Mental health services for war-affected children

Abstract

Background

In war-affected societies it is assumed that the major mental health problem facing the population will be stress reactions.

Aims

To describe the creation of a child and adolescent mental health service (CAMHS) in Kosovo after the military conflict ended in 1999, and to establish the range of problems and diagnoses that presented.

Method

Data were collected on 559 patients over 2 years, including their referring problems and diagnoses.

Results

Stress-related disorders constituted only a fifth of the case-load in year 1. A substantial number of patients were symptom-free but attended because they had been exposed to atraumatic event, and believed it might make them ill. Non-organic enuresis and learning disability were the most common diagnoses in year 2. Many patients had a complex mix of social and psychological difficulties that did not fit conventional diagnostic categories.

Conclusions

Mental health services that only address traumatic stress may fail to meet the needs of war-affected children. A comprehensive, culturally appropriate CAMHS is needed to address a wide range of problems including learning disability. It should be developed through local actors, and build on existing local infrastructure. Services can also have an educational role in ‘depathologising’ normative responses.

There is an assumption that the major mental health problem facing war-affected societies will be stress reactions related to the exposure to political violence (Eyber & Ager, 2002; Thabet et al, 2002). The response to this perceived need is often rapid assessments using clinical rating scales to count trauma symptoms (de Jong et al, 2000), followed by the establishment of short-term counselling programmes, independent of existing health structures (Jones, 1995). However, there has been growing criticism of this approach. Symptom counting, using rating scales that have not been culturally validated, may not be the best means of assessing the mental health of the population (Richman, 1993; Jones & Kafetsios, 2002). Trauma counselling programmes have been criticised for being culturally inappropriate, ‘pathologising’ normative responses and undermining local systems of coping (Summerfield, 1999). A preoccupation with post-traumatic stress may obscure the social origins of suffering, and neglect the structural causes of the conflict (Muecke, 1992). Research data from a variety of conflict areas suggest that the majority of children exposed to traumatic events do not develop lasting traumatic reactions (Perrin et al, 2000), and that many recover spontaneously (Thabet & Vostanis, 2000). General mental health problems and disabling psychiatric conditions might not receive the attention they deserve, if the focus is on the identification of acute stress reactions (Jones, 2000; Silove et al, 2000).

METHOD

A locally led child and adolescent psychiatric service was established in Kosovo in the immediate aftermath of the North Atlantic Treaty Organization (NATO) air strikes. Systematic patient data collection and audit over 2 years provided the opportunity to examine the range of problems and diagnoses that would present to a comprehensive child and adolescent mental health service (CAMHS) in the immediate aftermath of war, and to describe the problems confronting the creation of a new mental health service following such military conflict. The project grew out of an emergency mental health programme established for children during the conflict in 1998 by a non-governmental organisation (NGO), Child Advocacy International.

Historical background

Kosovo, formerly an autonomous province of Yugoslavia, had a population of 2 million, 90% of whom were Albanians. In 1989 the province lost its autonomy and many Albanians were dismissed from public service jobs. Health care and medical education were provided through private, usually poorly resourced, parallel structures (Jones, 1993). Full-scale war began in mid-1998. When NATO air strikes ended in June 1999, at least 3500 people had been killed and 800 000 expelled by Yugoslav security services and paramilitary groups (Human Rights Watch, 2001). Since then the United Nations Interim Administration Mission in Kosovo (UNMIK) has administered the province. Albanians returned home and began to come to terms with the destruction and their personal losses. Many Serbs accompanied their own security forces as they withdrew from Kosovo. The remaining Serbian population has found itself in the position of a minority under attack and is fearful for its own security. Other ethnic communities – Roma, Turks and Bosniaks – also had difficulties. Inter-ethnic violence has continued to be a problem.

Mental health needs

Kosovo lacked adequate public health data on mental health needs prior to the conflict. After the air strikes, rural ‘health houses’ (clinics providing a combination of primary health care and regular specialist out-patient clinics) reported seeing two or three children with serious psychological difficulties each day. Many of these children had had psychological difficulties of some kind prior to the period of conflict. Approximately 55% of the total population in Kosovo is under 19 years old (Spiegel & Salama, 1999). The perinatal mortality rate is 33 per 1000 (Gloeb, 2001), putting it on a par with the developing rather than the developed world. It is estimated that 25–30% of all children attending primary health care facilities in developing countries have psychiatric disorders, although less than 20% are identified (Giel et al, 1981). Even without the effects of conflict and the harsh living conditions of the previous decade, one might expect a significant number of children to be in need of mental health services.

Kosovo regards itself as part of western Europe and is culturally complex. It is multi-faith: the largest group being Muslim, many of whom are secular. There are significant Catholic and Orthodox Christian minorities. In urban areas the younger generation have a lifestyle and aspirations similar to their western European counterparts. The rural community espouses values that are more traditional. Throughout the area, the patriarchal extended family is the most significant means of social support.

In the past many children's mental health problems, particularly behavioural disturbance, were seen as primarily the concern of the family. Children with severe disorders were taken to general practitioners or paediatricians. In rural Muslim areas people consulted the local hoxha, a Muslim religious teacher, who for a small donation would provide a specific prayer to be burnt and dissolved in a tea to be taken by the child. Catholic families might also consult their local priest. Neurodevelopmental problems often remained unassessed and untreated. Recent upheavals left many families feeling that their capacity to cope was insufficient for the severity of the problems, while the displacement of large numbers from countryside to city left them without their usual networks of support.

Mental health services

Kosovar psychiatric services, as in the whole of the former Yugoslavia, concentrated resources on a biological and institutional approach to serious mental disorder in adults. In the summer of 1999 the health services were in disarray. Most Serbian doctors chose to leave, or to move to one of the Serbian enclaves. There were 15 Albanian neuropsychiatrists in the province, one of whom had a special interest in children. There were no functioning social services, and only two clinical psychologists. Institutional facilities were degraded and understaffed. Over the subsequent 2 years, with the assistance of UNMIK, the university department of neuropsychiatry and the public psychiatric service were re-established. By 2001 there were 25 residents training in psychiatry. A number of NGOs had set up psychosocial programmes to provide some training in the identification of psychological disorders in children, but there was no local service to which children with problems could be referred. Such programmes gave little attention to serious psychological difficulties.

Child Advocacy International therefore decided to develop a community-based child and adolescent mental health service, rather than a psychotrauma service. The aim from the outset was to create a sustainable, culturally appropriate service to meet the locally identified mental health needs of children and adolescents throughout Kosovo, and to provide a training base for future specialists, as well as residents in general psychiatry. It was to be integrated with paediatric and primary health care services and with adult psychiatric services, which were also being transformed into community-based services. We wished to attend to severe unaddressed needs and to avoid an overextended role that could arise from treating the whole population as traumatised. Education and support for other health professionals and NGO staff were an essential part of the service.

Setting up the CAMHS

The attraction of high rates of pay and additional training means that NGOs can recruit the best qualified professionals, draining the public sector but leaving these staff without employment when funding dries up. To avoid this pitfall and ensure sustainability, Child Advocacy International negotiated with UNMIK and the department of neuropsychiatry to second two psychiatry residents who would work with an expatriate specialist supervisor. They would remain on hospital contract, and return to hospital posts when their training was completed. They would then be in a position to embark on the training of others. Two additional part-time residents and four nurses joined in the second year on a similar basis.

Clinics were located in primary health care facilities and, initially, in the Child Advocacy International office in Prishtina. This allowed for a more accessible, less stigmatising, community-based service with close connections with primary health care. The residents and nurses made home and school visits as needed. The service was open to children of all ethnic backgrounds up to the age of 18 years and to their parents. We also saw older children in higher education, and adults when no other psychiatrist was available. In each town local health professionals, schools and NGOs were informed of our presence. We advertised the service on local radio by providing talks on children's mental health problems. In the first year, the clinics were located in small towns in two of the most conflict-affected areas and in Prishtina, which had doubled in size because of the displaced population. In the second year, with the expansion of the medical team to four doctors, these two former clinics moved to four main towns in order to be integrated with the overall development of community-based mental health services in Kosovo, and to provide access to the greatest number of people.

The aim was to provide both a clinical service and a training opportunity. Training took the form of supervision and mentoring in the clinics and a regular weekly programme of seminars, lectures and case discussions for all the psychiatry residents. Funding was also used to set up internet access, create a comprehensive library, and provide two residents with the opportunity to study for 1–3 months in the UK. Because the consultant supervisors changed every 3 months, the residents encountered a wide variety of approaches, and engaged in a two-way exchange as to the appropriateness of Western systems of diagnosis and treatment in the Kosovar context.

Collection of data

The clinic contact data were collected on attendance sheets at every clinic. They included numbers of new appointments, of follow ups and, in the second year only, of non-attenders. All attending patients and their families completed a simple data collection sheet regarding biographical data, living circumstances, education, the source of referral and referring problem. At discharge the form was completed with ICD–10 diagnosis if any (World Health Organization, 1992), mode of treatment, number of sessions attended and disposal. The data were entered on a computer database and audited.

RESULTS

Figures 1 and 2 illustrate the number of new and follow-up appointments in all the communities over 2 years: 174 new patients were seen in year 1 and 385 were seen in year 2 (559 in total). Data on non-attendance were collected in the second year. Figure 3 shows the age and gender of the patients attending the clinics, combined over the whole period. The majority of adult patients were seen in the first year. Adult data have been excluded from the analysis that follows. The remaining data refer only to the population aged 20 years and under (154 patients in year 1, 376 patients in year 2). The main sources of referral (Figs 4, 5) changed over the 2 years as large numbers of NGOs left, and the medical community and patients themselves became more familiar with what the service could offer. The majority of self-referrals came because families heard about the service on the radio.

Fig. 1 New patient and follow-up appointments in year 1 of the survey.

Fig. 2 Patient contacts in year 2 of the survey.

Fig. 3 Age and gender of patients attending, years 1 and 2 combined. NA, not ascertained.

Fig. 4 Source of referral of patients in year1 (NA, non-attenders; NGO/Int, non-governmental/international organisation).

Fig. 5 Source of referral of patients in year 2 (NGO/Int, non-governmental/international organisation).

Families brought a wide variety of problems to the clinic (Tables 1 and 2). Table 3 shows these problems defined in terms of ICD–10 diagnoses for years 1 and 2. These two tables illustrate a shift in the pattern both of problems referred and of diagnoses made. In the first year the most common reason for referral was exposure to a traumatic event, even if the child was symptom-free, because NGOs (the most common source of referral of this problem) or families were concerned that there might be future problems. In the second year, as the number of NGOs in Kosovo declined, bedwetting (usually primary enuresis) and behavioural problems became the most common reasons for attendance. Special needs and a variety of neurodevelopmental difficulties also became more significant. Behavioural problems summed up by the term nervoz in Kosovo took the form of irritability, or disobedience and aggression. This was often combined with sleep problems, and was distinguished by parents from fear (frike). Table 2 shows that stress-related disorders were the most common diagnosis in the first year but learning disability and non-organic enures is superseded this in the second. In spite of large numbers presenting with behavioural problems, relatively few children met the criteria for conduct disorders. A substantial number did not warrant any psychiatric diagnosis; in others, behavioural problems were a marker for other difficulties – most commonly mild learning difficulties or mood disturbances.

Table 1 Reasons for patients attending the child and adolescent mental health service in year 1

Reason	n	(%)
Exposure to traumatic events	32	18.4
Unspecified	21	12.1
Behavioural problems	17	9.8
Low mood or tearfulness	11	6.3
Bedwetting	10	5.8
Somatic complaints	7	4.0
Fear	7	4.0
Learning difficulties	7	4.0
Nightmares	6	3.5
Sleep problems	5	2.9
Fits	3	1.7
Poor appetite	3	1.7
Strange movements	3	1.7
Family difficulties	2	1.1
Developmental delay	2	1.1
Forensic assessment	2	1.1
Poor communication with others	2	1.1
Poor concentration	2	1.1
Speech problems	2	1.1
Stuttering	2	1.1
Withdrawn	2	1.1
Deaf and dumb	1	0.6
Fainting	1	0.6
Head injury	1	0.6
Mute	1	0.6
Overactivity	1	0.6
Strange behaviour	1	0.6
Total	154

Table 2 Reasons for patients attending the child and adolescent mental health service in year 2

Reason	n	(%)
Bedwetting	61	15.8
Behavioural problems	50	13.0
Fear	29	7.5
Exposure to traumatic events	25	6.5
Learning difficulties	23	6.0
Unspecified	21	5.5
Nightmares	15	3.9
Somatic complaints	15	3.9
Speech problems	15	3.9
Stuttering	13	3.4
Developmental delay	12	3.1
Poor appetite	11	2.9
Fainting	8	2.1
Fits	8	2.1
Poor concentration	8	2.1
Low mood and tearfulness	8	2.1
Assessment for family placement	5	1.3
Sleep problems	5	1.3
Anger	3	0.8
Mute	3	0.8
School refusal	3	0.8
Separation difficulties	3	0.8
Sleepwalking	3	0.8
Specific fear	3	0.8
Strange movements	3	0.8
Medication	2	0.5
Strange feelings	2	0.5
Suicidal ideas	2	0.5
Tremor	2	0.5
Withdrawal	2	0.5
Breathing difficulties	1	0.3
Financial support	1	0.3
Hairpulling	1	0.3
Masturbating	1	0.3
Overactivity	1	0.3
Poor communication with others	1	0.3
Poor growth	1	0.3
Severe mental retardation	1	0.3
Soiling	1	0.3
Strange behaviour	1	0.3
Talking to himself	1	0.3
Tidiness	1	0.3
Weight loss	1	0.3
Total	376

Table 3 Diagnoses and ICD—10 codes for patients attending the child and adolescent mental health service

ICD—10 code	Diagnosis	Year 1		Year 2
		n	(%)	n	(%)
F07.2	Post-concussional syndrome	2	1.3	0	0.0
F20—F23	Schizophrenia, schizotypal and delusional disorders	0	0.0	3	0.8
F31—F34	Mood disorders	7	4.5	5	1.3
F41—F42, F44—F49.3	Neurotic and somatoform disorders	9	5.8	12	3.2
F43.1—F43.2	Stress-related disorders	33	21.4	15	4.0
F50.0	Anorexia nervosa	1	0.6	1	0.3
F51	Non-organic sleep disorders	5	3.2	11	2.9
F60	Personality disorders	0	0.0	1	0.3
F70—F79	Mental retardation	11	7.1	48	12.8
F90	Hyperkinetic disorders	2	1.3	4	1.1
F91—F92.8	Conduct disorders	0	0.0	5	1.3
F93—F94	Disorders of social and emotional functioning	2	1.3	13	3.5
F95—F95.2	Tic disorders	3	1.9	2	0.5
F98.0	Non-organic enuresis	10	6.5	58	15.4
F98.2	Feeding disorder of infancy and childhood	1	0.6	0	0.0
F98.5	Stuttering	2	1.3	15	4.0
	Emotional and physical abuse	1	0.6	0	0.0
	No psychiatric diagnosis	53	34.4	139	37.0
	Assessment not completed	7	4.5	11	2.9
	Still attending	0	0.0	33	8.8
	Missing data	5	3.2	0	0.0

The reduction of cases to problem lists or ICD–10 diagnoses does not illustrate this complexity. Once engaged, many families revealed a mix of difficulties that often required social as well as psychological interventions, as the following case vignette illustrates (all vignettes are fictitious but are drawn from real experiences).

Case vignette 1

Four members of this family were killed when they escaped from their village on a tractor during the war. The youngest daughter, then 7 years old, witnessed the deaths of her two older sisters, her father and an aunt. The mother brought this daughter, now 9, to the clinic because she was refusing school, bedwetting, irritable, crying without reason and having nightmares about her dead relatives. Initially the mother said that these symptoms had begun after the war. Later she stated that the irritability had begun 4 years previously, after an accident at work had left the eldest son (now 22 years old) in a wheelchair. This young man was depressed and occasionally violent and threatening, particularly towards his mother. The family had had social housing before the war. They were now allocated one room in a collective centre in Prizren, a long way from their village. It had a bathroom and kitchen shared with ten families. Their lack of a home prior to the conflict meant that they gained nothing from the aid agencies engaged in rebuilding programmes, and they were low on the list of rehousing priorities for the municipality. The lack of a father and the distance from their remaining extended family meant that the mother felt particularly vulnerable and uncertain of her role and ability to lead her family, while her eldest son felt frustrated at his inability to take on that leadership role.

The daughter's difficulties improved through a combination of play and family therapy conducted on home visits. However, the eldest son's depression and frustration continued. He tried fluoxetine briefly, but found it increased his irritability. We were aware that symptomatic relief was of little significance without improvement in their social circumstances. The mother felt that rehousing would resolve 90% of their problems. The son hoped for some means of active employment. Much of our time was spent pursuing the social agencies responsible, without success. One humanitarian agency was able to offer the daughter a holiday abroad, and at the family's request we also provided transport to a traditional healer who had a reputation for curing paralysis, but he had little effect. Mother and son continue to attend intermittently to discuss family issues, and to pursue rehousing and rehabilitation with our support.

War acted as a precipitant to psychological problems in a variety of ways, not simply through exposure to trauma. Post-war living conditions, the lack of material resources and the destruction of networks of support made previously manageable difficulties seem insuperable. The sudden improvement in the political and security conditions, combined with the influx of humanitarian agencies, allowed families who had had no previous opportunity to do so to access health care for longstanding problems. Sometimes they would initially label the onset as ‘traumatic’, but discussion would reveal the problem as pre-dating the war, or the war would exacerbate pre-war problems such as speech and learning difficulties (case vignette 2).

Case vignette 2

The patient was 7 years old and lived with her father, mother and five siblings in a burnt and partly destroyed house in western Kosovo. A teacher referred the child, who apparently suffered from elective mutism precipitated by the war. The parents described an occasion at the beginning of the war when Serbian soldiers had come to the village, lined the families up in the street and threatened to shoot them. When the child began to cry a soldier had put a gun in her mouth and threatened to shoot her if she did not shut up; the parents said she had not spoken since. A home visit was conducted, which all the family and additional relatives attended. The family lived in one room in very poor conditions with mattresses on the floor. Through careful interviewing, a more complex story of the child's complaint emerged. The child had never learnt to speak, a part from occasional words; she communicated by pointing, and although sociable and friendly with her siblings and friends, she had other developmental problems. She had been bed-wetting until 3 months previously. She was able to help with simple household tasks. There had been behaviour changes since the war: she had become irritable with her friends, was frightened of anyone in uniform and had sleep difficulties in the form of night terrors. However, the parents stated that all these problems had resolved themselves in the past few months except for the speech difficulties, which were the main concern, as the child had just started school. The family was given advice and referred to a speech therapist and for psychological assessment with regard to the developmental delays. Referral to a special school is being considered.

Children and families were offered a wide range of treatment (Fig. 6). The majority of those treated received psychological therapy. Family work was particularly suited to Kosovar culture, because of the importance of the extended family. The effects of the conflict had often thrown even larger groups into close living relationships with one another. Vignette 3 illustrates how patients often combined visits to us with seeing the local hoxha, a practice we supported.

Fig. 6 Treatments provided by the child and adolescent mental health service in years 1 and 2 combined.

Case vignette 3

A 12-year-old boy was referred to a paediatrician because of panic attacks, which had started when a shell had fallen by his school; the child had been knocked over but was uninjured. The local hoxha had seen him, as had an adult neuropsychiatrist. The patient felt the hoxha had been more effective because he had told him he was not suffering from a physical illness but was simply frightened, and provided a clear explanation for his feelings. This had made the patient feel better, but the attacks had persisted and he wanted help in stopping them. He was treated symptomatically with a cognitive–behavioural approach, combining education with relaxation, which resulted in some improvement. The patient stopped attending after six sessions.

The humanitarian community and an active women's movement had also raised awareness of physical and sexual abuse and neglect as problems requiring protection and intervention. For example, there had been a long-standing and continuing problem with young mothers of illegitimate or disabled children abandoning their babies (especially girls) after birth. These babies remain in hospital, and become institutionalised. Child psychiatrists from Child Advocacy International became involved in assisting a multi-agency programme to assess and care for these babies, and promote appropriate fostering and adoption.

When auditing the disposal of patients in both years, we included ‘did not complete treatment’ as one possible means of assessing patient satisfaction. This is a separate measure from the ‘did not attend’ rate counted at clinic contact in year 2 (Fig. 1). That ‘did not attend’ rate includes patients who may have missed some appointments, but returned on a later date and continued their treatment to discharge. In year 1, 29/154 did not return to complete treatment after one appointment. In year 2, the figure was 101/376. In both years the majority of these patients had no psychiatric diagnosis and a significant number had non-organic enuresis. In the second year a number of non-Albanian patients began to attend the clinic (2.3% of the total) and two joint training seminars for Serbian and Albanian mental health nurses were held. Concerns of the Serbian population about their security when travelling in Albanian majority areas remained the major bar to their attendance.

DISCUSSION

Limitations

This is a report of clinical work conducted under difficult conditions. Frequently during the first year clinics had to be closed for lack of power or security. Access was sometimes difficult. The assessment and treatment of patients and the residents' training were the priorities. The supervisors and trainees developed the database over the 2-year period, thus data were not collected or not completed on every patient. For some of the more serious cases (including anorexia, major depression, psychoses and sexual abuse) presenting in the latter part of year 2, the patients are still in attendance, and their final diagnoses are thus not included. However, the data provided are a reasonably accurate reflection of the range and numbers of cases.

What kind of service is needed?

The range of diagnoses, and the age and gender of the patients, are in keeping with Western psychiatric norms. The attendance of a number of adults at child psychiatry clinics, particularly in the first year, highlights the need for post-conflict services to be flexible enough to accommodate the needs of the wider population in the aftermath of an emergency, until normal services are re-established. The high percentage of children with learning disability, enuresis and neurodevelopmental difficulties resembles findings from developing countries. Poor access to health services, poor educational facilities and poor obstetric care have all contributed (Pillay & Lockhat, 1997; Somasundaram & van de Put, 1999). The substantial groups of children with more biologically based problems, and the cluster of older teenagers with serious psychopathological disorders, justify the creation of an inclusive mental health service rather than just a psychotrauma service. However, the large group with stress-related problems and mood disturbances, particularly in the first year, show that any post-conflict mental health service must have the capacity to cope with traumatic reactions, grief and loss, while at the same time recognising that these difficulties are often markers for complex social problems which need to be addressed. The fall in the percentage presenting with stress-related disorders in year 2 is in keeping with the view that many post-traumatic stress reactions in children are self-limiting in the absence of further stresses.

The large number of self-referrals in both years suggests that there were unmet needs in the community. In both years, the largest category was those with no psychiatric diagnosis. In the first year many of these cases were NGO referrals made simply because the child had been exposed to a traumatic event. Both the international and the general community had ‘learnt’ that there might be a reaction, and were concerned. In the second year the service saw a much wider range of problems, including behavioural difficulties and somatic complaints, none of which was serious enough to warrant a psychiatric diagnosis, but most of which benefited from explanation, education and the mobilisation of social support. A psychiatric service can play a significant part in ‘depathologising’ and normalising a war-affected society, and in treating minor disturbance. It is an educational role that we hope other professionals within the community will take on in the future. Many of the less serious presenting problems were difficulties with which families might have coped in normal circumstances, but the difficulties of postwar life, displacement and crowded living conditions had made them insurmountable. Behavioural problems often reflected greater stress in parents rather than an increase in conduct disturbance. Nocturnal enuresis is a much worse problem when beds are shared and there are no adequate facilities for washing sheets.

The most difficult problem was how to provide an adequate response for the large number of children with special needs, in the absence of adequate social services and with limited educational facilities. Our aim was to support the family in dealing with the numerous behavioural problems that occurred at home, as well as diagnosing and treating any accompanying mental illness.

These results are based on a clinical audit, not a community study, but they are likely to be representative of children and adolescents in Kosovo. Political violence was widespread during the war in Kosovo in both rural and urban areas, and in both years our clinics served mixed rural and urban populations in areas that had suffered a great deal and were typical of the wider population. A large number of displaced families had been rehoused by year 2. The greater stability of the population and greater accessibility of the clinics in this second year may have encouraged those with more long-standing, less acute problems to come forward, and increased the proportions of diagnoses such as learning disability.

Cultural challenges

Kosovars are familiar with Western-style adult psychiatric services. The local professionals felt that the main cultural challenge was in reforming a Soviet-style institutionalised and biological service which was inappropriate to the needs of the children. The substantial number of non-attenders after one appointment was thought to be the result of two cultural innovations: the introduction of an appointment-based system, when Kosovars are accustomed to waiting long hours at the clinic door and then being seen that day; and the use of psychological therapies, as opposed to offering a drug-based ‘quick fix’. The non-attenders were mainly patients who had no psychiatric diagnosis. Thus, it is also possible that non-attendance reflected satisfaction with the advice offered in one appointment, so that families saw no need to return (sometimes over long distances) for a follow-up. Similarly, those with primary enuresis received psychoeducation in the first session. Some might have been shy of attending for further group-based therapy. However, we found that among those with the most serious difficulties or behavioural problems, psychological therapies such as family, cognitive–behavioural and play approaches were very acceptable and popular. Neither staff nor patients felt that the service undermined traditional approaches, but rather that it offered an alternative where these had not worked or were not seen as appropriate.

Paradoxically, post-conflict societies may offer their populations improved opportunities, through access to humanitarian aid. Families who could not previously access help can now do so. Problems such as domestic violence and sexual abuse, previously little discussed, begin to be recognised. The ability to identify abuse raises challenges when there are as yet no established mechanisms or facilities for child protection or for dealing with the perpetrators.

The CAMHS has so far failed to provide an adequate service to all the ethnic communities in Kosovo. The lack of resolution of the political situation fosters continuing distrust on both sides. However, the engagement of the Serbian community in local elections suggests that bridge-building might be possible. The new child psychiatry residents are actively engaged in considering how to improve outreach to all Kosovars.

Ensuring sustainability

Many international psychotrauma programmes do not endure after the funding dries up. Banatvala & Zwi (2000) have argued that mental health interventions in complex emergencies should be affordable, effective and culturally valid; they should be based at the community level, and not bypass or undermine established health services; and they should be audited and reviewed to improve the standard of care. We have tried to meet these standards. At the end of the second year of the project, the four seconded staff returned to their hospital posts. One is due to become Kosovo's first child psychiatrist once the Ministry of Health has accredited her training. She is currently responsible for organising the academic and clinical training of eight child psychiatry residents, who between them run out-patient child and adolescent psychiatric clinics in primary health care facilities in the six main towns of Kosovo. Currently the Prishtina clinic is based in the adult psychiatric unit, but the hope is to establish a purpose-built department. In year 3, Child Advocacy International refurbished the newly established clinics and provided an international faculty to teach the academic component of the residents' training. The most significant difficulties at present arise from the lack of trained professionals and resources in social work, psychology and nursing, which would be required for a comprehensive service including a specialised capacity for learning disability. The main challenge in a divided society remains how to reach all sections of the community.

Clinical Implications and Limitations

CLINICAL IMPLICATIONS

1. ▪ In a war-affected society, the development of a sustainable, community-based child and adolescentmental health service that attempts to address the full range of mental health problems may be a more appropriate humanitarian intervention than a psychotrauma service focusing on a single diagnosis.

2. ▪ Many of those attending will have been exposed to trauma but may not be unwell. Psychiatric services working in this context may have a significant role in ‘depathologising’ normal reactions.

3. ▪ The effectiveness of such services will be limited if sufficient attention is not given to the simultaneous development of social, educational and psychological services.

LIMITATIONS

1. ▪ Data collection in the aftermath of conflict is difficult because of limited access and harsh working conditions. The audit does not cover every presenting case.

2. ▪ The final diagnoses of the more seriously affected patients still in attendance at the end of the second year are not included.

3. ▪ The service was largely restricted to the Albanian population of Kosovo because of security concerns among non-Albanian citizens.