Effectiveness of family support for children in homeless families

The whole issue of parents and children in need is a vast, complex and ethically challenging one. This review is specifically charged with an examination of those issues which impinge upon the stresses and strains that are experienced by parents of children in need.

A superficial examination of these issues that are involved in this particular area would suggest that there are a number of “sub-texts “which can all give rise to this particular situation. Firstly, to have child in need is clearly a stressful situation for any parent.(Meltzer H et al. 1999)

This can clearly be purely a financial concern and a reflection of the fact that the whole family is in financial hardship, perhaps due to the economic situation or perhaps due to the actions of the parents themselves. Equally the need of the child can be a result of anon-financial need, so we should also consider the child who is in some way handicapped, ill, emotionally disturbed or perhaps in need in some other way. This produces another type of stress on the parent, and these stresses are typically longer lasting and, in general, less easily rectified than a purely financial consideration of need. (Hall D1996).

It is part of the basic ethos of the welfare state that it should look after its less able and disadvantaged members. (Welsh Office 1997).Parents of children in need will often qualify in this definition. We shall therefore examine the various aspects of this problem.

Literature Review

Effectiveness of family support for children in homeless families

We will make a start by considering one type of child in need. The first paper that we will consider is that of Prof. Vostanis (Vostanis2002), which looks at the mental health problems that are faced by deprived children and their families together with the effectiveness of the resources that are available to them.

It is a well written and well researched paper, if rather complex and confusing in places. We will consider this paper in some detail as it provides an excellent overview of the whole area.

The paper starts with a rather useful definition for our purposes. It qualifies the deprived child, initially in terms of a homeless family, that being :

A family of any number of adults with dependent children who are statutorily accepted by local authorities (housing departments) in teak, and are usually accommodated for a brief period in voluntary agency, local authority or housing association hostels.

This period of temporary accommodation can vary enormously depending on the time of year and the area considered, and can range from a few days to perhaps several months. The target in Greater London is currently storehouse homeless families within 4-6 weeks. In London particularly, the homeless families can be placed in Bed & Breakfast accommodation.(D of H 1998)

In this respect, the immediate family support mechanisms do appear tube in place. Vostranis however, goes on to make the observation that despite the fact that the definition of the homeless family is rather broad, it does not cover all of the potential children in need, as those children and their carers who have lost their homes but have managed to live with relatives, on the streets or perhaps live as travellers, are not covered by the statutory obligation to provide housing. The official figures therefore, he observes, are generally an underestimate of the true situation. The official figures for the homeless families are put (in this paper) at 140,000. (Vostanis &Cumella, 1999)

The authors give us further information in that many families will become homeless again within one year of rehousing and the typical family seen is the single mother and at least two children who are generally under the age of 11 yrs. They also observe that the typical father and adolescent child tend to be placed in homeless centres. (Doff H 1995)

In exploration of the particular topic that we are considering, the authors give us the situations that typically have given rise to the degree of parental stress that may have led to the homelessness. They point to the fact that a homeless family is usually homeless for different reasons to the single homeless adult. Vostanis (et al 1997)is quoted as showing that 50% of the cases studied were homeless as direct result of domestic violence and 25% as a result of harassment from neighbours. The authors observe that the numbers in this category(and therefore the problems), are rising. (Welsh Office 1999).

There are a number of section to this paper which are not directly referable to our considerations. We shall therefore direct our attention purely to those parts that have a direct bearing on the subject. One particularly useful and analytical part of the paper is the section that details the characteristics and needs of the target group. This is a very detailed section, but it makes the point that the children in need in this group are particularly heterogeneous, generally all with multiple and inter-related needs. Homelessness is seldom a one off event. This particular observation, (say the authors),is crucially important for the development and provision of services.

Most families have histories of previous chronic adversities that constitute risk factors for both children and parents (Bassuk et al,1997). Such events include family conflict, violence and breakdown; limited or absent networks for family and social support; recurring moves; poverty; and unemployment. Mothers are more likely to have suffered abuse in their own childhood and adult life and children have increased rates of placement on the at-risk child protection register, because of neglect, physical and/or sexual abuse.

If we specifically consider the health needs of this population, the authors categorise them thus:

The children are more likely to have a history of low birthweight, anaemia, dental decay and delayed immunisations, to be of lower stature and have a greater degree of nutritional stress. They are also more likely to suffer accidents, injuries and burns. (BPA 1999)

Some studies have found that child health problems increase with the duration of homelessness, although this finding is not consistent. Substantial proportion of homeless children have delayed development compared with the general population of children of a similar chronological age. This includes both specific developmental delays, such as in receptive and expressive language and visual, motor and reading skills, as well as general skills and educational status (Webbet al. 2001).

It is for this reason specifically, that it has proved extremely difficult to assess the effectiveness of the family support services because of the multivariate nature of the problems that are presented.

The authors point to the fact that one of the prime determinants of the degree of support available, is the actual access that the families have to these services. Many sources (viz. Wilkinson R 1996), equate the poor health of the disadvantaged primarily with the lack of access to services. One immediate difficulty is the current registration system in the UK. In order to be seen in the primary healthcare team setting, one must be registered with a named doctor. In the majority of cases that we are dealing with here, they have moved area and registration is probably not high on their list of priorities. One can argue that there is the access to the A & E departments of the local hospitals but there is virtually no continuity here and they arena geared up to provide anything other than immediate treatment. (HallD 1996).

This fact restricts their access to primary healthcare team procedures such as immunisations and other preventative medicine health clinics.(Lissauer et al, 1993) . By the same token these groups also have restricted access to the social services, whether they be the access teams, the family teams or the family support units and other agencies.

The authors also point to other more disruptive trends in this group such as an inability to attend a particular school for fear of being tracked by an abusive partner. It follows that these children do not have a stable social support of a school. They are denied such factors as peer groups, routines and challenges which are both important protective and developmental factors. (Shankleman J et al2000).

The summation of all of these factors, and others, is that the effectiveness of the family support services is greatly reduced by the mobility and the transient nature of the family unit. Quite apart from the difficulties outlined above relating to the problems of access to avenues of help open to the child in need and their families there are the problems engendered by the fact that social service departments indifferent areas of the country may not have immediate access to the previous records giving rise to many potential, and real, problems with continuity of care. This problem is brought into more immediate focus when one considers the increased frequency of child protection registrations in this particular group. (Hall D et al 1998).

One specific analysis of the family support services of this particular group comes in the form of the psychiatric services. In the context of the title of this piece, it demonstrates how these particular services,(but not these alone), are failing to deal with the totality of the problem.

All of the aspects that we have outlined so far are conspiring to dilute the effectiveness of the services provided. The fact that they are a mobile population with no fixed address means that some of the services may choose to invoke this as a reason for not making provision for them, particularly if resources are stretched. If more resources are given, then they are typically preferentially targeted at the single adult homeless population where the need is arguably greater.

The authors of this paper point to the fact that this may not actually be true as some studies have shown that homeless single mothers and their children have a 49% psychopathy rate and only an 11% contact with the support services. (Cumella et al, 1998). The impact of this fact on the children can only be imagined. To an extent however, it can be quantified as the authors cite other studies which show a 30% need rating for children, (they do not actually define exactly what their perceived level of need was), contrasted with a 3% contact rate for children and adolescents in this area. (viz. Power S et al. 1995).

Suggestions for improvement

Putting these considerations together, the authors outline a set of proposals which are designed to help improve the access to some of the essential services. The model that they propose could, if successful and with a degree of modification, prove suitable for adaptation to other areas of the family support services. It is not appropriate to discuss this model in detail, but suffice it to say that it has tiered structure so that the degree of distress and need is titrated against the degree of input generated.

One of the reasons that we have selected this particular paper to present in this context is for its last section. It proposes a “family support services model” which has been developed and pioneered in the Leicester area. In the context of our review, it is worth considering in some detail.

A service provided through a family support team (four family support assistants).This is designed to detect a range of problems at the time of crisis; manage a degree of mental health problems (behavioural and emotional); provide parenting-training; support and train housing(hostel) staff; co-ordinate the work of different agencies; and provide some continuity after rehousing by ensuring intake by appropriate local services.

The family workers are based at the main hostel for homeless children and families. Other, predominantly voluntary, services have established alternative posts, such as advocates and key workers. Whatever the title of the post, it is essential that the post-holder has some experience and on-going training in mental health and child protection, so that he or she can hold a substantial case-load, rather than merely mediate between already limited services.

The family support workers have direct access to the local child and adult mental health services, whose staff provide weekly outreach clinics. Their role is to work with the family support workers another agencies, assess selected children and families, and provide treatment for more severe problems or disorders such as depression, self-harm and PTSD. A weekly inter-agency liaison meeting at the main hostel is attended by a health visitor, representatives of the local domestic violence service and Sure Start, There are also close, regular links with education welfare and social services. The aim is to effectively utilise specialist skills by discussing family situations from all perspectives at the liaison meeting.

A bimonthly steering group, led by the housing department, involves senior managers representing these agencies, as well as the education and social services departments and the voluntary sector, and they oversee and co-ordinate the service.

This appears to be something of an exemplar in relation to services provided elsewhere. The paper does not provide any element of costing sin this area neither does it provide any figures in relation to its success rates, contact rates or overall effectiveness. In conclusion this paper is an extremely well written and authoritative overview of the situation relating to the stresses of the homeless parent with children and the effectiveness (or lack of it) in its ability to reduce the stresses experienced by the homeless children in need and their parents. It proposes remedies but sadly it does not evaluate the effectiveness of those remedies.

The effectiveness of the support services on families of children with psychiatric morbidity

In order to address these shortcomings we can consider another paper by Tickler (et al 2000). This looks at a similar outreach set up which has been designed to capture the families of children in need who might otherwise slip through the net. This paper is written from different perspective and specifically analyses the effectiveness of these services as they pertain to an entry cohort of 40 families.

This particular study was set up after preliminary work was done in the Birmingham area with 114 homeless families and this study defined the needs of the families but did not quantify their support systems.(Vostanis et al 1998). This paper set out to identify and measure the support systems available and their effectiveness as far as the families were concerned. The stresses encountered were partly reflected by the incidence of psychiatric morbidity. The mothers in the group were found to have over 50% more morbidity than a matched control group. The children in the group were found to have “histories of abuse, living in care, being on the at-risk protection register, delayed communication and higher reported mental health problems.” Allot which adds to the general background stress levels. (Kerouac S etal. 1996).

This particular study found that despite the psychiatric morbidity in the children, (estimated to be about 30%), and the psychiatric morbidity in the parents, (estimated at about 50%), only 3%of the children and 10% of the parents had had any significant contactor support from the social services. In this respect, this paper is very useful to our purpose as it quantifies the levels of intervention and access to healthcare resources that this particular group has. By any appreciation, it would be considered woefully inadequate in any society that calls itself civilised. In the terms of the title of this piece, the effectiveness of the family support services is minimal.

Suggestions for improvement

Like the last paper discussed, this one also considered how best to tackle the problem, and this one is of much greater value to us, as it specifies a response, or intervention, to the problem in much the same way as the Vostranis 2002 paper did, but it makes the same measurements as it did prior to the intervention, and therefore allows us an insight into the actual effectiveness of the intervention.

The way this particular study worked was to assess the problem (as it has been presented above), devise an intervention strategy and then to measure its effect. This particular study goes to great lengths to actively involve all the appropriate agencies that could help the situation by having a central assessment station that acted as liaison between all of the other resources. In brief, it actively involved liaison with the following:

Education, social services, child protection, local mental health services, voluntary and community organisations to facilitate there-integration of the family into the community, and particularly their engagement with local services following rehousing; and training of staff of homeless centres in the understanding, recognition and management of mental illness in children and parents. This is essential, as hostel staff often work in isolation and have little knowledge of the potential severity and consequences of mental health problems in children.

It was hoped that, by doing this, it would maximise the impact that the limited resources had on reducing the levels of morbidity and stress in the families of the children in need.


The post intervention results were, by any estimate, impressive considering the historical difficulty of working with this particular group (O'Hara M 1995). 40 families (including 122 children) were studied in detail. The paper gives a detailed breakdown of the ethnic and demographic breakdown of the group. By far the biggest group were single mothers and children (72%)

The results showed that the majority of referrals were seen between1-3 times (55%), with a further 22% being seen 4-6 times. It is reflection of the difficulty in engaging this type of family in need that over 25% did not actually keep their appointments despite the obvious potential benefits that could have been utilised. The authors investigated this group further and ascertained that a common reason for nonattendance was the perception that the psychological welfare of the children was not actually the main concern. The families perceived that their primary needs were rehousing and financial stability. Other priorities identified were that physical health was a greater priority than mental health.

The authors also identify another common failing in the social services provision, and that is the general lack of regular contact. They cite the situation where some families cope well initially, apparently glad to have escaped an abusive or violent home situation, but a prolonged stay in a hostel or temporary accommodation may soon precipitate a bout of depression in the parents and behavioural problems in the children of such parents. (Brooks RM et al 1998). They suggest that regular re-visiting of families who have been in temporary accommodation for any significant length of time should be mandatory.

This paper takes a very practical overview by pointing out that workability of the system is, to a large extent, dependent on the goodwill of a number of committed professionals. The authors state that this has to be nurtured and they call for sufficient funding must be given to enable this particular model to be extended to a National level.

Thus far in the review we have considered the effectiveness of the service provision in the support of the families of the children in need in one specific target grouping, those who are stressed by virtue of the fact that they are homeless. We will now consider the literature on a different kind of family stress, and that is when a parent dies. This leaves the children with a considerable amount of potential emotional “baggage” and the surviving parent with an enormous amount of stress. (Webb E 1998).

Effectiveness of support services in the case of parental bereavement

An excellent paper by Downey (et al 1999) tackles this particular problem with both sensitivity and also considerable rigour. It is a long and complex paper, but the overall aims and objectives are clear from the outset.

The structure of the paper is a prospective case study which aims to assess whether the degree of distress suffered by a family during a time of bereavement is in any way linked to the degree of service provision that is utilised.

The base line for this study is set out in its first two paragraphs. Parentally bereaved children and surviving parents showed a greater than predicted level of psychiatric morbidity. Boys had greater levels of demonstrable morbidity than did girls, but bereaved mothers showed more morbidity than did bereaved fathers. Children were more likely to show signs of behavioural disturbance when the surviving parent manifested some kind of psychiatric disorder. (Kranzler EM et al 1990).

The authors point to the fact that their study shows that the service provision is statistically related to a number of (arguably unexpected[Fristad MA et al 1993]) factors namely:

The age of the children and the manner of parental death. Children under 5 years of age were less likely to be offered services than older children even though their parents desired it. Children were significantly more likely to be offered services when the parent had committed suicide or when the death was expected. Children least likely to receive service support were those who were not in touch with services before parental death.

Paradoxically the level of service provision was not found to be statistically significantly related to either the parental wishes or the degree of the psychiatric disturbance in either the parent orchid. (Sanchez L et al 1994) The service provision did have some statistical relationships but that was only found to be the manner of the parental death and the actual age of the child at the time.

The authors therefore are able to identify a mismatch between the perceived need for support and the actual service provision made. Part of that mismatch is found to be due to the inability of the social services and other related agencies to take a dispassionate overview. Elsewhere in the paper the authors suggest that there are other factors that add to this inequality and they include lack of resources and a lack of specificity in identifying children at greatest risk.(Harrington R 1996)

The authors examine other literature to back up their initial precept that bereaved children have greater levels of morbidity. They cite many other papers who have found distress manifesting in the form of “anxiety, depression, withdrawal, sleep disturbance, and aggression.”(Worden JW et al. 1996) and also psychological problems in later life(Harris T et al. 1996).

In terms of study structure, the authors point to methodological problems with other papers in the area including a common failing of either having a standardised measure or no matched control group(Mohammed D et al 2003). They also point to the fact that this is probably the first UK study to investigate the subject using a properly representative sample and certainly the first to investigate whether service provision is actually related to the degree of the problems experienced.

The entry cohort involved nearly 550 families with 94 having children in the target range (2-18). With certain exclusions (such as two families where one parent had murdered the other etc.) and no respondents, the final cohort was reduced to 45 families and one target child was randomly selected from each family.

It has to be noted that the comparatively large number of on-respondents may have introduced a large element of bias, insofar as it is possible that the families most in need of support were those who were most distressed by the death of a family member and these could have been the very ones who chose not to participate. (Morton V et al2003) The authors make no comment on this particular fact.

The authors should be commended for a particularly ingenious control measure for the children. They were matched by asking their schoolteacher to complete an inventory of disturbed behaviour on the next child in the school register after the target child.

A large part of the paper is taken up with methodological issues which ( apart from the comments above) cannot be faulted.


In terms of being children in need, 60% of children were found to have “significant behavioural abnormalities” with 28% having scores above the 95th centile.

In terms of specific service support provision, 82% of parents identified a perceived need for support by virtue of the behaviour of their children. Only 49% of these actually received it in any degree.

Perhaps the most surprising statistic to come out of this study waste fact that of the parents who were offered support 44% were in the group who asked for it and 56% were in the group who didn’t want it.

The levels of support offered were independent of the degree of behavioural disturbance in the child.

Suggestions for improvement

As with the majority of papers that we have either presented here or read in preparation for this review, the authors call for a more rationally targeted approach to the utilisation of limited resources. The study also provides us with a very pertinent comment which many experienced healthcare professionals will empathise with, (Black D1996), and that is:

Practitioners should also be aware that child disturbance may reflect undetected psychological distress in the surviving parent.

While not suggesting that this is a reflection of Munchausen’s syndrome by proxy, the comment is a valid reflection of the fact that parental distress may be well hidden from people outside of the family and may only present as a manifestation of the child’s behaviour. (Feldman Met al. 1994)

The conclusions that can be drawn from this study are that there is considerable gap in the support offered ( quite apart form the effectiveness of that support) in this area of obvious stress for both parents and children. (Black D 1998). This study goes some way to quantifying the level of support actually given in these circumstances.

Effectiveness of support in families where there is domestic violence

We have considered the role of the effectiveness and indeed, even the existence, of adequate support services for the children in need and their parents in a number of different social circumstances. The next paper that we wish to present is an excellent review of the support that is given to another specific sub-group and that is women and children who suffer from domestic violence. Webb and her group (etal 2001) considered the problem in considerable (and commendable) depth

The study itself had an entry cohort of nearly 150 children and their mothers who were resident in a number of hostels and women’s refuges that had been the victims of family violence at some stage in the recent past. The study subjected the cohort to a battery of tests designed to assess their physical, emotional and psychological health, and then quantified their access to, and support gained from, the primary healthcare teams and other social service-based support agencies. This study is presented in a long and sometimes difficult tread format. Much of the presentation is (understandably) taken up with statistical, ethical and methodological matters – all of which appear to be largely of excellent quality and the result of careful consideration.


The results make for interesting and, (in the context of this review), very relevant reading. Perhaps one of the more original findings was that nearly 60% of the child health data held by the various refuges was factually incorrect. This clearly has grave implications for studies that base their evidence base on that data set(Berwick D 2005).

Of great implication for the social services support mechanisms was the finding that 76% of the mothers in the study expressed concerns about the health of their children. Once they had left the refuge there was significant loss to the follow up systems as 15% were untraceable and25% returned to the home of the original perpetrator.

The study documents the fact that this particular group had both high level of need for support and also a poor level of access to appropriate services. In the study conclusions, the authors make the pertinent comment that the time spent in the refuge offers a “window of opportunity” for the family support services to make contact and to review health and child developmental status.

This is not a demographically small group. In the UK, over 35,000children and a parent, are recorded as passing through the refuges each year, with at least a similar number also being referred to other types of safe accommodation. Such measures are clearly not undertaken lightly with the average woman only entering a refuge after an average of 28separate assaults. One can only speculate at the long term effects that this can have on both the mother and the children.

Suggestions for improvement

In common with the other papers reviewed, this paper also calls for greater levels of support for the families concerned as, by inference, the current levels of effectiveness of the family support services is clearly inadequate.


This review has specifically presented a number of papers which have been chosen from a much larger number that have been accessed and assessed, because of the fact that each has a particularly important issue or factor in its construction or results.

The issue that we have set out to evaluate is the effectiveness of the family support services which are specifically aimed at reducing the stress levels for the parents of children in need. Almost without exception, all of the papers that have been accessed (quite apart from those presented) have demonstrated the fact that the levels of support from the statutory bodies is “less than optimum” and in some cases it can only be described as “dire”.

Another factor that is a common finding, is that, given the fact that any welfare system is, by its very nature, a rationed system, the provision of the services that are provided is seldom targeted at the groups that need it the most. One can cite the Tickler (et al 2000)and Downey (et al 1999) papers in particular as demonstrating that substantial proportion of the resources mobilised are actually being directed to groups that are either not requesting support or who demonstrably need it less than other sectors of the community. Some of the papers (actually a small proportion) make positive suggestions about the models for redirecting and targeting support. Sadly, the majority do little more than call for “more research to be done on the issue”.

In overview, we would have to conclude that the evidence suggests that the effectiveness of the family support services in reducing stress and poverty for the parents of children in need is poor at best and certainly capable of considerable improvement.


Bassuk, E., Buckner, J., Weiner, L., et al (1997)
Homelessness in female-headed families: childhood and adult risk and protective factors.
American Journal of Public Health, 87, 241–248 1997

Berwick D 2005
Broadening the view of evidence-based medicine
Qual. Saf. Health Care, Oct 2005; 14: 315 - 316.

Black D. 1996
Childhood bereavement: distress and long term sequelae can be lessened by early intervention.
BMJ 1996; 312: 1496

Black D. 1998
Coping with loss: bereavement in childhood.
BMJ 1998; 316: 931-933

BPA 1999
British Paediatric Association. Outcome measures for child health.
London: Royal College of Paediatrics and Child Health, 1999.

Brooks RM, Ferguson T, Webb E. 1998
Health services to children resident in domestic violence shelters.
Ambulatory Child Health 1998; 4: 369-374.

Cumella, S., Grattan, E. & Vostanis, P. (1998)
The mental health of children in homeless families and their contact with health, education and social services.
Health and Social Care in the Community, 6, 331–342

D of H 1995
Department of Health. Together we stand: the commissioning, role and management of child and adolescent mental health services.
London: Stationery Office, 1995.

D of H 1998
Department of Health. The quality protects programme: transforming children's services.
London: Department of Health, 1998. (Local authority circular LAC (98)28.)

Dowdney L, Richard Wilson, B Maughan, M Allerton, P Schofield, and D Skuse 1999
Psychological disturbance and service provision in parentally bereaved children: prospective case-control study
BMJ, Aug 1999; 319: 354 – 357

Feldman MD, Ford C, Reinhold T 1994
Patient or Pretender: Inside the Strange World of Factitious Disorders
London: Wiley, ISBN 0-471-58080-5

Fristad MA, Jedel R, Weller RA, Weller EB. 1993
Psychosocial functioning in children after the death of a parent.
Am J Psychiatry 1993; 150: 511-513

Hall D. 1996
Health for all. 3rd ed.
Oxford: Oxford University Press, 1996.

Hall D, Lynch MA. 1998
Violence begins at home.
BMJ 1998; 316: 1551-1560

Harrington R. 1996
Childhood bereavement: bereavement is painful but does not necessarily make children ill.
BMJ 1996; 313: 822

Harris T, Brown GW, Bifulco A. 1996
Loss of parent in childhood and adult psychiatric disorder: the role of lack of adequate parental care.
Psychol Med 1996; 16: 641-659

Kerouac S, Taggart M, Lescop J, Fortin M. 1996
Dimensions of health in violent families.
Health Care Women Int 1996; 7: 412-426.

Kranzler EM, Shaffer D, Wasserman G, Davies MA. 1990
Early childhood bereavement.
J Am Acad Child Adolesc Psychiatry 1990; 29: 513-519

Lissauer, T., Richman, S., Tempia, M., et al (1993)
Influence of homelessness on acute admissions to hospital.
Archives of Disease in Childhood, 69, 423–429.

Meltzer H, Gatward R. 1999
Mental health of children and adolescents in Great Britain. In:
Norwich: Stationery Office, 1999.

Mohammed, D Braunholtz, and T P Hofer 2003
The measurement of active errors: methodological issues
Qual. Saf. Health Care, Dec 2003; 12: 8 - 12.

Morton V, Torgerson DJ. 2003
Effect of regression to the mean on decision making in health care.
BMJ 2003 May 17;326: 1083-4.

O'Hara M. 1995
Domestic violence and child abusemaking the links.
London: National Children's Bureau, 1995.

Power S, Whitty G, Youdell D. 1995
No place to learn: homelessness and education.
London: Shelter , 1995.

Sanchez L, Fristad M, Weller R, Weller EB, Moye J. 1994
Anxiety in acutely bereaved prepubertal children.
Ann Clin Psychiatry 1994; 6: 39-43

Shankleman J, Brooks R, Bryan C, Davies M, Webb E. 2000
Emotional and behavioural disturbance in pre-school and school-aged children resident in refuges.
Arch Dis Child 2000; 82: A55

Tickler V, S. Cumella, T. Bellerby, and P. Vostanis 2000
Service innovations: a mental health service for homeless children and families
Psychiatr. Bull., September 1, 2000; 24(9): 339 - 341.

Vostanis, P. , Grattan, E., Cumella, S., et al (1997)
Psychosocial functioning of homeless children.
Journal of the American Academy of Child and Adolescent Psychiatry, 36, 881–889.

Vostanis P, Grattan E & CumellaS et al 1998
Mental health problems of homeless children and families: longitudinal study. British Medical Journal, 316, 899-902.

Vostanis P, Cumella S, Grattan E. 1999
Psychosocial functioning of homeless children.
J Am Acad Child Adol Psychiatry 1999; 36: 881-889

Vostanis P 2002
Mental health of homeless children and their families
Advan. Psychiatr. Treat., November 1, 2002; 8(6): 463 - 469.

Webb E. 1998
Children and the inverse care law.
BMJ 1998; 316: 1588-1591

Webb E, Judith Shankleman, Meirion R Evans, and Rachel Brooks 2001
The health of children in refuges for women victims of domestic violence: cross sectional descriptive survey
BMJ, Jul 2001; 323: 210 - 213 ;

Welsh Office. 1997
The health of children in Wales.
Cardiff: WO, 1997.

Welsh Office. 1999
The Children First programme in Wales: transforming children's services. Cardiff: Welsh Office, 1999. (Circular 20/99.)

Wilkinson R (1996)
Unhealthy Societies: The Affliction of Inequality,
Routledge, London 1996

Worden JW, Silverman PR. 1996
Parental death and the adjustment of school-age children.
Omega J Death Dying 1996; 33: 91-102.