The Effectiveness of Community Mental Health Nursing

Original article

The effectiveness of community mental health nursing: a review Charlie Brooker, Julie M. Repper, Andrew Booth Sheffield Centre for Health and Related Research (SCHARR), Sheffield, UK

In view of the importance currently attached to evidence-based health care, we present a systematic review of publications about the effectiveness of community mental health nursing interventions. Only 11 studies were identified which used an experimental design, focused solely on the nursing intervention, and were conducted in the UK since 1965. Not only is the evidence limited, but it does not examine those areas of work in which most community mental health nurses are involved, and the methodological rigour of the identified studies can be questioned. Other types of research with the potential to inform community mental health nursing practice are suggested. Research into the outcome of community mental health nursing interventions has made disappointing progress over the past decade. Journal of Clinical Effectiveness Vol. 1 No. 2, 1996: 44-50

Introduction In the present drive towards evidence-based health care, it is essential to define and measure the nursing contribution to health to ensure that professional and training resources are deployed most effectively. Thomas and Bond 1 have undertaken a review of the effectiveness of nursing in general that presents a disappointing picture of research in this area, with a lack of rigour in study design, small sample sizes, and no studies of cost-effectiveness. Despite a limited search strategy, nine of the 29 studies considered in their review were concerned with people experiencing mental health problems, which suggests a relatively rich vein of outcome research in this particular branch of nursing. In this paper we report a systematic review of rigorously selected evidence of the effectiveness of community mental health nursing. The problems found in conventional reviews of publications have been summarised by Cullum:2 reviewers fail to acknowledge the strengths and weaknesses in the primary research; they use only a subset of available material and fail to identify the criteria for including research in their review; discussion is limited to published research; assessment of the validity of cited research is difficult as methods are rarely described; and reviewers often draw inaccurate conclusions from study findings. Various guidelines have been compiled by which literature can be assessed systematically according to pre-agreed

Charlie Brooker PhD, MSc, BA, RAMN, Professor of Mental Health Nursing, Deputy Director, J. M. Repper MPhil, BA, RMN, RGN, Research Fellow, A. Booth BA, DipLib, ALA, Director of Information Section, Sheffield Centre for Health and Related Research, Sheffield University, 30 Regent Street, Sheffield S1 4DA, UK. Correspondence to: CB.

44

J. Clin. Effect. Volume 1 Number 2 1996

© Pearson Professional Ltd 1996

criteria to arrive at a judgement of a study's relative value and place it within the wider corpus of literature. 2-6 The key points to emerge from these can be summarised as a process of critical review. This is shown in the Figure.

Clinical effectiveness of mental health care The notion of clinical effectiveness has arisen, most notably, since the inception of the health service reforms in the early 1990s and the introduction of the purchaser/provider split. Once political and local imperatives have been met there are compelling reasons for purchasers with limited resources to contract for interventions that are known to be most clinically effective. That is, interventions at either the service or practitioner level that are based on research evidence of improved outcome for the client/patient. In view of the problems associated with the generalisability of single studies, clinical protocols are most usefully based on the aggregated results of carefully appraised studies in the area. A number of reviews and meta-analyses in the area of mental health have been published over the last 5 years and the work of the Cochrane Collaboration has prioritised schizophrenia as an area for critical review of existing research results. However, the findings of these reviews have been slow to influence purchasers or providers and there is a marked lack of reference within advisory documents and purchasing contracts to the activities of individual professions. Although nurses make up the largest professional group supporting people who have mental health problems,8-9 neither the focus of their work nor the skills which they require to be effective seem to be specified in service contracts. This is perhaps not surprising as the clinical effectiveness of community mental health nurses does not seem to have been subject to critical review.

The effectiveness of community mental health nursing

• Identify the specific questions to be answered (questions must specify the patient group or problem, the intervention being considered, the comparison intervention where appropriate, and the clinical outcomes of relevance) • Define the scope of the review and the criteria for inclusion (what years/type of study to be included, any aspects to be excluded) • Translate the concept of the review into search terms and keywords (use variants in terminology and spelling) • Define the search strategy (what databases/unpublished sources to be used) • Define the scope of the bibliography (is volume/quality as expected; does it need limiting by more stringent inclusion criteria, or expanding by looking at parallel areas of study?) • Filter the papers for studies relevant to the scope of the review (select studies that meet pre-determined criteria, may use hierarchy of evidence type to accept/reject studies, may follow-up other relevant studies from references of retrieved items) • Critical appraisal of the papers (describe methodological strengths and weaknesses of selected studies; may use a checklist to identify good quality studies and present hierarchy of selected studies; may present results in tabular form) • Systematic synthesis of results and contextual summary of review (identify areas with adequate supporting evidence, lack of evidence and conflicting evidence; identify areas for further research and review) Figure The critical review process (adapted from SCHARR7).

Their particular contribution to mental health care has not been carefully defined nor has its effect been systematically examined. As the recent National Review of mental health nursing suggests,9 this may be because of the difficulty in separating the work of nurses from that of other professional groups in terms of defining expertise and measuring the impact of individual interventions. In this regard nurses are no different from other professionals as measuring the outcome of mental health care interventions is a particularly complex process.

Assessing the effectiveness of mental health care As the treatment, care, and support of people with mental health problems targets physical, psychological, and social aspects of the person, outcome measurement must cover a range of dimensions. It must also take account of the fact that changes in a person's symptoms or functional abilities are points along a continuum rather than absolutes. Progress may take place over many years and is decided not only by treatment but by social, economic, and inter-related factors. Difficulties also exist in the selection or development of appropriate measures that can measure the impact of the intervention; and are reliable, valid, sensitive to change, and specific to the situation concerned. ° Even with the existence and assiduous use of such measures, attribution of improvement to specific interventions is problematic. The only conclusive way of proving association between treatment and outcome is by a controlled trial.11 In the present paper we review research on the effectiveness of mental health nursing. Our aim was to assess

Original article

the quality and quantity of experimental or quasi-experimental research and to synthesise the evidence that is currently available. The resulting conclusions may inform: purchasers in contracting for appropriate services and deploying resources more effectively; providers in setting up services and training which promote maximum benefit for targeted client groups; and the research community in assessing the focus and type of research which must be undertaken in mental health nursing.

Method The question addressed by this systematic review is 'How effective is community mental health nurse intervention?' Although it would have been preferable to specify the intervention, target group, and outcomes of relevance (Figure) the paucity of studies in the area required a broad and somewhat all-inclusive question. Nevertheless, the criteria for inclusion were carefully specified. A decision was made to include only those studies which fulfilled the following criteria: • Undertaken in the UK (as CMHNs may have a different training and role in other locations) • Undertaken in or after 1966 (the year that MEDLINE was established) • Studies of CMHN intervention rather than multidisciplinary teams in which CMHNs work • Used a randomised controlled trial or quasi-experimental design. The strategy used to search the publications databases is crucial to the success of any review and as Cullum2 has pointed out more than one indexing service should be used. This review aimed to be comprehensive both in the number of databases used and in the sensitivity of its search strategies. A broad range of databases from health and related disciplines was interrogated; MEDLINE, EMBASE, PsycLit, CINAHL, RCN Nurse-ROM, Science Citation Index, Social Science Citation Index, DHSSDATA and ASSIA. The Cochrane Database of Systematic Reviews and the embryonic databases of the NHS Centre for Reviews and Dissemination (DARE and NEED) were also searched. Hand searching of the International Nursing Index and a community mental health nursing bibliography 1 " were used to complement database coverage. Although cut-off dates vary according to database, the use of MEDLINE and PsycLit ensured the inclusion of materials from 1966 and 1967 onwards, respectively. A sensitive search strategy was used that searched for papers across four dimensions; the psychiatric or mental health discipline, the involvement or inclusion of nurses or nursing, the community setting and, a randomised controlled trial, meta-analysis or quasi-experimental study design. For each of these dimensions a mixture of index terms and free text were used. For example, in CINAHL, the subject heading community-mental-health-nursing was used as well as the free-text expression 'mental adjacent to health, or psychiatry or psychiatric' combined with

J. Clin. Effect. Volume 1 Number 2 1996 45

original article

either 'nurse or nursing'. These included non-specific terms such a random [truncated] and control [truncated]; the broadest level of searching advocated by the Cochrane Collaboration, 13 as well as specific keywords such as 'clinical-trial' or 'randomised-controlled-trial' in the publication type. As a further check the community setting concept was dropped from the search strategy and a review of resultant abstracts was undertaken to establish either their specific setting or to identify where a setting was not explicit. Similarly, institutional addresses were also reviewed to establish whether or not studies were likely to have been conducted in the UK. The methodological rigour of studies published in medical books and journals has been the subject of much attention over the past 25 years or so and errors are reported to exist in between 35-90% of trials published." More specifically in the field of mental health a recent survey of numeric methods used in all 12 issues of the 1993 British Journal of Psychiatry found that 40% (n = 65) of papers contained at least one error in design, analysis, or presentation of data.15 The quality of the methods of each of the studies selected for inclusion in this review was assessed using the following guidelines which are adapted from the Cochrane Collaboration: 4 • Was the process of randomisation explained? • Was a power calculation reported for estimation of the sample size? • What percentage of the sample was retained at followup? • What was the length of follow-up? • Was there blind assessment of outcome?

Results A total of 11 studies which met the preset criteria were identified (Table 1). They were concerned with five broad clinical topic areas: serious mental illness, 16 " 1 ' neurotic disorder, 18-21 behavioural nurse therapists, 22-23 liaison in Accident and Emergency departments, 24-25 and problem drinking. 26 The study by Brooker et al 16 was excluded from further consideration as clients had not been allocated at random to groups. The various aspects of methodological rigour examined in the remaining 10 studies is outlined in Table 2. In summary, the process by which clients were allocated at random to groups was explained in half the reports (5/10) and in two of these five papers the only description of allocation was the use of a minimisation procedure. None of the studies reported using a power calculation to assess the adequacy of sample size and associated error. However, blind assessment of outcome was reported to have occurred in eight of the ten studies. The length of follow-up varied greatly, with a range of 0-336 weeks. Three studies used a simple pre- and posttest design with no follow-up at all. A 7-year follow-up study was reported by Burns et al 18 and achieved an impressive 83% coverage of the original sample, however, variation in drop-out at follow-up varied from

46

J. Clin. Effect. V o l u m e 1 N u m b e r 2 1996


11-36%. The latter figure is reported by Gournay and Brooking 21 at 6-month follow-up. It is also worth noting that an economic analysis was also reported in four of the ten studies identified.

Discussion Ten RCTs of the outcome of CMHN intervention were identified using a comprehensive search strategy and subsequently the rigour of each trial was assessed. The number of studies was small and had a narrow range of coverage in comparison with the rich and diverse nature of CMHN clinical work (especially so-called 'generic roles'). The identified studies focused largely on CMHN outcome for clients with depression and anxiety-based disorders (n = 6) but also included CMHN liaison roles in Accident and Emergency departments (n = 2), with a single study each on outcome for problem drinkers and clients with serious mental health problems. White 29 identified CMHN specialisation (that is 'working intentionally and predominantly with one client group') as most common in work with older people, resettlement/rehabilitation, children/adolescents, and HIV/AIDS (these groups constituting the work of 42% of the total role for all CMHNs). Yet two of the selected studies concerned the outcome of trained nurse therapists rather than CMHNs. In addition, a further two examined liaison roles in Accident and Emergency departments and one concerned outcome for problem drinkers (both far from common roles). There is therefore little research to inform the work of generic CMHNs and a disproportionately small amount of research into the work of CMHNs who specialise. Not only is the coverage of CMHN work afforded by the 10 studies scant, but concern must be expressed about the methodological rigour of the research that has been undertaken in the area. For example, none of the studies reported the use of a power calculation to estimate sample size and the associated error, and only half of the identified studies explained the method used to allocate clients at random to groups (Table 2), so none of the studies would have been rated as Category A (adequate) if the Cochrane Collaboration Guidelines' had been applied, as the details of randomisation were not sufficient to allow assessment of the extent of bias. Notwithstanding these criticisms, the studies did show higher quality methods in terms of the length of followup, with some studies achieving high levels of sample retention over long periods of time. 18,26 In addition most of the studies made 'blind' assessments of outcome. In view of the limited evidence of clinically effective practice offered by RCTs which examine the outcome of CMHN intervention, it is important to consider other sources of evidence of effective practice. 30 Rigorously conducted RCTs that examine the clinical outcome of interventions undertaken by multidisciplinary community mental health teams certainly have implications for CMHNs. In two such studies, CMHNs constituted up to 70% of such teams. 31-32 Another potential source for evidence of effective practice are meta-analyses of RCTs that

The effectiveness of community mental health nursing Original article

Table 1 Summary of randomised controlled trials undertaken to assess the outcome of community mental health nursing in the UK (1966-1995) Study

Interventions

Comparison groups

Sample

Measures

Outcome

Catalan et al 24

Assessment of self-poisoners

CMHNs vs doctors

n = 120 (nurses = 75 doctors = 45)

1. 2. 3. 4.

1. 2. 3. 4.

Marks23

Behavioural psychotherapy in primary health care

Nurse therapists vs routine GP care

Clients with phobic and obsessional problems (n = 92)

1. Fear questionnaire 2. General Health Questionnaire 3. Service use (economic valuation)

Significant improvements and costs for nurse therapists

Gordon and Gordon' 9

CMHN-led cognitive group therapy

Cognitive therapy group vs controls

Depressed women (40-60 yrs) (n = 20)

1. 2. 3. 4.

Significant gains in depression, self-esteem and hopelessness for treatment group

Atha25

Problem-solving therapy

Intervention from CMHN vs standard care

Self-harmers presenting at A&E (n = 20)

1. Beck depression scale 2. Beck hopelessness scale

Significant gains for treatment group on both measures

Gournay and Brooking21

Counselling for minor psychiatric disorder

CMHNs vs routine GP care

Mostly women with an average age of 39 ( n = 104)

1. 2. 3. 4.

General Health Questionnaire Beck depression scale Spielberger trait anxiety Service use (economic evaluation)

No differences beetween groups in terms of clinical outcome or costs. Client satisfaction greater for CMHN group

Burns et al 18

'Supportive' intervention by CMHN

CMHNs vs appointments in psychiatric outpatients

Mostly women with chronic neurotic disorders (n = 71)

1. 2. 3. 4.

Major symptom score Social adjustment scale Service contact Economic evaluation*

No differences between groups for clinical outcome or service contacts BUT in 18-month follow-up study higher client satisfaction and lower costs for CMHN group

Gournay22

Behavioural psychotherapy

Home-based vs outpatient clinic-based

Mostly women with agoraphobia (n = 88)

1. Fear questionnaire 2. Phobic severity

Significant gains for both groups

Patterson et al 26

Supportive counselling for problem drinkers

CMHNs vs standard outpatient aftercare

n = 127

1. Abstinence 2. Blackouts 3. Marital discord

1. In CMHN group abstinence greater at 12 months 2. Fewer blackouts in CMHN group 3. Marital discord lower in CMHN group

Waterreus et al 20

Supportive interventions and behavioural psychotherapy from CMHNs

CMHNs vs routine GP care

Women with an average age of 76 years (n = 90)

1. 'Caseness' as defined by CARE 2. Consumer satisfaction

No change in 'caseness' but higher consumer satisfaction for CMHN group

Brooker et al 1 6 **

Family intervention in schizophrenia

'Trained' CMHNs vs controls

Mostly men with diagnosis of shizophrenia (n = 34)

1. Krawieka, Goldberg and Vaughan 2. Social adjustment schedule 3. General Health Questionnaire

Significant gains in clinical outcomes and consumer satisfaction for treatment families

Muijen et al 17

CMHNs using assertive outreach/case-management strategies

Assertive outreach CMHNs vs generic CMHNs

Men and women with a serious mental illness (n = 58)

1. 2. 3. 4. 5. 6.

No difference in clinical outcome between groups but control group costs much higher

Adequacy of assessment Consumer views of rapport Further episodes of self-poisoning Type of management following assessment

SCL-90-R inventory Beck depression inventory Beck hopelessness scale Social adjustment self-report questionnaire

Present State Examination Brief Psychiatric Rating Scale General Anxiety Scale Social Adjustment Scale Consumer views Economic evaluation

CMHNs and doctors assess equally well Management similar Consumers' views equivalent No difference in further episodes

* Economic evaluation not reported in 7-year follow-up study (Burns et al18) but reported in original 18-month follow-up (Mangen et al27); ** Brooker et al's study was considered and rejected by Cochrane Collaboration study of family intervention and schizophrenia (Mari and Steiner28) and so is included here but similarly rejected.

J. Clin. Effect. Volume 2 1996 47

Original article The effectiveness of community mental health nursing

Table 2 Methodological quality assessment of randomised controlled trials undertaken in community mental health nursing in the UK (1966-1995) Allocation

Follow-up

Study

Population

Catalogue*

Method

Report of power calculation

Catalan et al 24

Self-poisoners referred to a psychiatric unit

B

Random number table

No

Patient in general practice

B

No details

No

52

28

Yes

Listeners to national radio broadcast

B

No details

No

0

n/a

Not reported

Atha25

Attenders at an Accident/ Emergency department

B

No details

No

0

n/a

Not reported

Gournay and Brooking21

Patients in general practice

B

Minimisation procedure

No

24

36

Yes

Burns et al 18

Chronic neurotic patient referred to psychiatric services

B

Minimisation procedure

No

336

17

Yes

Gournay22

Agoraphobics referred to specialist services

C

Coins toss

No

0

n/a

Yes

Patterson et al26

Those discharged from an alcohol treatment unit

C

Availability of CMHN at discharge

No

52

11

Yes

Waterreus et al 20

All people aged over 65 in one electoral ward

B

No details

No

12

15

Yes

Muijen et al 17

People with a serious mental illness discharged from a psychiatric hospital

B

No details

No

72

30

Yes

Marks23 Gordon and Gordon

19

Length (wks)

% drop-out

Blind assessment

n/a

n/a

Yes

* Studies are allocated to either category A, B or C according to Cochrane Collaboration guidelines.4

48 J. Clin. Effect. Volume 1 Number 2 1996


have b e e n c o n d u c t e d in the field of mental health, for example the meta-analysis of family intervention u n d e r taken by t h e C o c h r a n e Collaboration which has clear implications for CMHN practice. 2 8 Although RCTs have b e e n h e r a l d e d as the sole conclusive m e a n s of proving that an intervention has the desired impact, 1 1 , 3 3 purchasing guidelines a n d clinical decision-making will never be totally reliant o n the findings of such studies. RCTs may be ethically i n a p p r o p r i a t e where interventions strongly suspected to be effective are withheld from the control g r o u p , a n d practically inappropriate in the study of rarely used interventions or infrequently occurring problems. Observational m e t h o d s have b e e n responsible for h u g e p a r a d i g m shifts in the history of mental health care (for example, Goffman 34 suggesting that over-dependence o n the results of r a n d o m i s e d controlled trials would limit innovation in practice. Indeed, in view of the lack of consensus over either the most desirable or the most appropriate areas of o u t c o m e to b e assessed, RCTs may n o t result in interventions that are most a p p r o p r i a t e or effective from the perspective of the individual patient or client. A n u m b e r of observational studies have b e e n published recently that d o take users' views into account; 3 5 - 3 6 or broadly review the existing publications in an area to make suggestions for practice a n d r e s e a r c h , 3 7 - 3 8 or e x a m i n e actual practice in the light of policy a n d recomm e n d e d g o o d practice. 3 9 - 1 0 This small selection of research shows the potential of observational m e t h o d s for informing the n a t u r e of c o m m u n i t y mental health nursing practice.

Conclusion Purchasers a n d service providers n e e d to base their activity o n key clinical questions: is an intervention safe, does it work, is it valued by the consumer, a n d is it worth it? However, as this p a p e r shows, little evidence of the effectiveness of C M H N interventions exists, a n d the quality of such evidence is questionable. This leads to a situation in which purchasers a n d providers must look to o t h e r sources of information for suggestions of effective interventions a n d must acknowledge the value of observational research. Although the practical limitations of the wholesale a d o p t i o n of the 'clinical effectiveness' a p p r o a c h have b e e n discussed, r a n d o m i s e d controlled trials remain the most conclusive evidence o n which to base clinical guidelines, a n d carefully c o n d u c t e d metaanalyses of such trials could inform the p u r c h a s i n g of effective mental health services. Ironically, the most salient finding of this review of the effectiveness of C M H N intervention is the paucity of high quality research into the effectiveness of CMIINs. In 1984 this was summarised as follows: In short, with a few notable exceptions, the literature does nothing to dispel the uneasy notion that the efficacy of CPN services in relation to patient outcome may well he a myth. Twelve years on, o n e is forced to ask w h e t h e r anything has c h a n g e d .

Original article

References 1. Thomas L, Bond S. The effectiveness of nursing: a review. Journal of Clinical Nursing 1995; 4: 143-152 2. Cullum N. Critical reviews of the literature. In: Hardey M, Mulhall A, eds. Nursing research: theory and practice. London: Chapman and Hall, 1994 3. Mulrow C D. The medical review article: state of the science. Annals of Internal Medicine 1987; 10G: 485-488 4. Cochrane Collaboration. In: Oxman A, ed. The Cochrane Collaboration handbook. ScctionVI. Oxford: The Cochrane Collaboration, 1994 5. Moher D. Assessing the quality of randomised clinical trials: an annotated bibliography of scales and checklists. Controlled Clinical Trials 1995; 16: 62-73 6. Richardson W S, Wilson M C, Nishikawa J, Hayward R S A. The well-built clinical question: a key to evidence-based decisions. ACP Journal Club 1995; Nov-Dec: A12-A13 7. Sheffield Centre for Health and Related Research, Critical Reviews Advisory Group. Critical reviews at SCHARR: a position paper. Sheffield: SCHARR Information Section, 1995 8. Onyett S, Pillinger T, Muijen M. Making community mental health teams work. London: Sainsbury Centre for Mental Health, 1995 9. Department of Health. Working in partnership: a collaborative approach to care. Report of the Mental Health Nursing Review Team. London: HMSO, 1994 10. Jenkins R. Towards a system of outcome indicators in mental health care. British Journal of Psychiatry 1990; 157: 500-514 11. Andrews G, Peters L, Teesson M. The measurement of consumer outcome in mental health: a report to the National Mental Health Information Strategy Committee. Sydney: Clinical Research Unit for Anxiety Disorders, 1994 12. White E. Community psychiatric nursing: a selected bibliography. In: Brooker C, White E. Community psychiatric nursing: a research perspective, Vol 3. London: Chapman and Hall, 1995 13. Dickersin K Scherer R, Lefebvre C. Identifying relevant studies for systematic reviews. British Medical Journal 1994; 309: 1286-1291 14. DerSimonian R, Charette J, McPeek B, Mosteller F. Reporting on methods in clinical trials. The New England Journal of Medicine 1982; 306: 1332-1337 15. McGuigan S M. The use of statistics in the British Journal of Psychiatry. British Journal of Psychiatry 1995; 167: 638-688 16. Brooker C, Falloon I, Butterworth A, Goldberg D, Graham-Hole V, Hillier V. The outcome of training community psychiatric nurses to deliver psychosocial intervention. British Journal of Psychiatry 1994; 165: 222-230 17. Muijen M, Cooney M, Strathdee G, Bell R, Hudson A. Community psychiatric nurse teams: intensive support versus generic care. British Journal of Psychiatry 1994; 165:211-217 18. Burns T, Paykel E S, Ezekiel A, Lemon S. Care of chronic neurotic outpatients by community psychiatric nurses: a long-term follow-up study. British Journal of Psychiatry 1991;158:685-690 19. Gordon C, Gordon E. Short-term group treatment of depressed women: a replication study in Britain. Archives of Psychiatric Nursing 1987; 1: 111-124 20. Wattereus A, Blanchard M, Mann A. Community psychiatric nurses for the elderly: well tolerated, few side effects, and effective in the treatment of depression. Journal of Clinical Nursing 1994; 3: 299-306 21. Gournay K, Brooking J. A prospective randomised controlled trial of the efficacy of CPNs and GPs in treating patients with minor psychiatric disorder in primary care. Unpublished Report to the Department of Health, 1991


49

Original article

22. Gournay K. The base for exposure treatment in agoraphobia: some indicators for nurse therapists and community psychiatric nurses. Journal of Advanced Nursing 1991; 16: 82-91 23. Marks I. Psychiatric nurse therapists in primary care: the expansion of advanced clinical roles in nursing. London: Royal College of Nursing, 1985 24. Catalan J, Marsack P, Hawton K E, Whitwell D, Fagg J, Bancroft H J. Comparison of nurses and doctors in the assessment of deliberate self-poisoning patients. Psychological Medicine 1980; 10: 483-491 25. Atha C. The role of the CPN with clients who deliberately harm themselves. In: Brooker C, ed. Community Psychiatric Nursing: A Research Perspective, Vol 1. London: Chapman and Hall, 1990 26. Patterson D G, McCourt M W, Shiels R A. Community psychiatric nurse aftercare for alcoholics. Irish Journal of Psychological Medicine 1991; 8: 4-14 27. Mangen S P, Paykel E S, Griffith J H, Burchell A, Mancini P. Cost-effectiveness of community psychiatric nurse or out-patient psychiatrist care of neurotic patients. Psychological Medicine 1983; 13: 407-416 28. Mari J de J, Steiner D. The effects of family intervention for those with schizophrenia. The Cochrane Database of Systematic Reviews (Issue 1), 1995 29. White E. Community psychiatric nursing: the 1990 national survey. Bradford: CPNA Publications, 1990 30. Slavin R E. Best evidence synthesis: an intelligent alternative to meta-analysis. Journal of Clinical Epidemiology 1995; 48: 9-18 31. Marks I, Connolly J, Muijen M. Home-based versus hospital-based care for people with serious mental illness. British Journal of Psychiatry 1994; 165: 179-194 32. Ford R, Beadsmoore A, Ryan P, Repper J, Craig T, Muijen M. Providing the safety net: case management for people

50



with a serious mental illness. Journal of Mental Health 1995; 1:91-99 33. Sheldon T A. Please bypass the PORT. British Medical Journal 1994; 309: 142 34. Goffman E. Asylums. Harmondsworth: Penguin, 1960 35. Day J C, Wood G, Dewey M, Bentall R. A self-rating scale for measuring neuroleptic side effects: validation in a group of schizophrenic patients. British Journal of Psychiatry, 1995 (in press) 36. Birchwood M, Smith J, Macmillan F et al. Predicting relapse in schizophrenia: the development of an early signs monitoring system using patients and families as observers: a preliminary investigation. Psychological Medicine 1989; 19: 649-656 37. MacLean L. Community psychiatric nurses in relation to diversion schemes. In: Brooker C, White E, eds. Community Psychiatric Nursing: A Research Perspective, Vol 3. London: Chapman and Hall, 1991 38. Repper J, Brooker C. Public attitudes towards community facilities for people with mental health problems. Journal of Health and Social Care in the Community, 1996 (in press) 39. Turner G. Client/CPN contact during the administration of depot medication: implications for practice. In: Brooker C, White E, eds. Community Psychiatric Nursing: A Research Perspective, Vol 2. London: Chapman and Hall, 1993 40. White E. Community psychiatric nursing 1980-1990: a review of organisation, education and practice. In: Brooker C, White E, eds. Community Psychiatric Nursing: A Research Perspective, Vol 2. London: Chapman and Hall, 1993 41. Brooker C. Some problems associated with the measurement of community psychiatric nurse intervention. Journal of Advanced Nursing 1984; 9: 165-174