessed risk in each case.

18. We asked what the policy was for people who were planning to move from the area where they were on the CPA. Two areas (Ayrshire and Arran and Orkney) had not yet developed a formal policy. In 8 areas the policy was to have meetings with key personnel from the new area. Eight areas said they sent written information. Four areas left these arrangements to the key worker who would refer to the new agency. Three areas discussed it with the person concerned and/or their carer. Twelve areas informed all "relevant agencies" of the move. Some areas used a number of strategies. The variability of these arrangements is a cause for concern. People who are on the CPA are by definition vulnerable. It is essential that any handover between services is well managed to ensure that care is provided continuously.

Key workers

19. One of the central tenets of the CPA is the role of the key worker in organising and monitoring the care and supervision of the recipient. Almost all (24) areas had defined the roles and responsibilities of key workers in writing. Only 2 areas had not done so (see annex, column 14). Social workers and community psychiatric nurses were eligible to be keyworkers in all 26 areas. Mental health nurses and occupational therapists were also eligible to fulfil the role in all but one area, as were psychiatrists in all but two areas. Table 4 shows the individuals able to act as key workers:

People on the CPA and their carers

20. The 1996 circular stated that "users should be given a copy of their care plan". All but one of the respondents (Stirling) stated that it was their policy for people on the CPA to have a copy of their care plan. Stirling stated that people could have a copy on request.

21. Thirteen out of 21 operational CPA areas collected information on the number of people who received a copy of their care plan. In ten areas everyone on the CPA received a copy of their care plan (column 15, annex). Twenty-one areas stated that people on the CPA and carers were given information about the CPA. The remaining 5 areas did not provide this information.

22. One of the objectives of the CPA is that people using the service should be involved in all aspects of the CPA. Fifteen areas collected information on the number of people who attended their own CPA meetings. Twelve areas provided figures as follows:

These findings suggest that people on the CPA are not yet sufficiently involved in their own care planning.

Funding

23. In seven areas the CPA manager posts were funded by health; 6 by social work; 4 jointly by social work and health and one through revenue support grant. Two of the social work funded posts used the Mental Illness Specific Grant.

Administration

24. Twenty of the 26 areas provided administration time to support the CPA. Of the 17 areas who gave details, the hours attached to this task were not specific or were negotiable in two. Other areas varied in the amount of administrative support provided.

25. We asked respondents to calculate the annual costs of management and administrative support. Half the areas had either not yet determined or estimated these costs or did not reply. The other 13 areas replied as follows:

Practice issues

26. We asked areas if they recorded review dates on CPA forms. Twenty-two areas did so. It proved more difficult to get information on the percentage of CPA reviews which were held within 7 days of the dates set at the previous review, as only half the areas responded (itemised in the annex, column 17).

27. One indicator of the effectiveness of the CPA is the appropriateness of hospital admissions during the year. In sixteen areas a total of approximately 304 people receiving the CPA were admitted to hospital during the previous year. Responses from twelve areas indicated that 80 of these were unplanned or crisis admissions. There were no unplanned or crisis admissions in the other four areas (see annex, column 19).

28. The survey sought information about particularly vulnerable groups of people who may have been on the CPA. The figures in Table 9 are based on responses from sixteen areas only.

29. We were interested to know who managed the CPA on a day-to-day basis. Responsibility rested with the CPA coordinator/manager in 8 areas, with service managers in a further eight, with community mental health team in six areas and various other members of staff (including a divisional mental health officer and a senior care manager) in the 3 remaining areas. In one area the steering and implementation group had this responsibility. Seven areas had full-time managers and 11 had a part-time post covering the task. Eight areas either thought the question did not apply to them or gave no response. The amount of hours dedicated to managing the CPA in areas where the post of manager was part-time, varied considerably from 3-6 hours per week in six areas, between 7 and 10 hours per week in four, and between 10 and 15 hours per week in one area. In only 10 areas did the manager have a clear job description with lines of accountability. Five areas did not provide this. Nine areas thought the question did not apply to them. The City of Edinburgh and Midlothian gave no response.

Declined/discharged/lost to follow up

30. As column 20 in the annex identifies, a total of approximately 61 people across 17 areas were reported as having declined the offer of the CPA. Eighteen areas reported that a total of 225 people had been discharged from the CPA in the past year (for details see annex, column 21). Fourteen areas reported that they had not lost contact with anyone on the CPA in the last year. Four areas (Dumfries and Galloway, Greater Glasgow, Angus and Perth and Kinross) reported they had lost contact with between 1-4% of people in the last year. This would suggest that key workers have been largely successful in maintaining contact with people on the CPA.

31. We were interested to see if systematic attempts were made to identify people who were not in touch with services who met the criteria for the CPA. Fifteen of the 21 responding areas did not attempt this. Six areas did do so, half of them by reviewing caseloads.

Monitoring the progress of people on the CPA and data security

Monitoring

32. Nineteen areas reported that CPA data could be used to monitor the progress of people on the CPA. Eleven areas said that CPA data could be used in clinical service management, e.g., in balancing and limiting CPA key worker case loads. Nineteen areas said that CPA data could be used for clinical audit and monitoring of the CPA process and 20 areas said that it could be aggregated for planning purposes.

33. It is important that this CPA information is distributed to key people. Nineteen areas provided CPA information to the purchasers of services. Seventeen areas reported to Trust boards, 18 to service managers, including housing. Nine areas did use the information for clinical audit. Nineteen areas said they could do so. Eighteen areas reported the results to clinicians and clinical teams and 16 of them reported to users’ and carers’ representatives also. Sixteen areas said they had developed or were developing local outcome measures. The most commonly used measure (in 5 cases) was the Health of the Nation Outcome Scale.

Data security

34. Because of the sensitive nature of the information held on people on the CPA it is important that data is held securely. Fifteen respondents had clear written policies on disclosure of CPA information to all health and social care workers involved in individual care. Thirteen areas had a written policy on disclosure of information to housing workers; 12 for the police and probation services; 14 for relatives and carers and 14 to people using the service about their own care (complying with the Access to Health Records Act 1990). Only 4 areas held CPA data on a joint health/social work information system. Thirteen respondents reported CPA data were held on an electronic database and 18 on a manual system (some used both). All of the areas which stored data on an electronic database controlled access by means of a password or encryption.

35. The collection and availability of data is an important element in implementing the CPA effectively. People on the CPA may need help at any time of the day, not just during office hours. In 21 areas CPA data was readily available to those who needed to know during office hours but only 12 areas could provide it outside office hours. All 21 areas felt their CPA data were kept in a secure environment. They all confined access to CPA data to groups of workers on a need to know basis.

Training

36. Our survey sought information about staff for whom CPA training was provided.

The responses were as follows:

Nineteen areas stated that professionals from different agencies attended the same CPA training courses. In nine areas all key workers were trained for the CPA. One area said 80% of key workers were trained. In 2 areas 60% received training and one area trained half its key workers. Two areas could not answer this question; 2 did not respond and in 4 areas, key workers were not receiving any CPA training (see annex, column 22).

37. It is important that staff working with people on the CPA should know what criteria to use in defining severe and enduring mental illness. The survey asked whether criteria were included in training packages and for which groups. In 18 areas criteria were included in training for people carrying out needs assessments. Seventeen areas included them in CPA key worker training and 14 areas included them with users and carers involved in the CPA. Thirteen areas included criteria in training for local user and carer organisations. Only 4 areas incorporated information about the Data Protection Act into their training.

Developing the CPA

38. We were interested in the positive aspects of a local approach to the CPA and asked for comments. Table 11 shows the most common responses:

39. We asked what would help to improve the implementation of the CPA locally. The most common responses were resources (12) and training/joint training (7). Other areas identified administration/support (4), more involvement from GPs (3), and a common database (3).

40. Nineteen respondents thought the administration process could be improved. Ten areas said that more dedicated administration time would help the process. Others identified improved IT support/database/computerised systems (6) and a national network for tracking those people lost to follow-up (2).