Developing needs led child and adolescent mental health services: issues and prospects

European Child & Adolescent Psychiatry 8:1±10 (1999) Ó Steinkop Verlag 1999 REVIEW ARTICLE FOR DEBATE R.C. Harrington M. Kerfoot C. Verduyn Developing needs led child and adolescent mental health services: issues and prospects Accepted: 16 September 1998 R.C. Harrington (&) Department of Child and Adolescent Psychiatry Royal Manchester Children's Hospital Pendlebury Manchester M27 1HA, United Kingdom e-mail: r.c.harrington@man.ac.uk M. Kerfoot Department of Psychiatric Social Work Maths Tower University of Manchester Oxford Road, Manchester C. Verduyn Department of Psychology Royal Manchester Children's Hospital Pendlebury Manchester M27 1HA, United Kingdom Abstract For many years mental health services for children have been developed incrementally with little attention to the needs of the local population. However, over the past decade there have been attempts to develop more rational ways of planning child mental health services. This paper describes the information required to develop a needs-led child mental health service and, within that context, discusses how priorities should be set. It will be suggested that although the assessment of needs for child and adolescent mental health services is still very haphazard, there is now a clear trend for the evaluation of clinical practice to become more systematic. At an individual level we know quite a lot about the e cacy of treatment and the measurement of outcomes. At the service level, several models of good practice are being speci ed and evaluated. Key words Needs assessment ± child mental health ± service planning Introduction Child and adolescent mental health services (CAMHS) have historically been neglected as a priority area within many health services. In the UK, for example, even the government admits that service development is `patchy' (33). Services for children with emotional or behavioural problems have been especially vulnerable to planning `blight' as a result of failure to reconcile di erences between professions and agencies who provide services. The planning process has been further undermined by the di culty in showing a clear relationship between level of expenditure on child mental health services and outcomes (7). Health professionals in many countries have therefore adopted an incremental approach to the development of services. As extra money becomes available, it has been spent bit by bit on an extra therapist here, a psychiatrist there, and so on. When funding has been reduced, or resources have been withdrawn by other agencies, cuts have been made on an ad hoc basis. The process of expansion and contraction of services has, by and large, been unplanned. The result is often great variation between districts in the nature and extent of child mental health services. Over the past decade, however, there have been attempts to develop more rational ways of planning child mental health services. For instance, the Department of Health in the UK has urged health authorities to develop partnerships with local authorities in order to de ne what they mean by `needs' and to develop a joint children's strategy to meet them (37). Similar kinds of

2 European Child & Adolescent Psychiatry, Vol. 8, No. 1 (1999) Ó Steinkop Verlag 1999 processes have occurred in other countries, such as the United States (46). In many areas, then, e orts are being made to assess the needs for mental health services of children and adolescents. In the UK several reports have recently been published to help in this process (12, 19, 28, 43). Since children with mental health problems commonly require input from several agencies (e.g., health and education) most of these reports propose a system for a comprehensive multi-agency assessment of need, followed by a strategic approach to the commissioning of services. Setting priorities on the basis of multi-agency needs assessments is great in theory and it is easy to see why both health providers and health purchasers are attracted to the idea. But will it work in practice? In this paper we shall focus on the information required to develop a needs-led child mental health service and, within that context, discuss how priorities should be set. The general perspective that will be presented is one of cautious optimism. It will be suggested that although the assessment of needs for CAMHS is still very haphazard, there is now a clear trend for the evaluation of clinical practice to become more systematic. At an individual level we have an increasing knowledge base about the e cacy of treatment and about the measurement of outcomes. At the service level, several models of good practice are being speci ed and evaluated. The paper is organized into three parts. The rst part deals with the needs of the individual and the e ectiveness of treatments. In the second part the focus is on the needs of populations. The third part deals brie y with the process of choosing between needs. Measuring the needs of individuals Need as emotional or behavioural symptoms Symptoms and signs of mental disorder in children can now be recognized with reasonable reliability by trained practitioners (4). The application of standardized rules (diagnostic criteria such as DSM-IV ± (3) or the ICD-10 (45)) then results in fairly reliable diagnoses. These rules are intended to sharpen the di erentiation between isolated emotional or behavioural symptoms, which are very common in children and often of no psychopathological signi cance (36), and mental or behavioural disorders. Criteria for assessing whether or not a disorder is present di er according to the nature of the presenting problems. They include the frequency of symptom/s, number of symptoms, duration, persistence, age appropriateness (departure from expected developmental course), intensity and controllability. Need as severity of impairment It will be appreciated however that the presence of symptoms does not, in itself, indicate a need for intervention. Indeed, recent epidemiological studies suggest that when disorder is de ned by symptomatic criteria alone, nearly 50% of children have a diagnosis (39)! Accordingly, in many diagnostic systems disorder is diagnosed only when symptoms lead to impairment in one or more areas of everyday functioning, such as school work or peer relationships. This usually leads to an overall prevalence of mental and behavioural disorder of between 5 and 10% (36, 39). Need can then be de ned by categorizing individuals according to severity of impairment. For instance, Wallace and colleagues (43) proposed four categories of need for CAMHS, based on the frequency and severity of impairment. These were a) common problems with a low risk of severe disorder, b) common disorders that are not necessarily severe, c) less common problems which indicate severe disorder, and d) potentially severe disorders. Using the concept of impairment in the planning of child mental health services has some problems. For example, it can be very di cult to determine which diagnoses or problems are causing the impairment. Moreover, in many instances the so-called impairment (e.g., problems with school work, poor intrafamilial relationships) may be a cause as well as a consequence of emotional or behavioural symptoms. Nevertheless, the severity and duration of social disablement are important determinants both of referral to child mental health services (36) and of response to treatment (23, 24). Need as the rate of contact with CAMHS `Need' can also be estimated from the combination of morbidity arising from disorder and rate of service utilization. These methods usually indicate substantial unmet needs. For example, in the Isle of Wight study (36) only 10% of children with a diagnosis that caused impairment were in contact with mental health services. Such ndings are often used to argue for increased spending on CAMHS, particularly community-based services. However, data based on service utilization are of only limited value in planning CAMHS for two reasons. First, they say little about the reasons for poor uptake of services or the factors that might improve take-up. Indeed, one of the ndings from the Isle of Wight study that is often forgotten is that the majority of parents of children diagnosed with a disorder did not see their child as in need of treatment! Second, such approaches do not deal with the situation in which children have symptoms for which no e ective intervention is available.

R.C. Harrington et al. 3 Developing needs led child and adolescent mental health services Needs as lack of e ective care or services Another way of de ning need is as the lack of e ective interventions. In this context, two types of need can usefully be distinguished (9). Needs for care are de ned as requirements for speci c interventions or activities that are potentially e ective in ameliorating disabling disorders or problems. Children with an intractable problem for which there is no e ective treatment can therefore be described as having no ``meetable need'' even when they are in contact with services. Needs for services are de ned as needs for speci c agents or agencies to deliver these interventions. It is important to distinguish between needs for care and needs for services because a child can be in contact with a service but have an unmet need for a treatment of proven e ectiveness. For example, a recent review suggested that some clinical services in the USA made too much use of psychological therapies of unproven bene t (such as play therapy and `counselling') and too little use of therapies that have been shown to work, such as the cognitive-behavioural therapies (21). In the UK there is concern that stimulant medication, which is of proven bene t in hyperkinetic syndrome, is under-prescribed (5). It is important, then, to consider the evidence-base for the provision of CAMHS. Recent reviews of the provision of child and adolescent mental health services (19, 28, 43) have tended to eschew the evidence base. For example, Wallace and colleagues (43) concluded that ``there are few examples of child and adolescent mental treatment that can unequivocally demonstrate their e ectiveness''. Their blueprint for CAMHS (pp. 91±95) made no mention of the use of the evidence base either as a guiding principle for developing a service or as an indicator of the kinds of treatments that should be included. Similarly, the Health Advisory Service in the UK (19) did not include the use of treatments of proven e ectiveness in their list of principles of providing CAMHS. Rather, priority was given to principles such as easy access and a multidisciplinary approach. Parents seeking help for their children need accessible and comprehensive child mental health services. However, they also have a right to expect that systematic steps have been taken to assess the merits and demerits of the treatments on o er. Unless such steps are taken, there is a danger that parents will have easy access to treatments that are either ine ective or perhaps even harmful! Table 1 shows a system for rating the quality of the evidence base. The system was adapted from the system of the US Task Force On Preventive Care (42) and from Stevens and Raftery (41). Table 2 shows the quality of the evidence supporting the use of selected treatments for some of the most common disorders that are seen by child mental health professionals. As the table indicates, there is in fact evidence for the e cacy of treatments for behavioural disorders (24, 44), depressive disorders (11, 14, 17, 38), hyperkinetic disorders (2) and phobias (26, 27). Less is known about how best to manage conditions such as anxiety disorder and obsessive compulsive disorder. However, for each of these conditions randomized trials have suggested that e ective treatments may exist (1, 6, 20, 25). Not only is progress being made in establishing which treatments are e ective, but evidence is also accumulating about treatments that are probably mostly ine ective. For example, there have been at least six randomized trials of tricyclic antidepressants in childhood depressive disorder and none has found a signi cant di erence between active drug and placebo (18). Need as demand for care or services One last de nition of need requires consideration ± need as demand for services either by parents or by professionals. It is not uncommon to nd that CAMHS are approached to provide treatment because `something must be done'. Levels of demand may be coloured by media responses ± there is now a prevailing view that counselling is automatically required after signi cant life events or traumas. Such requests often require a process of consultation to establish whether involvement of CAMHS is appropriate. Just because another agency or individual demands a service does not mean that a need for that service exists. For example, there are increasing demands for preventive interventions with high risk groups, but there is little evidence that such interventions are e ective in reducing the risk of either behavioural problems (32) or emotional disorders (15) later in life. Indeed, some preventive interventions could cause more harm than good (15). In planning mental health services for children, however, it is important to recognize that mental health professionals are expected to deal not only with disorders (as de ned above) but also with a host of other problems. These include school refusal, deliberate self-harm, and isolated but severe problems such as resetting. They may also be involved in assessment and in the preparation of reports for the courts. Conclusion: measuring the needs of individuals A variety of di erent methods exist for measuring the mental health needs of children and each has strengths and weaknesses. Methods that measure morbidity and service contact would seem particularly well suited for populations in low contact with services, such as school samples or children at high risk of psychopathology. In these samples a child with impairment arising from mental illness who was not in contact with services

4 European Child & Adolescent Psychiatry, Vol. 8, No. 1 (1999) Ó Steinkop Verlag 1999 Table 1 Analysis of service e cacy in child and adolescent mental health Strength of recommendation A B C D E F G Quality of evidence Ia Ib II III IV There is good evidence to support this procedure There is fair evidence to support this procedure There is some evidence to support this procedure There is poor evidence to support this procedure There is some evidence to reject this procedure There is fair evidence to reject this procedure There is good evidence to reject this procedure Consistent evidence obtained from at least ve systematic randomized controlled trials (RCTs) conducted by independent research groups. Consistent evidence obtained from between two to four systematic RCTs by independent research groups. Evidence obtained from one systematic RCT, or from several well designed cohort or case control studies, or from several RCTs in which the results were inconsistent. Opinions of respected authorities based on clinical experience, descriptive studies or reports of expert committees. Evidence inadequate because of problems in methodology, e.g., sample size, length of follow-up, or con ict of evidence. Table adapted from Stevens & Raftery (1997) and from the US Task Force on Preventive Health Care (1989) would often be seen as `in need'. However, estimates of need that are based on existing rates of service use will invariably be only a rough estimate of the actual level of need. Some children will improve rapidly whether treated or not. Others will su er from intractable problems for which no e ective treatment exists, and may therefore be regarded as having no meetable need. Moreover, such estimates assume that children who are in contact with services have had their needs met. As indicated above, this may not be the case. It is important then to establish that children who are known to services have had a needs-led assessment and that an intervention of demonstrable e cacy, if available, has been o ered. At present the type of care that is given to children with mental disorders and their families is often determined by the preferences of the clinician and by the resources of the local service. Without treatment protocols that are based on the existing evidence base there is danger that children will be given ine ective treatments. Measuring the needs of groups The de nitions of needs described in the previous section are clinical and intended to de ne the needs of individuals. In developing needs-led CAMHS for a given district, however, information is also required on the aggregated needs of the population, so-called epidemiologically based needs assessment. It will be appreciated that these two sources of information are not contradictory. In theory, the best way of establishing the needs of a population would be to assess the needs of each individual in it using one or more of the methods described above. However, in practice such information is seldom available, and it is therefore necessary to estimate the needs of the population using other methods. Epidemiology of disorder One of the most widely used methods is to extrapolate from epidemiological research conducted either in another area (36) or nationally (8) in order to estimate the rate of disorder in the local population. This approach can provide helpful information about the relative numbers of cases of one disorder compared with another, and about the age and gender distribution of certain kinds of problems. For instance, in deciding whether child or adult mental health teams should provide services for the 16 to 21 year-old age group, it is important to know that severe psychotic conditions that may require admission to hospital for intensive nursing become much more common during this age period. However, the use of prevalence rates from national surveys is of limited value in planning local CAMHS for three main reasons. First, the prevalence of many childhood mental disorders varies signi cantly between areas. For instance, rates of conduct disorder are higher in deprived areas (31). Second, for all the reasons given in the previous section, it cannot be assumed that the simple prevalence of disorder will equate with meetable need. Thus, even if the `true' prevalence of mental disorder could be estimated, it is not obvious how this would translate into the numbers of child mental health professionals required to provide services. Third, community surveys are seldom large enough to produce accurate estimates of the prevalence of rare disorders such as schizophrenia, anorexia nervosa or repeated

R.C. Harrington et al. 5 Developing needs led child and adolescent mental health services Table 2 E cacy of interventions for selected disorders seen by child mental health professionals Quality of evidence Comments Key reference or review Behavioural disorders Depressive disorder Parent skills training 3±10 A Ia Compliance may be poor in clinically referred samples Webster-Stratton (1991) Cognitive-behaviour therapy 11±17 C Ib Treatment is often lengthy Kazdin (1997) Systemic family therapy 11±17 F Ib Kazdin (1997) Family therapy 10±17 E Ib Two trials of family therapy and two of adding a family intervention ± all negative Cognitive-behaviour therapy 10±17 B Ib Thus far studied only in mildly depressed cases; 1/3 fail to respond Tricyclic medication 10±17 G Ib Six RCTs with negative results. Best used as a second line treatment. Fluoxetine 10±17 C II Two RCTs published thus far, one positive and one negative Harrington et al. (1998a) Harrington et al. (1998b) Hazell et al. (1995) Emslie et al. (1997) Simeon et al. (1990) Hyperkinetic disorder Anxiety disorders Problem Intervention Age Recommendation Obsessivecompulsive disorder Phobic disorder Stimulant medication 4±10 A 1a E ective in around 70% of cases only Behavioural management training 3±10 B 1a Behavior modi cation brings little additional bene t when added to stimulant medication Cognitive-behaviour therapy 10±17 C Ib Three RCTs, two from the same group, with positive results American Academy of Child and Adolescent Psychiatry (1997a) American Academy of Child and Adolescent Psychiatry (1997a) Kendall et al. (1997) Family management 7±14 C II One RCT with a positive result Barrett et al. (1996) Tricyclic antidepressants 9±16 F 1b Three out of four RCTs had a Serotonin reuptake inhibitors (SRIs) negative result 10±16 B Ib The SRIs that have been studied include clomipramine and uoxetine Systematic desensitization 4±16 B 1b Several trials but some with a negative result Behavioural family interventions American Academy of Child and Adolescent Psychiatry (1997b) Heyman (1997) King et al. (1994) 4±16 C III One RCT with a positive result King & Ollendick (1997)

6 European Child & Adolescent Psychiatry, Vol. 8, No. 1 (1999) Ó Steinkop Verlag 1999 deliberate self-harm. Yet it is often these rare problems that require the most intensive health and social interventions. Epidemiology of risk factors Several risk factors are robust correlates of emotional and behavioural disorders in children, including poverty, family discord, school non-attendance, mental retardation, reading retardation and organic brain dysfunction (34, 36). In many countries local data will be available on the epidemiology of some of these factors. For instance, in the UK national census data on social and demographic factors can be used to calculate indices of deprivation such as the underprivileged area score (22). Local agencies may also have information on rates of non-attendance at schools and on children who are looked after by, or known to, social services. Children in the care system have high rates of mental disorders (29). At present there is no reliable method for estimating need or service use from a district's demographic or social characteristics. However, since these data are available in most health districts they can be used for comparative approaches to needs assessment (see later). In theory epidemiological approaches are the ideal way of assessing the needs of groups. In practice, however, those planning CAMHS seldom, if ever, have access to accurate data on the epidemiology of mental disorder amongst children. Needs assessment is therefore much more in uenced by two other approaches: comparative assessments and the corporate approach (40). Comparative needs assessment Comparative approaches involve comparisons within, or between areas, of information on morbidity, service use and provision, costs, and outcomes. As noted earlier, in child mental health few direct measures of morbidity are presently available. However, proxy measures such as social deprivation or the rate of school non-attendance may provide useful pointers about di erences between areas within a district, or between districts. Information on service use and provision can be compared with other districts, or with national estimates of the need for certain kinds of services that have been provided by professional bodies (35) or specialist assessment teams (19). Wallace and colleagues (43) discuss some of these estimates. It will be appreciated that low levels of service provision in comparison with local or national norms do not necessarily mean that there are high levels of unmet need. However, signi cant discrepancies are often a catalyst for change. There are a large number of di erent information systems available for use by CAMHS. Most CAMHS can therefore supply basic information on current levels of activity, such as waiting times, numbers of new cases seen, etc. Such information can provide useful pointers about the functioning of CAMHS. However, a problem can arise when services use di erent methods of data collection (this is very common in the UK) or di erent de nitions of functioning (such as `waiting time'). Similarly, it can be di cult to obtain reliable comparative information about the costs of CAMHS (30). Hospitals may use di erent measures to calculate costs and overheads and although information on the total costs of CAMHS is often available, it is much harder to get reliable information on unit costs. Nevertheless, substantial discrepancies between services in activity levels, waiting lists or costs can highlight inequalities in access to services. There are a large number of ways of measuring the outcomes of interventions for children with emotional or behavioural disorders. Most of these measures, however, have been developed primarily for use in research settings and until recently there were few generic measures of outcome that were suitable for use in routine clinical practice. Such measures are now being developed. For example, the Health of the Nation Outcome Scale for Children and Adolescents (HoNOSCA) has been shown in a large national study in the UK to be both feasible and a valid measure of change (13). It is currently being used in a national survey to compare outcomes both within and between services. Corporate needs assessment Corporate approaches to needs assessment involve the synthesis of views about the mental health needs of children from those people and agencies involved in their care. These include health purchasers, mental health services, primary care physicians, social services, education departments, health service regional o ces, voluntary agencies, parents, and children. It will be appreciated that each of these sources of information will bring di erent perspectives. For instance, most mental health professionals will view needs in terms of services that they supply. Parents, however, may feel that the most important intervention for their child is a placement in a special school. Moreover it can be di cult to obtain a representative view from some groups, such as parents or children. Nevertheless, views of the users of services and of primary care physicians are assuming greater prominence in the planning of CAMHS. Setting priorities The needs assessment procedures reviewed above should help to describe the likely pattern of problems in the

R.C. Harrington et al. 7 Developing needs led child and adolescent mental health services local population, the priorities of professionals and parents, and the areas of greatest unmet need. It is likely, however, that the assessment will identify a multitude of needs for which the capacity to bene t from interventions will be greater than available resources. It will be necessary therefore to set priorities. Identifying priorities Criteria can be set to prioritise needs. Commonly used criteria are severity, prevalence, evidence of e ectiveness of interventions, and feasibility. Many districts will wish to prioritise several groups and the following dimensions may be helpful: Problem groups: type of problem (e.g., deliberate selfharm, school non-attendance), diagnosis (e.g., conduct disorder), severity of problem (e.g., leading to school failure), duration of problem. Socio-demographic groups: gender, age range, geographical area, ethnicity. High risk groups: children known to social services or In Care, children of mentally ill parents, children with mental retardation, children with chronic physical illness. Intervention needed: treatment, education, prevention, accommodation, family support. In the past, many services have prioritised children who experience the greatest burden of formal mental illness, or those who impose the greatest burden on others. Thus, one or more of the following groups have been targeted: children with life-threatening mental disorders (e.g., deliberate self-harm, anorexia nervosa, severe depression) children whose problems cause signi cant problems for other people (e.g., chronic conduct disorder) children with markedly distressing symptoms (e.g., severe obsessive-compulsive disorder) children with chronic mental disorder children at high risk of mental disorder, especially children with mental retardation and children in contact with social services. There has therefore been an emphasis on the provision of interventions for children with acute mental disorders. Recently, however, it has been suggested that there should be greater emphasis on primary prevention and on supporting professional groups who work with children in educational or social services settings (19, 28, 43). For example, the Health Advisory Service model explicitly includes schools in a strategic approach to mental health problems in children. This could lead to the prioritising of the following groups or interventions: children in the general population children at high risk support for health professionals in high contact with children support for teachers preventive school-based programmes. Establishing the philosophy and framework for CAMHS The next step is the development of a framework for providing child and adolescent mental health services. A variety of di erent models have been developed. The tiered community model (10) advocates that CAMHS are seen as consisting of a range of services extending from `tier one' (primary services ± primary care physicians, social workers, teachers, school nurses, etc.), through `tier two' (individual specialist mental health workers such as community psychiatric nurses and clinical psychologists), to `tier three' (specialist multidisciplinary CAMHS) and `tier four' (very specialised inpatient and outpatient interventions). Specialist CAMHS would have a key role in `supporting' or consulting with other professionals. The advantage of this approach is that it proposes a comprehensive model of service provision that could in theory provide coordinated services across agencies. A disadvantage is that its very comprehensiveness could pull services in too many directions at once. The model assumes capacities of analysis and powers of implementation that may not currently exist between health, education and social services. The core mental health service model is exempli ed by Goodman's (12) model in which health provision focuses on `core mental health problems' (such as anorexia nervosa and severe depression) using interventions of proven e cacy. This provision would have strong links with hospital outpatient and liaison services and with intensive services such as inpatient units. Education and social services would take the lead in managing children with behavioural problems. This model has the considerable advantage of using the evidence base as a guide to practice. However, it could lead to the fragmentation of services ± most so-called core mental health problems are strongly associated with social and educational di culties. Moreover, in the model proposed by Goodman (12) there is no formal assessment of needs, so it is unclear whether it would be supported by local health commissioners. Primary care models for CAMHS would locate outpatient work and assessment in health centres. Such models often propose that districts are divided into

8 European Child & Adolescent Psychiatry, Vol. 8, No. 1 (1999) Ó Steinkop Verlag 1999 sectors in which there would be liaison between primary health care, social and educational services, and CAM- HS. However, unless primary care models are very well funded there would inevitably be problems arising from small teams. There would be limited scope for multidisciplinary assessment and a lack of physical resources for some interventions. Assertive outreach treatment models aim to engage patients and families who are unable or unwilling to attend CAMHS. Treatment is conducted in the home (16), in children's homes, or in schools. Choosing between models The choice of model will be in uenced not only by the philosophy of the service but also by several other considerations. First, the availability and organization of existing facilities for the chosen model is likely to be an important pragmatic determinant of whether it can be implemented. For example, it could be di cult for a community service to metamorphose into a core mental health service located on a hospital site. Second, it is necessary to consider whether the model makes the best possible use of the resources available. For example, community-based models may be better able to meet the needs of some client groups. However, because they often lack a viable resourced infrastructure, they may be less e cient than hospital-based services in respect of training, continuing professional development, and sta recruitment or retention. Studies comparing these di erent models of service delivery are necessarily at an early stage. However, the studies conducted thus far have produced some surprises. For example, although it is widely believed that highly accessible community-based interventions will be more e ective than routine outpatient care, a recent randomized trial found this not be the case (16). Clinicians will need therefore to be able to call upon a range of di erent services, some hospital-based and others based in the community. Conclusions Assessing the needs for care of children with emotional or behavioural problems is a complex process. There is no universally accepted de nition of what the term `need' means in this eld. Moreover, in most areas both clinical and epidemiologically based statistical information about need is lacking. Initial estimates of need are likely, then, to be inaccurate and it is important that services are designed to be exible and capable of responding to changes in need. This will mean setting clear targets for services, measuring whether these targets have been met, and being prepared to change services in line with the ndings. Needs assessment exercises are likely to identify high levels of unmet need at both individual and population levels. At the individual level, there is much potential for developing more systematic and evidence-based treatment protocols. Recent research has identi ed several treatments that seem to be e ective (Table 2) yet there is concern that these treatments are not yet widely used. Without systematic treatment plans there is a danger that clinicians will persist with ine ective treatments for too long. At a population level, there is evidence that rates of mental disorder among children vary with local indices such as poverty and ethnicity. It should therefore be possible to set local targets in respect of allocation of resources rather than having to rely on national statistics. It is likely to be di cult to implement change within CAMHS. Many CAMHS are under-funded and professionals will therefore have low expectations of real change. It will be important therefore that all the relevant agencies are involved in the process of assessing needs and setting priorities. The joint planning team should include representatives of CAMHS, local agencies (social services, education, health), the voluntary sector and health purchasers. There should be consultation with other groups, particularly general practitioners and service users. The planning process could thereby become a vehicle of change. References 1. American Academy of Child and Adolescent Psychiatry (1997b) Practice parameters for the assessment and treatment of children and adolescents with anxiety disorders. Journal of the American Academy of Child and Adolescent Psychiatry 36(10 Supplement):69S±84S 2. American Academy of Child and Adolescent Psychiatry (1997a) Practice parameters for the assessment and treatment of children, adolescents, and adults with attention-de cit/hyperactivity disorder. Journal of the American Academy of Child and Adolescent Psychiatry 36(10 Supplement):085S± 121S 3. American Psychiatric Association (1994) Diagnostic and Statistical Manual of Mental Disorders ± DSM-IV (fourth edition). Washington DC: American Psychiatric Association 4. Angold A, Costello EJ (1995) A testretest reliability study of child-reported psychiatric symptoms and diagnoses using the Child and Adolescent Psychiatric Assessment (CAPA-C). Psychological Medicine 25:755±762 5. BAP Consensus Statement (1997) Child and learning disability psychopharmacology. Journal of Psychopharmacology 11:291±294

R.C. Harrington et al. 9 Developing needs led child and adolescent mental health services 6. Barrett PM, Dadds MR, Rapee RM (1996) Family treatment of childhood anxiety: a controlled trial. Journal of Consulting and Clinical Psychology 64:333±342 7. Bickman L (1996) A continuum of care: more is not always better. American Psychologist 51:689±701 8. Bone M, Meltzer H (1989) The Prevalence of Disability among Children. OPCS Surveys of Disability in Great Britain, Report 3. London: HMSO 9. Brewin CR, Wing JK, Mangen SP, Brugha TS, MacCarthy B (1987) Principles and practice of measuring needs in the long-term mentally ill: the MRC Needs for Care Assessment. Psychological Medicine 17:971±981 10. Department of Health (1995) A Handbook on Child and Adolescent Mental Health. London: HMSO 11. Emslie G, Rush A, Weinberg W, Kowatch R, Hughes C, Carmody T, Rintelmann J (1997) A double-blind, randomized placebo-controlled trial of uoxetine in depressed children and adolescents. Archives of General Psychiatry 54:1031±1037 12. Goodman R (1997) Child and Adolescent Mental Health Services: Reasoned Advice to Commissioners and Providers. London: Maudsley Discussion Paper No. 4 13. Gowers SG, Harrington RC, Whitton A, Lelliot P, Beevor A, Wing J, Jezzard B (In press) A brief scale for measuring the outcomes of emotional and behavioural disorders in children: the Health of the Nation Outcome Scales for Children and Adolescents (Ho- NOSCA). British Journal of Psychiatry (still in press) 14. Harrington R, Whittaker J, Shoebridge P, Campbell F (1998b) Systematic review of e cacy of cognitive behaviour therapies in child and adolescent depressive disorder. British Medical Journal 316:1559±1563 15. Harrington RC, Clark A (1998) Prevention and early intervention for depression in adolescence and early adult life. European Archives of Psychiatry and Clinical Neuroscience 248:32±45 16. Harrington RC, Kerfoot M, Dyer E, McNiven F, Gill J, Harrington V, Woodham A, Byford S (1998) Randomized trial of a home based family intervention for children who have deliberately poisoned themselves. Journal of the American Academy of Child and Adolescent Psychiatry 37:512±518 17. Harrington RC, Whittaker J, Shoebridge P (1998a) Psychological treatment of depression in children: a review of treatment research. British Journal of Psychiatry 173:291±298 18. Hazell P, O'Connell D, Heathcote D, Robertson J, Henry D (1995) E cacy of tricyclic drugs in treating child and adolescent depression: a meta-analysis. British Medical Journal 310:897±901 19. Health Advisory Service (1995) Child and Adolescent Mental Health Services. Together we Stand. London: HMSO 20. Heyman I (1997) Children with obsessive compulsive disorder. British Medical Journal 315:444 21. Hibbs ED (1995) Child and adolescent disorders: issues for psychosocial treatment research. Journal of Abnormal Child Psychology 23:1±11 22. Jarman B (1983) Indenti cation of underprivileged areas. British Medical Journal 286:1705±1709 23. Jayson D, Wood AJ, Kroll L, Fraser J, Harrington RC (1998) Which depressed patients respond to cognitivebehavioral treatment? Journal of the American Academy of Child and Adolescent Psychiatry 37:35±39 24. Kazdin AE (1997) Practitioner review: psychosocial treatments for conduct disorder in children. Journal of Child Psychology and Psychiatry 38:161±178 25. Kendall PC, Flannery-Schroeder E, Panichelli-Mindel SM, Southam-Gerow M, Henin A, Warman M (1997) Therapy for youths with anxiety disorders: a second randomized clinical trial. Journal of Consulting and Clinical Psychology 65:366±380 26. King J, Ollendick TH (1997) Annotation: treatment of childhood phobias. Journal of Child Psychology and Psychiatry 38:389±400 27. King NJ, Hamilton DI, Ollendick TH (1994) Children's Phobias. A Behavioural Perspective. Chichester: John Wiley 28. Kurtz Z (1996) Treating Children Well. London: Mental Health Foundation 29. McCann JB, James A, Wilson S, Dunn G (1996) Prevalence of psychiatric disorders in young people in the care system. British Medical Journal 13:1529±1530 30. NHS Executive North West (1997) Child and Adolescent Mental Health. A Review of Services in the North West. Warrington: NHS Executive North West 31. O ord DR, Boyle MH, Racine YA (1991) The epidemiology of antisocial behavior in childhood and adolescence. In: Pepler DJ, Rubin H, (eds) The Development and Treatment of Childhood Aggression. (pp Hillsdale, NJ: Erlbaum 32. O ord DR, Kraemer HC, Kazdin AE, Jensen PS, Harrington RC (In press) Lowering the burden of su ering from child psychiatric disorder: trade-o s among clinical, targeted and universal interventions. Journal of the American Academy of Child and Adolescent Psychiatry 37:686±694 33. Parliamentary Select Committee. Number 4 (1997) Mental Health Services for Children. London: HMSO 34. Richman N, Stevenson J, Graham P (1982) Pre-school to School: a Behavioural Study. London: Academic Press 35. Royal College of Psychiatrists (1990) Child and Adolescent Psychiatry: Into the 1990s. Occasional Paper OP8. London: Royal College of Psychiatrists 36. Rutter M, Tizard J, Whitmore K (1970) Education, Health and Behaviour. London: Longmans 37. Secretary of State for Health (1998) Our Healthier Nation. A Contract for Health. London: The Stationery O ce Limited 38. Simeon JG, Dinicola VF, Ferguson HB, Copping W (1990) Adolescent depression: a placebo-controlled uoxetine treatment study and followup. Progress in Neuro-Psychopharmacology and Biological Psychiatry 14:791±795 39. Simono E, Pickles A, Meyer JM et al. (1997) The Virginia twin study of adolescent behavioral development. In- uences of age, sex, and impairment on rates of disorder. Archives of General Psychiatry 54:801±808 40. Stevens A, Raftery A (1992) The purchasers' information requirements on mental health needs and contracting for mental health services. In: Thornicroft G, Brewin CR, Wing J (eds) Measuring Mental Health Needs (pp 42±61). London: Gaskell 41. Stevens A, Raftery J (1997) Introduction. In: Stevens A, Raftery J (eds) Health Care Needs Assessment. The Epidemiologically Based Needs Assessment Reviews (pp xxxi-xliii). Oxford: Radcli e Medical Press 42. US Preventive Services Task Force (1989) Guide to Clinical Preventive Services. An assessment of the e ectiveness of 169 interventions. Baltimore: Williams & Wilkins 43. Wallace SA, Crown JM, Berger M, Cox AD (1997) Child and adolescent mental health. In: Stevens A, Raftery J (eds) Health Care Needs Assessment. The Epidemiologically Based Needs

10 European Child & Adolescent Psychiatry, Vol. 8, No. 1 (1999) Ó Steinkop Verlag 1999 Assessment Reviews (pp 55±127). Oxford: Radcli e Medical Press 44. Webster-Stratton C (1991) Strategies for helping families with conduct disordered children. Journal of Child Psychology and Psychiatry 32:1047± 1062 45. World Health Organization (1992) The ICD-10 Classi cation of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva: World Health Organization 46. Zahner GEP, Pawelkiewicz W, De- Francesco JJ, Adnopoz J (1992) Children's mental health service needs and utilization patterns in an urban community: an epidemiological assessment. Journal of the American Academy of Child and Adolescent Psychiatry 31:951±960