Mental health problems of young refugees: Duration of settlement, risk factors and community-based interventions

Abstract

Little is known about the characteristics of young psychologically-distressed refugees in mental health services, and how they vary according to the duration of settlement. This study of 102 young refugees referred to a community-based mental health service describes past adversities and current circumstances, referral problems, service utilization and treatment outcomes using the Strengths and Difficulties Questionnaire (SDQ). The more recently-arrived refugees had significantly higher levels of close exposure to war and violence, were more likely to have suffered separation from immediate family and to have insecure legal status. Those refugees settled longer were significantly more likely to be referred because of conduct problems while there was a trend in recent arrivals to present with internalizing pathology. A comparison of the teachers' and parents' mean SDQ scores of the study's young refugees sample and a national study representative of Great Britain as a whole showed that young refugees have higher scores in total problem and all subscales scores than the British scores. Community-based mental health services for young refugees appeared effective - significant improvement was found in SDQ scores for the sub-group (n = 24) who took up the treatments offered. The implications are discussed for service development and practitioners.

Full text

Clinical Child Psychology
and Psychiatry
18(4) 604 –623
© The Author(s) 2012
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1359104512462549
ccp.sagepub.com
Mental health problems of young
refugees: Duration of settlement,
risk factors and community-based
interventions
Glòria Durà-Vilà1, Henrika Klasen2, Zethu
Makatini3, Zohreh Rahimi4 and Matthew Hodes5
1University College London, UK
2De Jutters Centre for Youth, The Hague, The Netherlands
3East London NHS Foundation Trust, UK
4Medical Foundation for the Care of Victims of Torture, London, UK
5Imperial College London, UK
Abstract
Little is known about the characteristics of young psychologically-distressed refugees in mental
health services, and how they vary according to the duration of settlement. This study of 102
young refugees referred to a community-based mental health service describes past adversities
and current circumstances, referral problems, service utilization and treatment outcomes
using the Strengths and Difficulties Questionnaire (SDQ). The more recently-arrived refugees
had significantly higher levels of close exposure to war and violence, were more likely to have
suffered separation from immediate family and to have insecure legal status. Those refugees
settled longer were significantly more likely to be referred because of conduct problems while
there was a trend in recent arrivals to present with internalizing pathology. A comparison of the
teachers’ and parents’ mean SDQ scores of the study’s young refugees sample and a national study
representative of Great Britain as a whole showed that young refugees have higher scores in total
problem and all subscales scores than the British scores. Community-based mental health services
for young refugees appeared effective – significant improvement was found in SDQ scores for
the sub-group (n = 24) who took up the treatments offered. The implications are discussed for
service development and practitioners.
Keywords
Community-based interventions, risk factors, mental health, young refugees
Corresponding author:
Matthew Hodes, Academic Unit of Child & Adolescent Psychiatry, Imperial College London, St Mary’s Campus, Norfolk
Place, London W2 1PG, UK.
Email: m.hodes@imperial.ac.uk
462549CCP18410.1177/1359104512462549Clinical Child Psychology and PsychiatryDurà-Vilà et al.
2012
Article

Durà-Vilà et al. 605
Introduction
The United Kingdom (UK) has a centuries-old tradition of receiving immigrants (Kushner &
Knox, 1999; Winder, 2004). From 2000 to 2004, UK was the European country which had
most asylum applications submitted (United Nations High Commissioner for Refugees
[UNHCR], 2006). In 2004, there were 33,960 asylum applications in the UK. Most of the
applicants were from Iran, Somalia, China, Zimbabwe and Pakistan (Home Office, 2005). The
number of asylum applications to the UK has been reduced: in 2006 there were 23,610 appli-
cations, of which 6,225 (26%) resulted in grants of asylum, humanitarian protection or discre-
tionary leave (Home Office, 2007). Since then the number of asylum applications has been
steady, with 24,485 in 2009, but the proportion granted asylum has reduced, reaching 17% in
2009 (Home Office, 2010). Refugees are a heterogeneous group, as in addition to the usual
demographic variables, this population may be ethnically and linguistically diverse as well as
varied with respect to the kinds of adversities experienced. The legal status of those who leave
their own countries is also an important issue, as this will influence their economic, employ-
ment and housing conditions. Refugee status gives entitlement to work and benefits of the
welfare state (refugees are those whose legal rights are recognized, and asylum seekers are
those applying for refugee status) (UNHCR, 2006).
Refugee children’s psycho-social well-being is affected by a number of risk factors. Their risk
of developing psychopathology is influenced by the type and intensity of exposure to adversities in
their home countries. The killing of family members, or the torture or imprisonment of community
members witnessed first-hand, represent the maximum level of threat (Espino, 1991; Garbarino &
Kostelny, 1996). Organized violence may disrupt community life including the educational provi-
sion and economic activity resulting in privation (Pedersen, 2002; UNHCR, 2006).
Adverse experiences during the flight such as disruption of relationships and privation of food
and shelter – due to a rapid departure in pursuit of safety – are very stressful events. When the
asylum seekers or refugees finally arrive in their host countries, they may be subjected to other
stressful events such as economic hardship, culture and language change, social isolation, family
disruption, hostility and racism (Howard & Hodes, 2000; UNHCR, 2006). Refugee children may
also be exposed in the receiving countries to bullying, difficulties in peer relationships, genera-
tional differences in traditional values and family organization, and deterioration in parenting and
in family relationships (Almqvist & Broberg, 1999; Hodes, 2002a, 2002b).
The past adversities together with ongoing risk factors leads to an increase in the risk of
developing psychiatric disorders especially posttraumatic stress disorder (PTSD), depression
and conduct disorder (Fazel, Wheeler, & Danesh, 2005; Hodes, 2008; Tousignant et al., 1999).
There is evidence showing that the impact of exposure to war and persecution in refugees
decreases over time in resettlement countries (Almqvist & Broberg, 1999; Dyregrov, Gjestad, &
Raundalen, 2002; Weine et al., 1995). Furthermore, some of the adversities encountered in the
receiving countries might also diminish over time, with associated reduction in psychological
distress especially depression (Beiser, Dion, Gotowiec, Hyman, & Vu, 1995; Sack, Him, &
Dickason, 1999). Over time, asylum seekers obtain refugee status and have longer periods in
stable accommodation, learn the local language, build up social networks and take up educa-
tional or professional opportunities (Hauff & Vaglum, 1997; Laban, Gernaat, Komproe, van der
Tweel, & De Jong, 2005). The studies converge in suggesting changes in adaptation for children
and adults and reduced psychological distress one to two years after war experiences and resi-
dence in resettlement countries (Becker, Weine, Vojvoda, & McGlashan, 1999; Dyregrov et al.,
2002; Laban et al., 2005; Weine et al., 1995).

606 Clinical Child Psychology and Psychiatry 18(4)
There are a number of factors influencing young refugees’ use of mental health services
in resettlement countries. The barriers that asylum seekers and refugees experience in being
referred to specialist mental health services include language fluency, high mobility and dif-
ficulty in registering with general practitioners. Refugees with psychiatric morbidity may
also experience difficulties in accessing clinic based mental health services. They may per-
ceive mental health services as stigmatizing. Furthermore, the patient’s own description of
the symptoms may be in terms of a somatic rather than a psychological presentation. With
the difficulties that might prevent primary care referrals to specialist mental health services,
it is striking that refugee children have different referral pathways – they are more likely to
be referred by a non-medical agency (school, social services, etc.) – than non-refugee peers
(Howard & Hodes, 2000). Due to the limitations for refugee children in accessing mental
health services through primary care referral, attention has been given to services provided
in community settings. The benefits of school based services include the important role of
teachers in identifying psychologically distressed children, provision of access in non-
stigmatizing settings, without reliance on negotiating referral to secondary-level clinic-based
services from primary care (Fazel, Doll, & Stein, 2009; Hodes, 2002a, 2002b; Rousseau &
Guzder, 2008).
An account of the pilot study of the community-based child mental health service established
in London for children and adolescents from asylum-seeking and refugee backgrounds has previ-
ously been published (Hodes, 2002b; O’Shea, Hodes, Down, & Bramley, 2000). The current
report examines the work undertaken by this community service over three years. The aims of the
report are to investigate: firstly, whether amongst young refugees seeking treatment the back-
grounds and living situation differed according to the duration of settlement; secondly, to investi-
gate the importance of duration of settlement on the level of psychological distress and treatment
uptake and progress. The third aim was to investigate the outcomes of psychological interven-
tions. For the first and second aims, duration of settlement of two years was used as the threshold
for dichotomizing the group. The justification for this division is that studies converge in suggest-
ing changes in adaptation and reduced psychological distress two years after war experiences, and
residence in resettlement in countries (Becker et al., 1999; Dyregrov et al., 2002; Laban et al.,
2005; Weine et al., 1995).
Methods
The service
A community-based mental health service was established for children and adolescents from
asylum-seeking and refugee backgrounds. The feasibility of the service had been established in
the pilot project (O’Shea et al., 2000), and the service on which this report is based was pro-
vided during 1996–2001. The aims of the service were to provide direct therapeutic work to the
children and families, consultation with teachers and other staff about individual children,
training with a view to increasing the teachers and other staff’s competence in the recognition
and management of children’s psychological distress, and appropriate referral to specialist
child and adolescent mental health services. It was delivered by community-based child and
adolescent mental health professionals, functioning as primary mental health workers (Health
Advisory Service, 1995). The staff consisted of family therapists, child and adolescent psychi-
atric nurses, and a higher trainee in child psychiatry who provided sessional input to the service
over a six-year period. The service was based in institutions known to have contact with many

Durà-Vilà et al. 607
of the local refugee population: in a primary school, a secondary school and a voluntary home-
less family service situated in the borough of Westminster, London. Westminster has a multi-
cultural population with many residents from asylum seeking/refugee backgrounds. According
to the 2001 census, 52% of the Westminster population described themselves as white British
or Irish, 21% white other (this includes refugees from the Balkans), 8.9% as Asian, 7.4% as
Black or Black British, 6.3% as Chinese.
The intervention
Teachers and social or voluntary workers referred children with psychological distress or prob-
lems. In some instances consultations followed, and some children or families were seen after the
initial referral. An interpreter was provided when it was necessary. A variety of flexible treatment
and management options were available. They included liaison with other agencies, problem-solving
and practical help as well as direct therapeutic work. Individual psychotherapy based on narrative
work, supportive treatments, family therapy and cognitive work addressing issues of loss were
amongst the treatment models used.
The sample
A total of 102 children were referred to the community service, the majority (61 [60%]) by schools,
in particular through special educational needs coordinators. 43 (42%) had been in the UK less
than two years and 42 (41%) had been in the UK longer than two years. For 17 (17%) of the young
people the duration of settlement was unknown. The age range was 3–17 (mean age 10.1 years old)
and the male:female ratio was 3:1. The countries of origin reflect the areas where there has been
war during the study period: 45 (44%) of the young refugees were from the Middle East; 27 (27%)
were African; 23 (22%) were European and 7 (7%) from elsewhere.
Their main languages were Kurdish 25 (25%), Arabic 25 (25%), Albanian 16 (15%) and Somali
14 (14%). Of those seen directly interpreters were required in 63% (n = 54) by the family and in
42% (n = 37) by the child. Basic demographic information of age, sex and country of origin was
available for all children and there was no significant difference for these variables between chil-
dren who had recently arrived in the UK and those who had been in the country for two years or
longer. The sample concerns children and families referred to the service during 1995–2001, and
described in the pilot studies (Hodes, 2002b; O’Shea et al., 2000).
Measures and data collection
The measures used were those established in a pilot study carried out in a primary school, which
has been previously reported (O’Shea et al., 2000). This was a prospective study that relied on
clinical service data, obtained from teachers, and for the majority supplemented by data from
parents and children and adolescents themselves. Teachers completed a referral form that cov-
ered demographic information, fluency in English and their own perception of the child’s diffi-
culties in the form of a problem list and severity rating. A brief description of the teacher’s views
regarding the causes of the child’s difficulties was requested in the proforma. The children’s
symptoms were assessed with the Strengths and Difficulties Questionnaire (SDQ) (Goodman,
1997). This 25-item questionnaire has five subscales, each with five items covering: hyperactiv-
ity/inattention, conduct, emotional and peer problems, and also five items for prosocial behav-
iour. The scores of the problem subscales (prosocial subscale is not included as a problem

608 Clinical Child Psychology and Psychiatry 18(4)
subscale) can be summated to give a total score which gives the risk of disorder. The SDQ has
good psychometric properties (Goodman, 2001), and has been widely used in culturally diverse
settings (and translated into more than 40 languages) (Bourdon, Goodman, Rae, Simpson, &
Koretz., 2005; Goodman, Renfrew, & Mullick, 2000). The SDQ was completed by the referring
teacher in 47/102 (46%) cases, the parents in 27/102 (26%) cases or the adolescents themselves
in 5/102 cases (5%). When possible the SDQ was used again after treatment to assess outcome
(teachers’ SDQs were completed post-intervention in 27/102 [26%] cases and parents’ SDQs in
13/102 [13%] cases).
Clinical interviews by the child mental health practitioners were used to gain information on
demographic details, past adversities, current risk factors, family history and current mental health
problems. The family and child gave an account of their experiences of war and violence which
were classified into three categories adapted from the clinical rating scale for exposure to war and
violence (Espino, 1991; Howard & Hodes, 2000). These were: ‘personal experience of war’ which
included experiences like the bombing of their own or neighbouring houses or witnessing shoot-
ings, ‘distant experience of war’ which included for example exposure to distant shell fire or living
in an endemic war area and ‘no experience of war’. Similarly, ‘direct exposure to violence’ included
experiencing or witnessing torture, beatings, imprisonment or arrests. The second category of ‘vio-
lence to friends/family not witnessed’ means the young person knew of violence committed to
close family or friends, but was not present during these acts, and the final category was ‘no direct
or indirect exposure to violence’.
Information about basic socio-demographic details and treatments was available for practically
all the children referred to the service. The amount of data collected varied from case to case since
some children were referred but only discussed in consultation, while others had assessments and
some had treatments of variable numbers of sessions.
The data was collected in routine clinical practice as required by the commissioner that funded
the service (Kensington Chelsea and Westminster Health Authority). In view of this approval of the
project from a research ethics committee was not required.
Data analysis
Descriptive statistics were used for the demographic characteristics of the children and families,
past adversities and current risk factors (Table 1), referred problems and psychopathology (Table 2),
service provision and outcome (Table 3) and SDQ scores at referral, and SDQ scores pre- and post-
intervention (Table 4).
Fisher’s exact test and Pearson Chi-square test with or without Yates’ correction (‘2 x n’
tables) with 95% confidence intervals were applied, depending on the size of the cells, to com-
pare the data collected between the two groups of young refugees (children who had been two
years or less or more than two years in the UK), (Chi-Square values, degrees of freedom (df)
and 2-sided p-values are provided in Tables 1, 2 and 4). Independent t-tests were used to com-
pare teacher and parent SDQ scores at referral between the two groups of young refugees. The
SDQ scores were regarded as a continuous measure with each case as its own control, so a
Paired-Samples t-test was employed to compare the SDQ scores at baseline and after interven-
tion (pre- and post-intervention SDQ scores), (values of t, df and 2-tailed p-values are provided
in Tables 3 and 5). In addition to this, linear regression was used to elicit which risk factors
significantly contributed to the prediction of teachers’ pre-intervention SDQ scores (discussed
further in the outcome section).
The data was analysed using the Statistical Package for Social Sciences (SPSS 15.0, 2003).

Durà-Vilà et al. 609
Table 1. Past and current adversities.
Whole group Time in UK 2
years or less
Time in UK more
than 2 years
Chi-square
(value, df)
p-value
Past separation from
immediate family
(n = 87) (n = 38) (n = 38)
n (%) n (%) n (%)
-yes, political reason 49 (56%) 29 (76%) 17 (45%) 7.936, 2 0.019
-yes, marital or other 21 (24%) 5 (13%) 12 (32%)
-no 17 (20%) 4 (11%) 9 (24%)
Past exposure to violence (n = 70) (n = 33) (n = 34)
n (%) n (%) n (%)
-yes, direct or witnessed 22 (32%) 15 (45%) 5 (15%) 10.749, 2 0.005
-yes, to friends or family 29 (41%) 13 (40%) 13 (38%)
-no 19 (27%) 5 (15%) 16 (47%)
Past exposure to war (n = 76) (n = 35) (n = 37)
n (%) n (%) n (%)
-yes, personal experience 30 (39%) 23 (66%) 6 (16%) 18.353, 2 0.000
-yes, at a distance 37 (49%) 10 (29%) 25 (68%)
-no 9 (12%) 2 (6%) 6 (16%)
Family member disappeared
or killed
(n = 57) (n = 29) (n = 27)
n (%) n (%) n (%)
-yes 29 (51%) 17 (59%) 12 (45%) 1.125, 1 0.288
-no 28 (49%) 12 (41%) 15 (55%)
Refugee status (n = 79) (n = 34) (n = 36)
n (%) n (%) n (%)
-insecure 39 (49%) 25 (73%) 6 (17%) 22.914, 1 0.000
-secure 40 (51%) 9 (27%) 30 (83%)
English language skills (n = 86) (n = 39) (n = 37)
n (%) n (%) n (%)
-Poor, interpreter needed 54 (63%) 29 (74%) 15 (41%) 8.908, 1 0.003
-Adequate, no interpreter 22 (37%) 10 (26%) 22 (59%)
Family in the UK (n = 70) (n = 33) (n = 30)
n (%) n (%) n (%)
-supportive family 32 (46%) 13 (39%) 13 (43%) 0.101, 1 0.751
-no supportive family 38 (54%) 20 (61%) 17 (57%)
Friends in the UK (n = 67) (n = 31) (n = 29)
n (%) n (%) n (%)
-supportive friends 34 (51%) 9 (29%) 17 (59%) 5.342, 1 0.021
-no supportive friends 33 (49%) 22 (71%) 12 (41%)
Housing (n = 92) (n = 41) (n = 40)
n (%) n (%) n (%)
-temporary/inadequate 61 (66%) 29 (79%) 21 (52%) 2.849, 1 0.091
-stable/adequate 31 (34%) 12 (21%) 19 (48%)
Mental health of carer (n = 79) (n = 33) (n = 41)
n (%) n (%) n (%)
-poor, receiving psychiatric
treatment
10 (13%) 4 (12%) 6 (15%) 0.111, 2 0.946
-poor, no treatment 27 (34%) 12 (36%) 14 (34%)
-no reported problems 42 (53%) 17 (52%) 21 (51%)

610 Clinical Child Psychology and Psychiatry 18(4)
Table 2. Referred problem and psychopathology.
Whole group Time in UK 2
years or less
Time in UK more
than 2 years
Chi-square
(value, df)
p-value
Reasons for referral
(may be multiple)
(n = 99) (n = 42) (n = 39)
n (%) n (%) n (%)
-Concentration problem 49 (49%) 20 (48%) 24 (62%) 1.263, 1 0.261
-Peer relations difficulties 32 (32%) 18 (43%) 9 (23%) 3.844, 1 0.050
-Emotional problem 52 (52%) 26 (62%) 18 (46%) 2.355, 1 0.125
-Learning problem 31 (31%) 13 (31%) 14 (36%) 0.152, 1 0.697
-Conduct problem 36 (36%) 10 (24%) 19 (49%) 4.634, 1 0.031
Main problem/diagnosis
according to therapist
(n = 99) (n = 42) (n = 41)
n (%) n (%) n (%)
-Internalising 41 (41%) 21 (50%) 13 (32%) 3.567, 1 0.059
-Externalising 22 (22%) 6 (14%) 12 (29%) 1.286, 1 0.256
-Mixed or other 36 (36%) 15 (36%) 16 (39%) 0.057, 1 0.811
Table 3. Teacher and parent SDQ scores at referral.
Whole group Time in UK 2
years or less
Time in UK more
than 2 years
Independent
t-test (value, df)
p-value
Teacher SDQ scores
at referral
(n = 48) (n = 19) (n = 24)
Mean (SD) Mean (SD) Mean (SD)
-Total problem score 15.37 (6.63) 14.36 (7.07) 15.66 (6.83) −0.609, 41 0.546
-Hyperactivity score 6.38 (3.09) 6.00 (3.46) 6.54 (2.91) −.556, 41 0.581
-Peer problem score 2.70 (2.01) 3.21 (2.25) 2.29 (1.85) 1.469, 41 0.149
-Emotional problem
score
3.38 (2.74) 3.11 (2.82) 3.33 (2.87) −.261, 41 0.796
-Conduct problem
score
2.81 (2.47) 2.05 (1.81) 3.29 (2.88) −.634, 41 0.110
Parent SDQ scores
at referral
(n = 27) (n = 11) (n = 14)
Mean (SD) Mean (SD) Mean (SD)
-Total problem score 13.59 (5.70) 13.64 (6.45) 12.71 (4.78) 0.411, 23 0.685
-Hyperactivity score 5.33 (2.76) 4.64 (2.50) 5.50 (2.87) −0.788, 23 0.439
-Peer problem score 2.30 (2.12) 2.55 (2.69) 1.86 (1.61) 0.794, 23 0.435
-Emotional problem
score
3.30 (2.82) 3.82 (2.92) 3.07 (2.97) 0.628, 23 0.536
- Conduct problem
score
2.67 (2.52) 2.64 (2.83) 2.29 (2.05) 0.359, 23 0.723
Results
Risk factors
Most children who were referred lived with at least one natural parent (67/99 [68%]), and living
with single and two parents were equally frequent. It was expected that more children would be
living with two parents in the group that have been in UK longer, as families had more time to

Durà-Vilà et al. 611
Table 4. Comparisons between the teacher and parent SDQ scores at referral and teacher SDQ scores
from a national study.
Teacher SDQ
young refugees
(n = 48)
Teacher SDQ
British overall
norms (n = 8,208)
Parent SDQ
young refugees
(n = 27)
Parent SDQ
British overall
norms
(n = 10,298)
Total problem score
(normal range: 0–11)
Mean (SD) Mean (SD) Mean (SD) Mean (SD)
15.4 (6.6) 6.6 (6.0) 13.6 (5.7) 8.4 (5.8)
Hyperactivity score
(normal range: 0–5)
6.4 (3.1) 2.9 (2.8) 5.3 (2.8) 3.5 (2.6)
Peer problem score
(normal range: 0–3)
2.7 (2.0) 1.4 (1.8) 2.3 (2.1) 1.5 (1.7)
Emotional problem score
(normal range: 0–4)
3.4 (2.7) 1.4 (1.9) 3.3 (2.8) 1.9 (2.0)
Conduct problem score
(normal range: 0–2)
2.8 (2.5) 0.9 (1.6) 2.7 (2.5) 1.6 (1.7)
Note. National averages come from a national study with 8,208 teachers and 10,298 parents of children aged 5–15 years
representative of Great Britain as a whole (Meltzer et al., 2000; Youth in mind, 2001).
Table 5. Treatment and outcome.
Whole group Time in UK 2
years or less
Time in UK more
than 2 years
Chi-square (value, df)
or Fisher ex. (value)
p-value
Number of sessions (n = 92) (n = 39) (n = 37)
n (%) n (%) n (%)
-No face to face 18 (20%) 6 (15%) 0 (0%) 6.919, 3 0.075
-One or two 20 (22%) 7 (18%) 11 (30%)
-3 to 9 25 (27%) 11 (28%) 12 (32%)
-More than 9 29 (31%) 15 (39%) 14 (38%)
Type of treatment (n = 102) (n = 43) (n = 42)
n (%) n (%) n (%)
-Family 59 (58%) 29 (67%) 28 (67%) 9.176, 3 0.027
-Individual 15 (15%) 3 (7%) 10 (24%)
-Group 11 (11%) 6 (14%) 4 (10%)
-No face to face 17 (17%) 5 (12%) 0 (0%)
Type of discharge (n = 102) (n = 43) (n = 42)
n (%) n (%) n (%)
-Unplanned 34 (33%) 11 (26%) 9 (21%) 0.898, 3 0.826
-Negotiated 41 (40%) 20 (47%) 21 (50%)
-Ref. to specialist 12 (12%) 6 (14%) 4 (10%)
-Ongoing 15 (15%) 6 (14%) 8 (19%)
Change accord. to
therapist
(n = 48) (n = 21) (n = 23)
n (%) n (%) n (%)
-Improved 36 (75%) 14 (67%) 21 (91%) 0.993 0.064
-Same/worse 12 (25%) 7 (33%) 2 (9%)

612 Clinical Child Psychology and Psychiatry 18(4)
reunify, however, that was not the case (χ2 = 0.973, 1 df, p = 0.419). About 83/91 (91%) of the
children had siblings. 28/91 (31%) had three or more siblings.
Table 1 shows the past and current risk factors of the refugee children who had been in the UK
for less than two years compared to those who had been here longer than two years. Past adversities
included separation from close family, exposure to violence, personal experience of war and family
members being killed or disappearing. In all these areas levels of adversity were significantly
higher for the children who had arrived in the UK less than two years ago.
70 children (80%) had experienced a period of separation from their immediate family, in the
main, due to political reasons and fears for safety. Separation due to marital conflict was also com-
mon especially in those who had been in the UK longer than two years. Refugees who arrived less
than two years ago reported significantly higher levels of separation for political reasons than those
who had been in the UK for more than two years (29/38 [76%] compared to 17/38 [45%], χ2 =
7.936, 2 df, p = 0.019).
In terms of exposure to violence, statistically significant differences were found between those who
have been in the UK for two years or less and those who have been more than two years (χ2 = 10.749,
2 df, p = 0.005). 15/33 (45%) of those in the UK for two years or less reported direct or witnessed experi-
ence of violence compared with 5/34 (15%) of those who had been in the UK for more than two years
(χ2 = 6.164, 1 df, p = 0.013). An additional 13/33 (40%) of those in the UK for two years or less expe-
rienced violence indirectly as it was directed to their close friends or family without reaching an statisti-
cally different level. The numbers of young refugees who were not exposed to violence were: 16/34
(47%) for those who have been settled for longer than two years and 5/33 (15%) for those who have
been settled for two years or less. This difference was highly significant (χ2 = 7.936, 1 df, p = 0.004).
The difference in terms of past exposure to war is even more striking (χ2 = 18.353, 2 df, p =
0.000). 23/35 (66%) of the recent refugees had close experiences of war, while this is only the case
for 6/37 (16%) of those who have been settled in the UK for longer than two years (χ2 = 16.320, 1
Table 6. Strength and Difficulties Questionnaire (SDQ) score changes where both pre and post
intervention measures are available.
PRE-intervention
SDQ
POST-intervention
SDQ
Paired-Samples
t-test (value, df)p-value
Teachers’ SDQ scores n = 24 n = 24
Mean (SD) Mean (SD)
-Total score 14.67 (6.66) 11.04 (6.92) 2.787, 23 0.010
-Hyperactivity score 6.38 (3.29) 5.04 (3.43) 2.635, 23 0.015
-Emotional problems score 3.29 (2.97) 2.33 (2.39) 1.775, 23 0.089
-Conduct problem score 2.08 (2.15) 2.29 (2.26) −0.411, 23 0.685
-Peer problem score 2.77 (2.18) 1.58 (1.81) 2.551, 25 0.017
-Pro-social score 5.83 (2.58) 6.67 (1.99) −1.541, 23 0.137
Parents’ SDQ scores n = 11 n = 11
Mean (SD) Mean (SD)
-Total score 13.64 (6.02) 7.18 (5.76) 3.425, 10 0.006
-Hyperactivity 4.82 (2.75) 1.82 (1.47) 5.244, 10 0.000
-Emotional problem score 3.91 (3.11) 2.27 (1.95) 1.618, 10 0.137
-Conduct problem score 2.55 (2.77) 1.64 (2.06) 2.319, 10 0.043
-Peer problem score 2.36 (2.77) 1.55 (2.11) 1.014, 10 0.335
-Pro-social score 8.45 (2.58) 8.73 (2.41) −0.820, 10 0.434

Durà-Vilà et al. 613
df, p = 0.000). In terms of exposure to war at a distance, significant differences were also found:
25/37 (68%) of the more settled refugees and 10/35 (29%) for the more recent refugees (χ2 =
10.949, 1 df, p = 0.000). There were not statistically significant differences between the two groups
in reporting a family member killed or missing (χ2 = 1.125, 1 df, p = 0.288).
After arrival in the UK, the young refugees continued to experience serious adversities. These
included insecurity about their refugee status, lack of social networks, inadequate housing, poor
parental mental health, parental unemployment and poor English. In some instances, these adversi-
ties were equally experienced by those settled for more than two years but in other instances, recent
migrants experienced significant higher levels of current adversities.
In terms of refugee status, we categorized full asylum status and exceptional leave to remain as
‘secure’, while those with asylum seeker status, temporary leave to remain for six months or
refused status were classified as ‘insecure’. Of those in the country more than two years, 30/36
(83%) were in a relatively ‘secure’ position. This was only the case for 9/34 (27%) of those who
had been in the UK two years or less. Not surprisingly, there was a significant difference in security
of refugee status between the recent arrivals and those settled for more than two years (χ2 = 22.942,
1 df, p = 0.000). Of the recent arrivals, significantly fewer had a secure status.
The child’s fluency in English was significantly better in those cases where families had been in
the country for more than two years: 22/37 (59%) of these children spoke at least adequate English,
while this was only the case for just over a quarter of those who had arrived within two years
(10/39 [26%]) (χ2 = 8.908, 1 df, p = 0.003).
With regard to social networks, interestingly both recent as well as more settled refugees had
similar family networks in the UK. Almost half of the families had relatives in this country and
there was no significant difference between the two groups. There was a significant difference,
however, in terms of supportive friends: only 9/31 (29%) of recent migrants had supportive friends,
while this was the case for 17/29 (59%) of those settled for longer than two years (χ2 = 5.342, 1 df,
p = 0.021).
The social situation of the families was, as expected, very difficult. Of the whole sample, 61/92
(66%) families lived in overcrowded housing without adequate privacy or space (mainly hotel
accommodation, but also hostels and a refuge). Of those with more satisfactory housing, 31/92
(34%) families lived in council flats (that is, flats provided by the Local Authority, Westminster) or
in the case of unaccompanied children, in foster families or children’s homes. A large majority of
the recently arrived group, 29/41 (79%), lived in overcrowded hotels or hostels with no privacy,
compared with 21/40 (52%) of the comparison group (trend for significance, p = 0.091). It has to
be noted that only 19/40 (48%) had found adequate housing such as a council flat even two years
after arrival in the UK.
Parents of the children had also experienced considerable adversity and mental health risk fac-
tors. Almost half of them had mental health problems but only 10/79 (13%) of the whole sample
received psychiatric help. There was no significant difference with regard to parental mental health
problems and psychiatric input between the recent and more settled refugees. Equally, there was no
difference in their rate of unemployment which was over 90% for the head of the household.
Psychopathology
In accordance with the nature of the project which was based in community settings, referrals came
from schools, the homeless family service, social services, and health services. The education sec-
tor was particularly important in identifying problems with the young refugees. In those who had
been in the country for longer than two years, school teachers acted as referrer in almost three

614 Clinical Child Psychology and Psychiatry 18(4)
quarters of the cases and in the recent migrants they identified the problems in about half the cases.
The homeless family service was also significant in identifying problems with over 20% of refer-
rals coming from that agency. Health and social services accessed the project to a lesser extent even
though the mental health professionals worked closely with them. Overall referral rates by the dif-
ferent sectors did not differ significantly between recent migrants and those in the country for
longer. Referrers often named several problems that led to the referral (see Table 2).
These were predominantly problems with concentration, emotions, peer interaction, conduct
and learning. While there was no significant difference in the level of problems with concentration,
emotions and learning between the recent and more settled refugee children, there were significant
differences reported in difficulties with conduct problems and peer relationships. Recently-arrived
refugees had more peer problems than those who had been in the country for more than two years
(χ2 = 3.844, 1 df, p = 0.050). Interestingly, conduct problems were significantly more often reported
as the referred problem in those who had been in the country longer than two years (χ2 = 4.634, 1
df, p = 0.031). In the therapists’ judgement of the main problem, this difference in externalizing
behaviours between recent (6/42 [14%]) and more settled migrants (12/41 [29%]) was not signifi-
cant. There was a trend towards more internalising problems in recent migrants (21/42 [50%]
compared to 13/41 [32%]) (χ2 = 3.567, 1 df, p = 0.059).
At referral, 48 SDQs were received from teachers, 27 from parents and 5 from the young people
themselves. We analysed the teachers’ and parents’ questionnaires further and found that mean
total problem scores were in the ‘borderline’ area. This means that the scores lie between the 80th
and 90th percentile of the UK population. 40/48 (83.3%) of teachers’ SDQ scores and 16/27
(59.3%) of parents’ SDQ scores were in the abnormal range for total problem score, indicating high
risk of psychiatric disorder. Teachers’ and parents’ mean SDQ scores for the subscales of hyperac-
tivity, peer problem and conduct problem were also in the abnormal range. A comparison of the
teachers’ and parents’ mean SDQ scores between the study’s young refugees sample and a national
study representative of Great Britain as a whole1 showed that young refugees have higher scores in
total problem and all subscales scores than the British overall scores (teachers’ mean SDQ scores:
total problem [15.4 vs. 6.6], hyperactivity [6.4 vs. 2.9], peer problem [2.7 vs. 1.4], emotional prob-
lem [3.4 vs. 1.4], conduct problem [2.8 vs. 0.9], and parents’ mean SDQ scores: total problem [13.6
vs. 8.4], hyperactivity [5.3 vs. 3.5], peer problem [2.3 vs. 1.5], emotional problem [3.3 vs. 1.9],
conduct problem [2.7 vs. 1.6]) (Meltzer, Gatward, Goodman & Ford, 2000; Youth in mind, 2001)
(see Table 4). Independent t-test showed no significant difference in SDQ scores between the
recent and more settled refugees.
Linear regression showed that the risk factors that significantly contributed to predicting higher
teachers’ SDQ scores at referral were: current and past family discord (divorce, verbal and physical
violence were taken into account) (R = 0.372, p = 0.043 and R = 0.432, p = 0.017 respectively), and
current or past abuse to the child (R = 0.602, p = 0.000). The following risk factors were also inves-
tigated and found not to be predictors of teachers’ SDQ: mental health of the carer, separation from
immediate family, exposure to violence and war, and having had a family member killed.
Service utilization and treatment outcomes
In order to investigate service use and engagement with services we looked at the number of
sessions attended, type of treatment and type of discharge (that is, negotiated or unplanned) (see
Table 5).
In terms of attendance, it has to be borne in mind that not all patients were offered face-to-face
contact as the project was set up to provide consultations as well as direct work. Overall, 18/92

Durà-Vilà et al. 615
(20%) youngsters had no face to face contact; those who have been in the country for two years or
less were more likely to have no face to face contact (6/39 ([15%]) than those who have been for
more than two years for whom it was nonexistent (0/37 [0%]). Both the group that had arrived less
than two years ago and the group that had arrived more than two years ago engaged equally well
in treatment. Approximately one third of each group attended for consultation or assessment only
(0–2 sessions), around one third attended for brief interventions (3–9 sessions), while a further
third engaged in longer term work (10 or more sessions). No statistically significant differences
were found between the groups.
Treatment type varied significantly between the recent and more settled refugee children (χ2 =
9.176, 3 df, p = 0.027). Treatments offered were eclectic and based on individual and family needs.
They included liaison with other agencies, problem solving and practical help as well as direct
therapeutic work with individuals, groups of young people and families. The statistically signifi-
cant difference amongst the treatments was found in individual treatment, which occurred more
amongst those settled in the country for more than two years (10/42 [24%] compared with 3/43
[7%] for the more recent refugees) (F = 0.030, p = 0.038). A trend was observed in the ‘no face to
face” category: youngsters staying in the country for two years or less had 5/43 (5%) while being
nonexistent for those settled in the UK for more than two years 0/42 (0%) (F = 1, p = 0.055). Even
though patients often moved at short notice, and therefore about a third was lost in follow-up, there
was also no significant difference between the groups in frequencies of unplanned discharge.
Treatment outcome
For 48 patients we received therapists’ ratings as to whether the patient showed overall improve-
ment or remained the same/worse (see Table 5). For the third of patients lost to follow-up no out-
come data is available. In those cases where outcome data is available all ratings showed that the
young people benefited significantly from treatment. Therapists thought that overall, three quarters
of the patients had improved while 25% remained the same or got worse. According to the thera-
pists’ report, in the group of recent arrivals only 67% improved while over 90% of the group that
had been in Britain for more than two years improved (statistical trend for difference, F = 0.993, p
= 0.064).
Data is available using the SDQ for pre- and post-treatment from teachers in 24 cases and from
parents in 11 cases (only one child had SDQ reported by both teacher and parent). As the number
of those who had filled out pre- and post-treatment SDQs was low (n = 24 for teachers and n = 11
for parents), we did not split this into two groups according to duration of settlement (see Table 6).
Table 6 shows that according to teachers the young people showed significant improvement
with regard to Total Problem Scores (t = 2.787, 23 df, p = 0.010), Hyperactivity Scores (t = 2.635,
23 df, p = 0.015) and Peer Problem Scores (t = 2.551, 25 df, p = 0.017). According to parents, there
were significant improvements in Total Problem Score (t = 3.425, 10 df, p = 0.006), Hyperactivity
Score (t = 5.244, 10 df, p = 0.000) and Conduct Problem Score (t = 2.319, 10 df, p = 0.043).
Discussion
There are a number of important findings of this investigation of study of refugees referred to
a child mental health service. Firstly, past adversities were significantly higher for the young
refugees who had been in the UK for less than two years in the following areas studied: separa-
tion from immediate family (due to political reasons), direct exposure to violence and personal
experience of war. Especially striking was personal experience of war, with recently arrived

616 Clinical Child Psychology and Psychiatry 18(4)
refugee children being approximately four times more likely to report this than refugees who
had been settled longer. Secondly, although they continued to experience serious adversities
once in the UK, they adapted well in some areas: those who had been in the UK longer spoke
English well, had social networks and supportive friends, but nevertheless, no significant dif-
ference in family network was found between the two groups. Thirdly, the study indicates
clearly that the young refugees had significant psychopathology (SDQ scores were strikingly
high). It was found that those refugees settled longer were significantly more likely to be
referred because of conduct problems while there was a trend in recent arrivals to present with
internalizing pathology. Fourthly, current and past family discord significantly contributed to
predicting higher teachers’ SDQ scores at referral. Finally, community-based mental health
services for young refugees appeared effective as significant improvement was found in teach-
ers’ SDQ scores (total, hyperactivity/inattention, peer problem scores) and in parents’ SDQ
scores (total, hyperactivity/inattention and conduct problem). The only significant differences
observed in treatment modalities were that those who have been in the country more than two
years were more likely to receive individual treatment and those more recently arrived were
more likely to have had no face to face contact with services. Frequencies of unplanned dis-
charge did not differ significantly in the two groups.
Comparing recently-arrived refugees and those settled longer than two years
The majority of refugees report considerable past experience of war, persecution and vio-
lence, but this was more frequently experienced in the recently arrived group. There are
three likely explanations. The first is related to the progressive tightening of UK immigra-
tion and asylum laws; it might be that in recent times only those who experienced more
extreme forms of violence were able to enter and remain in the UK, while previously asylum
seekers where more likely to remain following experience of a lesser degree of war, violence
or persecution. A second explanation is that there is a memory effect and recent experiences
are much more vivid in people’s minds so would be more likely to be reported. More settled
refugees who are trying to forget their traumas and adapt to their new country may be less
likely to remember (and/or report) the details of aversive experiences. A final explanation is
that reports are linked to the refugee status; less than a quarter of the recently-arrived refu-
gees had secure status (full status or exceptional leave to remain). They might have been
accustomed to emphasizing past trauma to interviewers such as immigration officials and
also health professionals.
In some ways, young refugees adapted well to their new environment. Most of those in the
country for more than two years spoke English well and many of their families had built up social
networks of supportive friends. A significant number of those settled for more than two years had
gained a relatively secure refugee status, although this was not the case for all families. Nevertheless,
the difference in the housing situation between those settled longer was not statistically significant
with over half of refugee families with children who had been in Britain longer than two years (in
some cases considerably longer) remained in highly unsuitable overcrowded hotel and ‘bed and
breakfast” accommodation.
Somewhat surprisingly, there was no significant difference in family networks between the two
groups and almost half of the recent arrivals already had a supportive family network in England.
It might be that refugees choose their country of refuge according to the existence of family net-
works or that wider family groups flee together. It might also explain why many families travel
through relatively safe countries in continental Europe in order to seek asylum in Britain.

Durà-Vilà et al. 617
Much of the literature on young refugees’ mental health looks is concerned with PTSD in
their original communities, during flight or early resettlement (Lustig et al., 2004).
Interestingly, in this sample, PTSD symptoms were rarely identified as the main problem by
teachers. As the project did not include psychiatric diagnostic assessments, PTSD symptoms
may have been underreported. It is known that anxiety disorders including PTSD may be
associated with relatively low levels of psychosocial impairment, and so may not be as ‘vis-
ible’ to teachers in classroom settings, compared with disruptive behaviour disorders
(Goodman, 2001; Tousignant et al., 1999). The teachers’ main concerns were disruptive
behaviours and poor learning. With the passage of time since war exposure, and time to settle
in the new communities and schools, refugee children may be expected to have reduced anxi-
ety symptoms including posttraumatic stress symptoms (Almquist & Broberg, 1999; Sack et
al., 1999). However, settlement stressors, such as poor parental mental health, unemployment
and economic hardship, and overcrowding, may contribute to the onset of disruptive behav-
iour. For this reason, conduct problems may be more frequent in those referred to the service
if they have been in the UK for a longer time. This is consistent with the high level of conduct
disorder found in the community study of refugee adolescents living in Quebec (Tousignant
et al., 1999). Conduct disorder occurred at a higher level of prevalence than amongst the
comparison families, and was associated with high levels of psychosocial impairment. The
Quebec study highlighted the effects of socioeconomic stressors such as paternal unemploy-
ment and poorly functioning family organization.
A further finding from this London service is the strikingly high level of hyperactivity/inatten-
tion symptoms, consistent with the pilot study (O’Shea et al., 2000). Hyperactivity/inattention
symptoms would be associated with disruptive behaviour and also difficulties in learning. These
problems would be observed by teachers and result in selective referral to the service. However,
there may be specific reasons why refugee children may have high levels of hyperactivity/inatten-
tion symptoms. These may arise because of earlier disrupted relationships, stress associated with
migration, and inattention associated with hyper arousal that occurs as an aspect of PTSD. For
example, it has been shown that children exposed to war events may have high levels of hyperac-
tivity/inattention as well as emotional symptoms (Qouta, Punamäki, & El Sarraj, 2005). The rela-
tive persistence of the hyperactivity/inattention symptoms, and high levels at the end of intervention,
suggest that some of the background factors may persist. An alternative explanation is that the
referred refugee children had a predisposition towards hyperkinetic disorder, their problems were
exacerbated by migration and refugee experiences, and that they need further disorder specific
interventions for these problems (NICE, 2009).
Implications for service delivery
The young refugees had considerable mental health needs. However their parents may have little
knowledge of their host countries’ health systems. Access to services is often difficult and most
services do not have dedicated community-based child mental health professionals for refugee
work. Conventional Child and Adolescent Mental Health Services (CAMHS) receive most of their
referrals through primary health and social services, and some from education services (Howard &
Hodes, 2000). In this project we showed that education and a voluntary service are helpful in iden-
tifying the mental health needs of young refugees. Teachers may play an important role in identify-
ing psychological distress and impairment.
Evidence for benefit of the service came from the size of reduction of the SDQ. It was
greater than that found in some other reports of school-based services, which had also used the

618 Clinical Child Psychology and Psychiatry 18(4)
same instrument. However, there were significant differences in approach between these
reports, including recruitment. Some recruited all immigrant or refugee children in classrooms
(Fazel et al., 2009; Rousseau et al., 2007), and some investigated time-limited single therapeu-
tic modalities (Rousseau et al., 2007). In the London service, children were selected on the
basis of teacher observation of distress and impairment indicating high risk of psychiatric dis-
order (Hodes, 2002a). It is expected that higher initial SDQ scores will be associated with
greater reduction for those who are offered intervention, or spontaneous changes over time, and
with regression to the mean.
The heterogeneity of the difficulties of the refugee children mean that refugee mental health
services need to be able to offer a broad range of interventions, so that children with disruptive
behaviour disorders as well as anxiety symptoms and posttraumatic stress disorder can be treated
(Hodes, 2008; Hodes & Diaz-Caneja, 2007). This is in keeping with recent recommendations that
refugee mental health services should be able to provide care for people with disorders that might
occur in the absence of the experience of war trauma and displacement, as well as those who have
a high level of psychological distress and impairment resulting from those experiences (Inter-
Agency Standing Committee, 2007).
The reduction of psychological distress using the SDQ as an outcome measure and the varied
problems and approaches offered are illustrated in the following two vignettes (derived from chil-
dren seen in the service, and pseudonyms used to preserve anonymity):
Vignette 1
Arta is a 9-year-old girl from Kosovo, the eldest of two children, living with her divorced mother.
At the time of referral, the teacher’s main concerns about Arta were her anxiety about her mother,
unsettled friendships and inability to concentrate and organize her schoolwork. Arta and her mother
had arrived in the UK three years previously and were granted exceptional leave to remain for three
years. Her parents had separated prior to their departure from Kosovo. Her maternal grandparents
were killed in the war, but the family had no experiences of direct violence although they had lived
in a war zone. They had links with other refugee families from the former Yugoslavia and sufficient
competence in English to be seen without an interpreter. The main complaint of Arta’s mother was
that she was a timid, anxious girl who refused to travel in lifts alone. They were housed in a two-
bedroom council flat but this was on the 18th floor of a council block. Arta only went in the lift
with her mother, or alternatively walked up the stairs if unaccompanied to avoid anxiety symp-
toms. Arta complained about verbal bullying from her peer group and lack of confidence with
schoolwork. Arta had been seen in the local CAMHS, but she and her mother did not return after
the assessment. A diagnosis was made of specific phobia of lift travel.
The parent and child were seen over the course of a term, and work centred on the development
of social skills with peers and exposure to lift travel. Her schoolteachers commented on the
improvement in Arta’s attention with gradual settling of her friendships. During subsequent home
visits she was able to use the lift in a behavioural exposure programme. Her mother, then the thera-
pist accompanied her and finally she managed to use the lift on her own successfully. Arta’s SDQ
score decreased from 21 pre-treatment to 10 at the termination of treatment.
Vignette 2
Haydar is a 9-year-old boy, a refugee from the Kurdish area of Iraq. His mother and older brother
were killed by the regime in Iraq. The family had come to the UK three years previously, and

Durà-Vilà et al. 619
Haydar lived with his paternal grandmother and an aunt in hotel accommodation. They had full
refugee status on humanitarian grounds. An Arabic-speaking interpreter was present for interviews
with adults in the family.
Haydar was born in Iraq but never attended school there as his family was concerned that this
would not be safe. The family experienced army house searches and the arrest, imprisonment and
death of family members. In the UK, his father had remarried another refugee from Iraq, but was
receiving treatment from a community mental health team for anxiety and depression. Haydar
chose to live with his grandmother due to a poor relationship with his stepmother and her 5-year-
old daughter. During the two years that Haydar had been at the school, he made slow academic
progress, and was thought to be under-achieving. He was reported to have little idea of letter
sounds, reading difficulties and was unable to write independently. Haydar’s grandmother described
him as suffering from nightmares and being irritable and moody. Haydar confirmed that he expe-
rienced bad dreams but was unable to remember them. He felt he could not discuss missing his
mother and sister with his grandmother for fear of upsetting her.
A diagnosis was made of an adjustment reaction. Psychological treatment involved grief work
(cognitive-behaviour therapy with exposure to loss) in individual sessions and family sessions with
Haydar’s grandmother and aunt (Walter & Bala, 2004). Photographs were used to discuss issues
around losses, both of people, lifestyle and culture. All the family experienced symptoms of depres-
sion, and after further meetings his aunt was referred to the adult psychological trauma service for
assessment. The family’s application for permanent housing was supported. Haydar’s academic
work improved significantly as his low mood lifted. Haydar’s SDQ score decreased from 21 pre-
treatment to 9 at discharge.
It is sometimes suggested that child mental health provision requires children and families be
settled in order to benefit from intervention. This group of patients was highly mobile and indeed
about one-third were lost to follow-up. Overall, however, patients did engage well with the child
mental health professionals who were placed conveniently close to them in schools and voluntary
centres where the refugees were likely to spend time. The community mental health project was
able to reach a relatively large number of young refugees with significant mental health needs, who
might otherwise not have been able to access services.
Guidelines for practitioners
In areas with high refugee populations, CAMHS should consider building up links with education
and voluntary services in order to create community-based services (Hodes, 2010). Our work has
found that the majority of children regarding whom teachers had concerns could be managed by
consultation with teachers and other professionals, and direct work in community settings.
A pragmatic approach to working in schools is needed given the range of clinical problems,
varied family support and involvement, and need to take into account the child’s age and develop-
ment. A combination of individual and family work is usually needed. In some settings, groups
may be helpful, but there are a number of considerations here in view of linguistic and cultural
diversity, including potential conflicts arising for political reasons. Providing group interventions
with more than one interpreter becomes impractical. Clearly, for all settings and interventions,
access to good interpreters is necessary and acquiring the skills to work with them is important
(Farooq & Fear, 2003; Raval & Tribe, 2002).
A further task for community-based child mental health practitioners is to facilitate appropriate
referral to tier-3 specialist CAMHS. This will be required for children who have more complex
problems including self harm risk. It is important to bear in mind that children from refugee

620 Clinical Child Psychology and Psychiatry 18(4)
backgrounds may have difficulties unrelated to their asylum or refugee status (Howard & Hodes,
2000). They may have problems and disorders including ADHD and autistic spectrum problems
that require the skills of the multidisciplinary team. Furthermore, it should not be assumed that for
these more impaired and distressed children, appropriate referral will be associated with poor
engagement or dropout. We have found that dropout was not higher amongst the refugee children
than it was for a white British group attending the same CAMHS in Westminster, despite the refu-
gee families being more socially isolated and having greater need for interpreters (Howard &
Hodes, 2000).
Limitations of the study and future research
The collected data sets were not complete. Only basic information was available for those patients
for whom our main intervention was consultation and liaison with the referrers. Background data
was often disclosed over a number of consultations, first with professionals and then with the fami-
lies. Furthermore, the population was highly mobile and sometimes families moved or stopped
attending, so there was considerable sample loss. The study was weakened by the absence of a
greater number of standardized instruments, and absence of independent investigator evaluation the
interventions. Nevertheless, this is the largest sample of refugee children in mental health services
described in the UK, and there is a dearth of similar studies from other countries.
Conclusions
The provision of refugee mental health services in schools and consultation with teachers and oth-
ers had many advantages. Further work involving the ongoing training of teachers in the detection
of children psychological distress would be important. The extent to which school-based mental
services would be able to meet the needs of refugee children remains unclear. Many children with
anxiety disorders and PTSD may not be detected by teachers and so alternative referral pathways,
including primary care, to specialist child mental health services are still appropriate. Further
investigation of the relative merits of primary care or education based services, or combined child
and adult mental health services is warranted.
Acknowledgements
Thanks are due to many teachers and other colleagues who contributed to the service.
Funding
This work is funded by Action for Peoples in Conflict and Kensington Chelsea and Westminster Health
Authority.
Note
1. National averages come from a national study with 8,208 teachers and 10,298 parents of children aged
5–15 years, representative of Great Britain as a whole (Meltzer et al., 2000; Youth in mind, 2001).
References
Almqvist, K., & Broberg, A. G. (1999). Mental health and social adjustment in young refugee children three
and a half years after their arrival in Sweden. Journal of the American Academy of Child and Adolescent
Psychiatry, 38(6), 723–730.

Durà-Vilà et al. 621
Becker, D. F., Weine, S. M., Vojvoda, D., & McGlashan, T. H. (1999). Case series: PTSD symptoms in ado-
lescent survivors of “ethnic cleansing”. Results from a 1-year follow-up study. Journal of the American
Academy of Child and Adolescent Psychiatry, 38(6), 775–781.
Beiser, M., Dion, R., Gotowiec, A., Hyman, I., & Vu, N. (1995). Immigrant and refugee children in Canada.
Canadian Journal of Psychiatry, 40(2), 67–72.
Bourdon, K., Goodman, R., Rae, D., Simpson, G., & Koretz, D. (2005). The Strengths and Difficulties Ques-
tionnaire: US normative data and psychometric properties. Journal of the American Academy of Child and
Adolescent Psychiatry, 44(6), 557–564.
Dyregrov, A., Gjestad, R., & Raundalen, M. (2002). Children exposed to warfare: A longitudinal study. Jour-
nal of Traumatic Stress, 15(1), 59–68.
Espino, C. M. (1991). Trauma and adaptation: The case of Central American children. In F. L. Ahearn
& J. L. Athey (Eds.), Refugee children: theory, research and services. The Johns Hopkins series in
contemporary medicine and public health (pp. 106–124). Baltimore, MD: The Johns Hopkins Uni-
versity Press.
Farooq, S., & Fear, C. (2003) Working through interpreters. Advances in Psychiatric Treatment, 9, 104–109.
Fazel, M., Doll, H., & Stein, A. (2009) A school-based mental health intervention for refugee children: An
exploratory study. Clinical Child Psychology and Psychiatry, 14(2), 297–309.
Fazel, M., Wheeler, J., & Danesh, J. (2005). Prevalence of serious mental disorder in 7000 refugees resettled
in western countries: A systematic review. The Lancet, 365(9467), 1309–1314.
Garbarino, J., & Kostelny, K. (1996). The effects of political violence on Palestinian children’s behaviour
problems: A risk accumulation problem. Child Development, 67(1), 33–45.
Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A research note. Journal of Child Psy-
chology and Psychiatry, 38(5), 581–586.
Goodman, R. (2001). Psychometric properties of the Strengths and Difficulties Questionnaire (SDQ). Journal
of the American Academy of Child and Adolescent Psychiatry, 40(11), 1337–1345.
Goodman, R., Renfrew, D., & Mullick, M. (2000). Predicting type of psychiatric disorder from Strengths and
Difficulties Questionnaire (SDQ) scores in child mental health clinics in London and Dhaka. European
Child and Adolescent Psychiatry, 9(2), 129–134.
Hauff, E., & Vaglum, P. (1997). Establishing social contact in exile: A prospective community cohort
study of Vietnamese refugees in Norway. Social Psychiatry and Psychiatric Epidemiology, 32(7),
408–415.
Health Advisory Service. (1995). Child and adolescent mental health services. Together we stand. London,
UK: HMSO.
Hodes, M. (2002a). Implications for psychiatric services of chronic civilian strife: Young refugees in the UK.
Advances in Psychiatric Treatment, 8(5), 366–376.
Hodes, M. (2002b). Three key issues for young refugees’ mental health. Transcultural Psychiatry, 39(2),
196–213.
Hodes, M. (2008). Psychopathology in refugee and asylum seeking children. In M. Rutter, D. Bishop, D. Pine,
S. Scott, J. Stevenson, E. Taylor & A. Thapar (Eds.), Rutter’s Child and Adolescent Psychiatry (5th ed.)
(pp. 474–486). Oxford, UK: Blackwell.
Hodes, M. (2010) CAMHS for refugees and recent immigrants. In G. Richardson, I. Partridge & J. Barrett
(Eds.), Child and adolescent mental health services. An operational handbook (2nd ed.) (pp. 214–225).
London, UK: Gaskell.
Hodes, M., & Diaz-Caneja, A. (2007). Treatment options for young people and refugees with posttraumatic
stress disorder. In A. Hosin (Ed.), Children, families and refugees of multiple traumas: contemporary
issues in mental health (pp. 40–65). Basingstoke, UK: Palgrave.
Home Office. (2005). Asylum statistics. United Kingdom 2004. 13/05. Retrieved from London Home Office:
http://www.homeoffice.gov.uk/rds/pdfs05/hosb1305.pdf
Home Office. (2007). Asylum statistics United Kingdom 2006 (3rd ed.). London, UK: ONS.
Home Office. (2010). Control of immigration. Statistics United Kingdom 2009. London: The Home Office.
Retrieved from http://rds.homeoffice.gov.uk/rds/pdfs10/hosb1510.pdf

622 Clinical Child Psychology and Psychiatry 18(4)
Howard, M., & Hodes, M. (2000). Psychopathology, adversity, and service utilization of young refugees.
Journal of the American Academy of Child and Adolescent Psychiatry, 39(3), 368–377.
Inter-Agency Standing Committee (IASC). (2007). IASC guidelines on mental health and psychosocial sup-
port in emergency settings Retrieved from http://www.who.int/mental_health/emergencies/guidelines_
iasc_mental_health_psychosocial_june_2007.pdf.
Kushner, T., & Knox, K. (1999). Refugees in an age of genocide: global, national and local perspectives dur-
ing the twentieth century. London, UK: Frank Cass.
Laban, C. J., Gernaat, H. B. P. E., Komproe, I. H., van der Tweel, I., & De Jong, J. T. V. M. (2005). Postmi-
gration living problems and common psychiatric disorder in Iraqi asylum seekers in the Netherlands. The
Journal of Nervous and Mental Disease, 193(12), 825–832.
Lustig, S. L., Kia-Keating, M., Knight, W. G., Geltman, P., Ellis, H., Kinzie, J. D., & Saxe, G. N. (2004).
Review of child and adolescent refugee mental health. Journal of the American Academy of Child and
Adolescent Psychiatry, 43(1), 24–36.
Meltzer, H., Gatward, R., Goodman, R., & Ford, F. (2000). Mental health of children and adolescents in
Great Britain. London, UK: The Stationery Office.
NICE. (2009). Diagnosis and management of ADHD in children, young people and adults. National Clinical
Practice Guideline Number 72. National Collaborating Centre for Mental Health commissioned by the
National Institute for Health & Clinical Excellence. Leicester: British Psychological Society & London:
Royal College College of Psychiatrists.
O’Shea, B., Hodes, M., Down, G., & Bramley, J. (2000). A school-based mental health service for refugee
children. Clinical Child Psychology and Psychiatry, 5(2), 189–201.
Pedersen, D. (2002) Political violence, ethnic conflict, and contemporary wars: Broad implications for health
and social well-being. Social Science and Medicine, 55(2), 175–90.
Qouta, S., Punamäki, R. L., & El Sarraj, E. (2005). Mother-child expression psychological distress in acute
war trauma. Clinical Child Psychology and Psychiatry, 10(2), 135–56.
Raval, H., & Tribe, R. (2002). Working with interpreters in mental health. London, UK: Brunner-
Routledge.
Rousseau, C., Benoit, M., Gauthier, M. F., Lacroix, L., Alian, N., Rojas, M. V., & Bourassa, D. (2007).
Classroom drama therapy program for immigrant and refugee adolescents: A pilot study. Clinical Child
Psychology and Psychiatry, 12(3), 451–65.
Rousseau, C., & Guzder, J. (2008). School-based prevention programs for refugee children. Child and Ado-
lescent Psychiatric Clinics of North America, 17(3), 533–49.
Sack, W. H., Him, C., & Dickason, D. (1999). Twelve-year follow-up study of Khmer youths who suffered
massive war trauma as children. Journal of the American Academy of Child and Adolescent Psychiatry,
38(9), 1173–1179.
Statistical Package for Social Sciences (SPSS). (2003). Base System User’s Guide, SPSS. Chicago, IL: SPSS
Inc.
Tousignant, M., Habimana, E., Biron, C., Malo, C., Sidoli-LeBlanc, E., & Bendris, N. (1999). The Quebec
adolescent refugee project: Psychopathology and family variables in a sample from 35 nations. Journal
of the American Academy of Child and Adolescent Psychiatry, 38(11), 1426–1432.
United Nations High Commissioner for Refugees (UNHCR). (2006). The state of the world’s refugees.
Human displacement in the new Millennium. Oxford, UK: Oxford University Press.
Walter, J., & Bala, J. (2004). Where meanings, sorrow, and hope have a resident permit: Treatment of fami-
lies and children. In J. P. Wilson & B. Drozdek (Eds.), Broken spirits. The treatment of traumatized asy-
lum seekers, refugees, war and torture victims (pp. 487–519). New York & Hove: Brunner-Routledge.
Weine, S., Becker, D. F., McGlashan, T. H., Vojvoda, D., Hartman, S., & Robbins, P. (1995). Adolescent sur-
vivors of ‘ethnic cleansing’: Observations on the first year in America. Journal of the American Academy
of Child and Adolescent Psychiatry, 34(9), 1153–1159.
Winder, R. (2004). Bloody foreigners. The story of immigration to Britain. London, UK: Abacus.
Youth in mind. (2001). Strength and Difficulties Questionnaires (SDQ): Normative SDQ data from Britain.
Retrieved from: http://www.sdqinfo.com/bb1.html

Durà-Vilà et al. 623
Author biographies
Glòria Durà-Vilà, MRCPsych MSc, is a consultant Child and Adolescent psychiatrist in Surrey and Borders
NHS Foundation Trust, and an Honorary Lecturer at the Department of Mental Health Sciences, University
College London. She was previously Honorary Clinical Lecturer at Imperial College London. Her research
interests focus on medical anthropology and cultural psychiatry including religion and psychiatry, idioms and
narratives of distress, intellectual disability and personal identity theory.
Henrika Klasen, MRCPsych MSc, is a Consultant Child and Adolescent Psychiatrist, De Jutters Centre for
Youth Mental Health, The Hague, and Associate Professor Curium, Leiden University Medical Centre, The
Netherlands. Her research interests include medical anthropology, access to care for migrants and global
mental health.
Zethu Makatini was previously a family therapist in the young refugees’ mental health service, St Mary’s
Child & Adolescent Mental Health Service, CNWL NHS Foundation Trust, Paddington Green, London, and
is currently a Senior Family Therapist in the East London NHS Foundation Trust.
Zohreh Rahimi was previously a family therapist in the young refugees’ mental health service, St Mary’s
Child & Adolescent Mental Health Service, CNWL NHS Foundation Trust, Paddington Green, London, and
is now a Senior Family Therapist in the Medical Foundation for the Care of Victims of Torture, London.
Matthew Hodes, MBBS BSc MSc PhD FRCPsych, Senior Lecturer in Child and Adolescent Psychiatry,
Imperial College London, has carried out research in the interface between physical and mental health, and
cultural psychiatry including psychopathology and health service utilisation amongst young refugees. He is
a Consultant in Child and Adolescent Psychiatry at St Mary’s Child & Adolescent Mental Health Service,
CNWL NHS Foundation Trust, Paddington Green, London.