Scottish Government

Findings

Our initial work resulted in over 10,000 titles or abstracts which might have been relevant. After removing duplicates and others that proved non-relevant, including those identified through the Level One scoping exercise, and the subsequent more detailed review at Levels Two and Three, we have approximately 2,600 entries in the Reference Manager database. Key criteria for the exclusion of entries from the database (unless an overlap with the overall aim and objectives of the scoping study could be identified) included: literature that pre-dated 1990; biochemical and pharmacological research into substance misuse; the genetics of substance misuse; substance misuse by adolescents (unless this related to use as an effect of parents' misuse, and we have retained material that confirms that early or problematic substance use is a risk factor resulting from having a substance misusing parent); literature related to the general areas of prevention of substance use/misuse in young people; domestic violence (if solely related to inter-spouse violence with no mention of children or family issues); interventions (again if child or family issues were not mentioned) and smoking.

All references are in the English language (even with foreign journals, where the title and/or abstract are published in English). The majority are from the USA. Quantitative work and basic exploratory work ( i.e. on the impact on, or risks for, children) dominate. Over half of the entries are journal articles, though this includes abstracts from the USA resource Dissertation Abstracts International, with a further 135 book chapters, 60 reports and 40 books. About 40 references, mainly journal articles and reports relate specifically to Scotland.

It was immediately apparent that there is a very considerable amount of material looking at impact and risk, whilst resilience and interventions are growing areas. Child protection and domestic violence are two highly relevant and overlapping areas of literature and these themes feature strongly in much of the literature, particularly the literature relating to child protection. Similarly, pregnancy, motherhood and fetal alcohol spectrum disorder were large areas. It was apparent that there was a lack of literature focusing on children's views, the father's role in parenting, service needs, service provision, mental health and ethnicity. There is a need for further work in these areas to establish how they may impact upon parental substance misuse and its impact on children and families.

The following sections contain the reports of the topics reviewed at levels two and three. Apart from the topic of fathers, which reviewed just over 100 abstracts, approximately 30-60 abstracts were reviewed for each topic. For the most part, literature tended to be American, published journal articles and based on quantitative work. Particular exceptions to this are highlighted in the sections that follow. Many of the points to emerge from the review of the service needs literature was felt to mirror other topics reviewed, and so this literature was integrated throughout the report (including the discussion). Similarly, we read through the Scottish specific references, and evidence and ideas contained within them were integrated into the appropriate section of the report, or into the discussion and recommendations sections.

Overall, the literature rarely differentiated between the substance misused by parents in terms of impact or need; this issue is explored in more detail in the Discussion (see particularly p37).

Prevalence

There is a dearth of literature that examines the prevalence of parental substance use at a general population level. Prevalence information usually comes at the start of articles, by way of background, and is rarely detailed enough to trigger further discussion. Estimates of the number of family members affected by someone else's substance misuse are rare: it is much more common for papers to estimate the prevalence of substance misuse more generally.

Prevalence estimates can be calculated from national survey data, from smaller samples of specific populations or by considering the measurement of problems faced by children and families of substance misusers. However, the majority of this work is American. Gomby & Shiono (1991) reviewed USA work that focus specifically on newborns. They discussed how prevalence calculations might be over- or underestimates, thus highlighting some of the key challenges and issues to consider.

Hidden Harm (2003) estimates that there are up to 60,000 children under 16 years old in Scotland who have a parent with a drug problem (approximately 5% of the total population group for this age). Further estimates indicate that 10,000-20,000 children live with a drug-using parent whilst the number of babies born to drug-misusing mothers rose to nearly 18 per 1,000 in 2000 ('It's Everyone's Job to Make Sure I'm Alright', 2003). 'Getting our Priorities Right' (2002) estimated that there are 40,000-60,000 children in Scotland affected by parental drug use and that 19% of the 10,798 adults reported to the Scottish Drug Misuse Database in 2001/2002 were living with dependent children.

At present, the method used by the ACMD in calculating the figure for Hidden Harm is the best estimate available for numbers of children affected by drug misuse in Scotland, where figures for substance misusing parents were calculated by using a method combining a number of sources: 1) the prevalence results reported by Hay et al. (2001), 2) the Drug Outcome Research in Scotland ( DORIS) figures and 3) SDMD figures. The latter two figures are based on new contacts known to treatment services only and are therefore limited. However, Hay et al. (2001) developed a hidden population model to attempt to estimate the people not in touch with services. Data sources were treatment contacts ( SDMD), drug related hospital admissions, police contacts and social work/social enquiry records and were identified by Council area or by postcode. This enabled figures to be re-calculated based on NHS Board area, DAAT areas and police force areas.

Pavis (2004) has highlighted variations in prevalence across areas / NHS regions and thus emphasised the importance of localised prevalence estimates when considering service delivery for those misusing substances, and their families. As an example of such a local calculation, Murray and Hogarth (2003) estimated that " there are an estimated 1,306 children and young people affected by substance misuse within the family" in the Borders region. This is 1.2% of the overall population of 109,270 at the time of the study. Limitations to this work are that it covers (some) illegal drugs only, focused on adults in direct contact with services, and did not employ a standard methodology for estimating prevalence; thus such calculations are certain to be underestimates. Hay, Gannon & McKeganey (2005) have done work on estimating the number of children affected by parental substance misuse in Glasgow, deriving an estimate of 6,142 such children (5.5% of the under 16 population), with slightly more living with at least one parent with an alcohol problem than with a drug problem. In both cases it is more likely that the father has the alcohol or drug problem.

For alcohol, the most recent and widely cited figures are found in the Plan for Action on Alcohol Problems, which indicates that 80,000-100,000 children in Scotland are affected by parental alcohol use. The 2005 Alcohol Manifesto (Alcohol Focus Scotland) also reports a figure of 100,000 children living with a parent with an alcohol problem. However, there may be particular problems in calculating estimates of numbers of children affected by alcohol misuse. First, the hidden population may be much greater and thus estimates based on treatment services are likely to be underestimates. Secondly, particular attention has to be paid to terminology and whether estimates relate to dependence, problem drinking, binge-drinking, and how these terms are defined.

Hidden Harm (2003) also reported that, on average, 25% of children on child protection registers were there because of parental alcohol or drug use. Other estimates of substance use in families in contact with social workers are much higher, ranging from 20-78% depending on sample size and point of contact with social services (see Forrester & Harwin, 2004; Hayden 2004). Figures for Scotland for 2004-2005 state that over 12,000 children were looked after, with 2,157 on the CPR at 31 ^st March 2005 (Scottish Executive, 2005).

Conclusions and recommendations

It is clear that there are broad estimates of the numbers of children affected by parental substance misuse across Scotland, and that taking alcohol and drugs misuse together these figures probably exceed 150,000. This is in line with the widely held belief that substance misuse (particularly alcohol) affects many people, and that children are a group most likely to be affected. However, commentary on these documents, and on calculating prevalence in general, suggests that these figures are likely to be underestimates. There is a dearth of specific knowledge (as opposed to generalised estimates) in this area, with a need for clearer national and local data, and estimates drawn from the most reliable methodologies in order to drive forward Scottish policy in this area. However, the availability of data, and the best method for making such calculations, remains unclear. Work by the EMCDDA (Frischer et al., 2001) has tested various methods for estimating problematic drug misuse prevalence in Great Britain. The work of Hay et al. (2005, 2001) in Glasgow is important and further consideration should be given to the application of their capture-recapture methodology (which can include calculations of the 'hidden' population, rather than limiting data solely to information contained within health, police and social services databases). Some work has been done to test the suitability of capture-recapture methodologies to the alcohol population (Corrao, Bagnardi, Vittadini & Favilli, 2000) but, as alcohol is a legal substance, existing information systems and data collection processes relating to alcohol misuse are unlikely to support the ready application of such methodologies. Further work to explore potential methods to calculate prevalence in relation to alcohol misuse would therefore be particularly useful.

Work on prevalence must consider its definitions of substance misuse ( e.g. alcohol or drugs, which drugs, what is defined as misuse) and who is being counted ( e.g. all family members, just children, age restrictions). The definition of parental substance use needs to include both mother's and father's use. Given that around a fifth of the Scottish population live in rural areas (and 6% of those in remote rural areas), having clearer prevalence estimates by area would further help define where adapted practice and policy approaches are needed.

The Scottish Executive response to Hidden Harm (2005) does not make specific recommendations relating to estimating prevalence. However, there are clear recommendations that the Scottish Drug Misuse Database ( SDMD) collect data on the children of those included in the database. This is a recommendation that has already been acted upon, with the decision that workers will be asked to collect information during initial assessment on the number of dependent children of the client (biological or those who they care for 'as a parent'); ages of these children; whether the client's children live with them or not; and whether the client or their partner is pregnant. At follow-up (usually 3 and 12 months), and at discharge, this information will again be collected. Workers will also be asked to record whether the client is receiving an intervention related to their children; and which type of organisation is providing the child orientated intervention. Whilst this is an important step, it remains to be seen how this change operates in practice. No information will be collected on the identified needs of the children or risks to them. Whilst we understand that a conscious decision was made not to collect information on risk at a national level, because data processing times would not help protect children, we recommend that these data continue to be collected at a local level by drug (and alcohol) treatment workers. Additionally, as the Scottish Drug Misuse Database is not designed to collect information about clients with solely an alcohol problem, alternative methods of collecting data on child-related issues from clients of services whose primary presenting problem is alcohol should be explored.

It is still the case, therefore, that further prevalence work is needed, both nationally and locally across Scotland, to understand the numbers of family members affected by alcohol or drug problems. Issues of methodology and definition, along with consideration of hidden, non-clinical populations, are vital.

Child Protection and Domestic Violence

The clear message from the literature on these subjects was that problematic substance use has a negative impact on parenting skills and parental attention to the child. In particular the literature identified the heightened risk to children of all forms of abuse where one or more parents were using substances problematically. It also identified how parents with substance problems often had histories themselves of child abuse and neglect. However, it is important to note that the literature focuses on maternal (as opposed to paternal) substance misuse.

An important finding was the co-existence of domestic violence as a key factor in the environments of parents with substance problems, and how the combination of both factors increased the negative impact in all areas: the child's development, their experiences in adolescence, relationships and parenting abilities as adults, and prediction of adolescent psychopathology, perpetrating child abuse, developing substance problems, or perpetrating or suffering domestic violence in adulthood. Experiences of domestic violence and sexual abuse were also highlighted among women in treatment and in prison populations. Suffering domestic violence was also identified as a risk factor for the mother, with some studies suggesting it increased her punitive response towards their children.

A number of studies highlighted how parental substance use, as a single factor, is not solely responsible for increased risks of child abuse. Environmental factors including poverty, social isolation, and lack of family or community involvement increased the risk of harm to children and/or the removal of children from parental care. A related issue was the perception by parents (usually mothers) that substance use treatment would help them to keep custody of the child but at the same time they feared that approaching treatment services might prompt child protection procedures. This fed in to recommendations by a number of studies that services needed to assess and address the holistic needs of parents and children. They suggested treatment not only needed to accommodate children but that supplementary services concurrently addressing non-substance specific needs demonstrated better outcomes for parenting and children and reduction in child abuse.

Within the broader group of substance using parents there was a range of specific parenting sub-sample populations. These included pregnant women, women in prison, and adult children of substance using parents who were now parents themselves. Pregnancy, in particular, was seen as an opportunity for risk assessment and intervention; with childbirth and children being both a motivator and demotivator for help seeking (see separate section on this topic for more detail). Some articles raised questions about mandatory assessment and/or testing of babies born to substance using mothers.

In relation to service provision and professional intervention, several studies identified social workers as needing better knowledge and support in order to address substance use within their child protection role. The failings of the child protection structures and staff to address substance use were raised in various forms, from lack of knowledge of social workers to the separation of child and adult services leaving gaps in service provision (within Scotland this was a clearly identified finding to emerge from the Caleb Ness Inquiry). One study showed that where there were good relationships between social worker and parent the outcome for the children was better. Some literature identifies guidance for existing services and describes models of service provision that are addressing parental substance use and its impact on children and families.

Some studies found differences in relation to the types of harm children suffered depending on the gender of the child and the gender of the parent with the problem, namely, how parental substance abuse and/or parental domestic violence impacted upon girls and boys differently. The impact on boys tended to manifest itself through externalised behaviours, such as increased aggression, whereas girls internalised the negative effects and were more prone to withdrawal and mental ill health. However, further work in this area is needed. Similarly, some studies paid particular attention to emotional neglect and abuse and identified how this had an equally negative impact on children and young people as the more commonly identified physical and sexual abuse. This supports some of the reports of parents being 'absent' during their childhood as a result of substance use. However, these categories are unlikely to be mutually exclusive and therefore the methodology needs further exploration.

One of the positives identified by adult children was the need for their own parenting to be different and to act as the role models of what parenting should be, rather than what their own parents presented. Finally, some articles looked at how grandmothers were often involved in taking care of children as a result of parental drug use and subsequent child neglect. They also identified the need for greater support for grandmothers who had to return to child rearing in later life.

There was very little evidence of qualitative work, nor of studies examining the views of parents and children. One study (Richter & Bammer, 2000) outlined the strategies mothers took to reduce the harm to their children when they were using and this could form the basis for a more positive and strength-based approach to developing services and interventions.

Some literature identified the need for, or recommendations about, tools to assess or support childcare professionals in their assessment of and intervention with parents using substances. Similarly, there was literature describing service initiatives that focus specifically on intervening with parents or families when children are at risk. However, there appears to be an absence of a review of 'what works' in relation to child protection. This may be filled to some degree with the completion of an ongoing project on this subject by Hedy Cleaver in the UK (2005). This work addresses both substance use and domestic violence, and includes service user and provider perspectives.

One study looked at social services responses in cases of both domestic violence and parental substance misuse (Cleaver et al., 2005) and found that social worker plans and training were more likely to cover domestic violence than parental substance misuse. Links between domestic violence and substance misuse were rarely made; awareness by managers of local services for domestic violence or substance misuse varied between and within authorities. Furthermore, working relationships between adult and children's services, particularly housing and substance misuse services, were generally poor. Three quarters of initial assessments indicated that parenting capacity was negatively affected and children had unmet developmental needs; yet few initial assessments resulted in a referral to services for domestic violence or parental substance misuse. Greater focus should be given to improving information sharing and collaborative working between services for domestic violence and substance misuse, and children's services. Agreed, robust protocols and procedures should ensure that timescales and ethical considerations do not hamper joint-working.

Many of the abstracts finished by stating they had implications for policy and service provision but these were not explicit in the abstracts. Scotland is quite advanced in developing such guidance, stemming largely from Getting our Priorities Right (2003) and the Caleb Ness Inquiry. At least three areas of Scotland, Edinburgh and the Lothians (2005), the North East of Scotland and the Borders (2004) have produced child protection guidelines that consider how to work with children in families with problem substance use.

Conclusions

It is clear that:

• Problematic substance use can have a seriously negative impact on parenting skills and parental attention to the child, with a corresponding impact on many areas of the child's life.
• Co-existing domestic violence usually increased the negative impact in all areas, both in terms of its impact on the mothers' ability to parent and the children's outcomes.
• There are common intergenerational continuities in this area with parents with substance problems often having histories themselves of child abuse and neglect.
• Many other factors besides parental substance use are important in determining the impact on children, including the environmental factors of poverty, social isolation, and lack of family or community involvement.
• There are many suggestions that interventions need to be holistic and integrated, and that these lead to better outcomes for both parents and children.

Key recommendations:

• Undertake further research on fathers, both in terms of their role as substance misusing parents and as non-using partners to substance misusing mothers.
• Undertake further research on fathers as perpetrators of substance-related domestic abuse and the father's responsibility for parenting in this context.
• Undertake further research or more detailed reviews on differences in harm children suffer, depending on their gender, the gender of the parent with the problem, and the substance misused.
• Develop a focus on more qualitative work, work that examines the views of parents and of children, and work which focuses on positive, resilient and strength-based approaches (for example, the strategies mothers take to reduce the harm to their children when they are using).
• Undertake a review of 'What Works' in relation to child protection, especially with overlapping issues of substance misuse, and of domestic violence.
• Ensure that services are provided more holistically, focusing on all aspects of parenting, substance misuse and domestic violence, and on interventions to help children in these situations; and in an integrated form so that child and adult services are not divided.
• Improve qualifying and post qualifying social work training to ensure that it includes training on alcohol and drug use and how this relates to working with children and families. The training of child and family social workers should be a priority; the role of adult social workers was not a focus of this review and would need further investigation.

Resilience

The majority of abstracts looked at resilience or coping with parental alcohol problems. Many studies used qualitative methods, but did focus primarily on exploratory work. Few studies used resilience or coping information to develop and evaluate interventions to increase resilience.

Many studies demonstrated that ' ACOAs' (adult children of alcoholics) do not have more problems (unemployment, difficulties at work or education, or in relationships, different personality profiles, etc.) compared with non- ACOAs. In fact, many studies started from a realisation that the older research and clinical literature which portrayed the ' COA' or 'ACOA' populations as homogeneous and ordained to negative outcomes was incorrect. Much of the research undertaken in the past 15 years has instead demonstrated that this population is very heterogeneous, and that there is no pre-ordination for negative outcomes. Instead, these children seem to have a wide range of coping behaviours and resilience features. Some children and young adults do demonstrate negative outcomes which seem related to their negative childhood experiences (family conflict, inconsistent parenting practices, etc.) whilst others seem more protected and resilient, which seems related to such factors as support from a non-problem parent, or intact family rituals (Gogineni, 1995). A number of studies within this set of abstracts (for example: Leahy, 1997, Harris, 1999, Hogan 2000) have shown that although adults who were the children of problem drinkers report (as compared to control groups) very disturbed early family environments, they score no differently to control samples on measures of current adult functioning (see also Velleman & Orford, 1999).

Coping

A number of studies examined coping strategies, comparing these children or young adults with control groups. Amongst the reported findings were that children of substance misusing parents used emotion-focused coping strategies at all ages whereas adolescents tended to also discuss problem-focused strategies (Amond-Berry, 2000). Females tended to discuss using more passive and internally referenced coping strategies whilst males tended to discuss using more aggressive and externally referenced coping strategies (Amond-Berry, 2000). Devine & Braithwaite (1993) showed that although 'parental alcoholism' contributed to children adopting the 'acting out' and 'placator' roles and was the sole predictor of the adoption of the 'responsible child' role, the adoption of these 'survival roles' appears to be as much a response to family disorganization as to 'parental alcoholism'. Pilowsky et al. (2004) showed that more resilient children of injecting drug users were less likely to use internalising and externalising avoidance coping strategies but it is unclear (from the abstract) how coping was defined or what coping strategies were used instead.

Resilience

Many studies have shown that there are highly resilient individuals who are young adult children of problem drinking parents, although these studies also show that significant numbers are not functioning as competently as young adults from other non-problem backgrounds.

Resilient children seem to be remarkably similar to children from non-substance misusing families. Mohr (2000) found that they did not differ on any measure from others without substance misusing parents who did not develop substance misuse. Similarly, Gordon (1995) showed that few differences were identified in any of the variables they examined (for example, measures of security of attachment, use of alcohol, affective characteristics, disturbances of the self, interpersonal functioning, psychological symptomatology, family environment, and coping styles), " suggesting that not all ACOAs suffer long term consequences of growing up with an alcoholic parent". They did show that their ' ACOA' sample had a more "avoidantly attached" attachment style, which they suggest may have served to protect these individuals from the deleterious effects of parental 'alcoholism'.

A number of studies have shown that, contrary to expectations, maternal attachment, security of attachment and quality of maternal parenting do not always operate in the hypothesized protective manner (Curran & Chassin, 1996; Gordon, 1995; Mohr, 2000). Some work (Cavell et al., 2002; Hill et al., 1992) reports on the positive influence of positive family relationships and dynamics.

Pilowsky et al. (2004) showed that the level of actual support received by resilient and non-resilient children did not differ significantly, but that perceived support was greater among resilient children. The work of Chandy et al. (1993, 1995, 1996) identified protective factors which included less worry about abuse from parents, the perception that school personnel cared about them, positive parental expectations, rating self as generally healthy, and religion.

One way that many children from this background demonstrate their resilience is by the professions they enter: there were a number of studies reporting adults who were children of substance misusers now successfully engaged in careers as therapists, social workers, medical students or doctors ( e.g. Coombes & Anderson, 2000). Such papers often make the point that the adversity experienced by such children is sometimes transferred into a positive outcome, in a way similar to that reported within the 'Post-Traumatic Growth' literature.

Risk

Although these abstracts were chosen to focus on resilience, a number also examined risk. Some of their conclusions were that: the number of problem drinkers in the household was the strongest predictor of adolescent substance misuse (Mohr, 2000); that family conflict predicts adult alcohol problems in ' ACOAs' (Gogineni, 1995); that dysfunctional family processes lead to greater negative impact on childhood self esteem than parental substance misuse itself (Godsall, 1995).

Interventions

Significantly, there were very few studies that took any of these findings on resilience, coping or risk and translated them into interventions to test out whether it was possible to alter these potentially predictive relationships. A number of studies presented 'frameworks' to help develop such interventions, or made suggestions of what these interventions might look like, without actually testing them ( e.g. Begun & Zweben, 1990). Mylant et al's. (2002) work proposed that mental health professionals teach core resiliency factors to promote healthy behaviours for this vulnerable population. Finkelstein et al. (2005) and Arman & McNair (2000) described their group focused interventions but no results or evaluative data were available or presented.

One of the two sets of studies that did provide and then monitor the effects of an intervention is Catalano et al's. (1997, 1999, 2002) 'Focus on Families' programme. This approach combines parent skills training and home-based management services to reduce parents' risk for relapse to substance misuse and children's risk for the development of problems with substance use while enhancing protection. In 2002 they summarised all the results of their intervention: parents in methadone treatment can be successfully engaged, and will participate in intensive family interventions; the risk- and protective-focused intervention increased parent relapse prevention skills; the intervention had effects on reducing parents' drug use, domestic conflict, and deviant peer networks, increased the number of family rules and meetings and influenced parental coping. However, little data is available that focuses on the children.

Aktan et al. (1996) evaluated the effectiveness of the Safe Haven Program, a family skills training program for African-American families where one parent is a substance abuser. The evaluation was conducted on 88 substance-using and non-using parents and 88 children (aged 6-12 yrs), and they showed that the program was effective in increasing parenting efficacy and behaviours toward children, improving the children's risk and protective factors and behaviours, and supporting treatment reductions in the parent and family illegal substance use.

Conclusions and recommendations

These studies demonstrate that many children will grow to be resilient, although many will remain at risk. However, there is still a considerable lack of clarity over what many of the resilience factors are which determine these positive outcomes, and whether the factors which relate to resilience in this area are similar to or different from those factors which have been shown to be effective in more general resilience research. It is also the case that very few interventions have been developed to alter the social dynamics within families such that protective factors are increased and risk factors are reduced. Newman (2004) summarises key points and key messages related to resilience in the early years, middle childhood and adolescence / early adulthood. There may be ideas here that can inform research, practice and policy. This is a major area for further research.

Pregnancy and Motherhood

This was not a specific area listed in the original objectives for the scoping review, yet it emerged as an important area in which a lot of work had been undertaken.

The majority of the work in this area has focused on alcohol and cocaine. Many studies appeared to take as their starting point a clear understanding that pre- and postnatal alcohol or drug misuse can have serious negative effects on babies and their subsequent development. There is quite a lot of work specifically on fetal alcohol spectrum disorder but little that we found on neonatal abstinence syndrome. Ornoy et al. (1996) reinforce the point that there is considerable evidence that the use of 'drugs' or 'alcohol' alone is not the important issue: it is the type of drug or alcohol, the amount taken at any one time, the number of times this amount is taken, and the stage of pregnancy at which this occurs, as well as a host of other, environmental, factors, which determine what, if any, negative impact substance misuse may have on a baby. As one example, and of relevance to the current national focus on binge-drinking trends, is a study that indicates a particular foetal risk if the mother binge-drinks, but the abstract is unclear on what constitutes a binge (Maier & West, 2001).

Many of the abstracts focused particularly on the impact of substance misuse on parenting (which usually means mothering), and on the mother-child interaction. However, very little of this work looked at the child's perspective, although one qualitative study reported on research with young mothers whose own mothers had misused substances. These young mothers demonstrated the long-term impact of maternal substance use and its impact on their own role as mothers, and wished to parent their own children differently. Linked to this is how becoming a parent ( i.e. a mother) when also a substance misuser can act as either a catalyst or a barrier to seeking help or treatment: being a substance misuser can either motivate or stop women accessing pre-natal care, and becoming a (prospective) parent can either motivate or stop women accessing substance misuse treatment services. Many women are fearful (see also the 'Child protection' section earlier in this report), that approaching treatment services may prompt child protection procedures.

The need for services to assist pregnant substance users is being increasingly recognised. Where work has focused on treatment services for women, positive outcomes have been demonstrated. These services are usually residential and sometimes accommodate the children. However, these outcomes relate mainly to the mother with few child-related outcomes reported. Additionally, few studies presented the views of the children. Some qualitative work has shown that, whilst women are aware of the risks of being a parent who is also a substance misuser, many women argue that they are still able to be good parents. It is the case that these views challenge traditional stereotypes of parenthood and mothering and such a challenge is appropriate; however, these mother's views do not appear to have been confirmed by additional work with children and others that might show whether or not substance misusing parents are still able to be good parents (see also Taylor, 1993 who conducted an ethnographic study with over 50 female injecting drug users in Glasgow).

It has been suggested that pregnant substance users could benefit from being managed using a shared care approach, involving obstetric services in conjunction with a substance misuse agency. Obstetric goals need to take account of pharmacological treatments, but should also shift towards a public health perspective, characterised by treating pregnant and postpartum substance misusers, protecting at-risk foetuses and children, and strengthening broken families. There is a need to educate pregnant women around alcohol, and the involvement of important people in mothers' social networks may be key to reducing substance misuse during pregnancy. Many studies have concluded that there need to be both women-specific and parenting components in existing treatment programs, where pregnant women who are substance misusers can benefit from comprehensive, family-centred treatment services and receive useful parenting advice. The benefits of specialist teams that treat addicted mothers and their babies have been demonstrated (Day et al., 2003), though there is a need for further work in this area with a view to developing more such services.

Many of the abstracts reviewed discussed fetal alcohol spectrum disorder. FASD covers the complete range of alcohol-related harm experienced by babies of mothers who drink during pregnancy. Children who have FASD are characterised by pre- and post-natal growth deficiency, distinctive facial features and moderate to severe learning difficulties and behaviour problems caused by central nervous system impairment. Families caring for children who have disorders associated with FASD have a particular and high need for support programmes and services. Higher rates of mortality and child custody disputes are particular issues. Problems regarding provision of services for FASD include: (1) lack of appropriate standards of care; (2) limited availability of programmes or services specific to FASD; (3) lack of clarity over what is most effective: a need for programme evaluations in this area; and (4) concomitant problems in the care giving environment. There is a need for services and programmes directed at FASD, and more research is needed to clarify and define the needs.

One study (Richter & Bammer, 2000), modelled directly from qualitative work with heroin-using females, describes a hierarchy of strategies that these mothers use to reduce harm to their children from maternal substance misuse. This hierarchy is:

1. Stop using;
2. Go into treatment;
3. Maintain stable small habit;
4. Shield children from drug-related activities;
5. Keep home environment stable, safe and secure;
6. Stay out of jail; and
7. Place with a caregiver and maintain as active a parental role as possible.

It would be interesting to see how useful such a hierarchy might be to working with families in Scotland and to what extent it might be possible to adapt such a model to the stepped provision of services. However, the work of McKeganey, Barnard and colleagues suggests that the process of keeping children shielded from drug-related activities and its negative impacts is by no means straightforward, so this issue would need to be carefully considered. Some work in Scotland has been done to specifically consider the clinical management of pregnant substance users (Scottish Borders region inter-agency children protection guidelines, 2004).

Conclusions and recommendations

• Becoming pregnant or being a mother whilst simultaneously misusing substances can be a barrier to accessing ante- or post-natal care, providing good parenting, and accessing substance misuse treatment. However, there is evidence that if treatment is accessed, this may be a good time to offer help, and that outcomes can be positive. A drawback is that the evidence on which this statement is based is largely from the USA, generally concerns residential treatment that is expensive, and has not adequately looked at experiences of and outcomes for children. Further work is needed to explore the extent to which these problems are mirrored in the UK / Scotland.
• A lack of attention to children, and other family members (particularly fathers - see Fathers section for more detail), is clear and addressing this must therefore be a recommendation for future work.
• Linked to the research on resilience is the identification of the importance of attending to protective and resilience factors and processes at all stages of the life cycle, including ante-, pre- and post-natally. Newman (2004) summarises key points and key messages related to resilience in the early years, middle childhood and adolescence / early adulthood. There may be ideas here that can inform research, practice and policy.

Fathers²

The literature in the database about fathers fell into three main categories.

• Level one: the impact on children of having a substance-misusing (primarily 'alcoholic') father. The majority of abstracts fell into this category.
• Level two: the impact of substance misuse (again primarily alcohol) on the father-child relationship and/or father characteristics / behaviours that mediate the impact of substance misuse on their children.
• Level three: the impact of substance misuse on fathers and fathering, exploring fathers' views and concerns and the fathering role. There were few abstracts here.

Level One

Several impacts for children of having a substance misusing (mainly alcohol) father were identified - their own substance use, psychopathology and psychiatric disorder, physical health, personality characteristics, psychosocial adjustment, adult attachment, cognitive functioning, school attainment or adjustment, behaviour conduct, risk and resilience factors. A small number of abstracts also looked at genetic transmission of addictive behaviours between substance-misusing fathers and their offspring. A small number of studies explored the differential findings associated with drug misuse and alcohol misuse ( e.g. Cooke et al., 2004; Kelley & Fals-Stewart, 2004). Both found that children of drug misusing fathers were at greater risk than alcohol misusing fathers of negative behaviours, psychosocial impairment and lifetime psychiatric disorder.

Level Two

The literature suggests a complex pathway between paternal substance misuse and unfavourable outcomes for children. Some studies focussed on the negative impact of substance misuse on various aspects of fathering (Das Eiden et al., 2002; Das Eiden & Leonard, 2000; Dumka & Roosa, 1995; Brooks et al, 1998), whilst others look at mediating factors ( e.g. family structure or paternal warmth) that may reduce the negative impact of (paternal) substance misuse on children. Some studies explored differences depending on whether fathers, mothers or both parents are substance misusers. Finally a number of studies focussed on the impact of substance misuse not only on the children, but also the (usually non substance-misusing) mothers in these families ( e.g. Frank et al, 2002; Fisher, 1998; Das Eiden & Leonard, 1999), and report negative psychological and physical outcomes associated with the fathers' substance misuse. These are all factors to take into account when considering intervention and service delivery.

Few abstracts made direct reference to paternal responsibility, and the importance of father-child relationships, and there appears to be a general sense that mother-child relationships are much more significant (Cavell et al., 2002; Tweed & Ryff, 1996). However, Tarter et al. (2001) take a unique angle, reporting the finding that children living with both parents have better outcomes, in terms of conduct problems and own substance misuse, than those whose fathers were absent - even when the father is misusing substances. They suggest this is because single men show more severe alcohol or drug misuse than those living with their families. Furthermore, they suggest that mothers with absent substance misusing partners have fewer resources for effective parenting. This at least seems to suggest a role for substance misusing fathers and reveal they can offer something to their children. However, given that these findings are from one study, further work would be needed.

Level Three

McMahon & Rounsaville (2002) assert that, "although a number of socio-economic forces have converged across cultures to make fathering one of the more prominent social issues of the new millennium, the status of substance-abusing men as fathers is rarely acknowledged in the conceptualisation of public policy, service delivery or research focusing on the adverse consequences of drug and alcohol abuse"' (p1109). The studies that do attempt to explore these issues further reveal that fathers are overwhelmingly placed, or place themselves, in a peripheral position where the care of their children is concerned.

In a study exploring drug addicted fathers' uncertainties about their importance to their children, Arenas & Greif (2000) describe how fathers often believe their children were better off without them. The authors found these men had a number of concerns about fatherhood, including, 'having no concept of what a father should be, confusing the roles of manhood and fatherhood, feeling inadequate as a provider and not knowing how to reconnect with children they have not seen, particularly daughters' (p339). The authors also describe the guilt the fathers felt if they abandoned their children. They go on to suggest possible interventions with fathers that focus on teaching them about positive fathering, and encouraging them to discuss their own parenting experiences. Again the study reveals the impact on parenting capacity, but more so, fathers' concerns about this. It is extremely positive to note the suggestion for working with such fathers on their fathering techniques, and their associated anxieties.

Only one study reported an impact on fathers where mothers were the substance misusers. Dumka & Roosa (1995) report mothers' problem drinking contributed to less positive father-child relationships. Further gaps in the literature include the role of fathers in families with substance misusing mothers. Studies looking into the efficacy of parenting interventions with fathers may also be useful in terms of finding strategies to build the types of relationship and family environment which are known to protect children from the risks associated with paternal substance misuse.

Conclusions and recommendations

There is a great deal of literature available that explores the impact of paternal substance misuse on children, but there is a major lack of research into fathering and fatherhood in relation to this area. Fathers are typically viewed as "entirely negative influences that need to be actively excluded from the lives of their children" (McMahon & Giannini, 2003, p337). Debates around the role of fathers within substance misusing families occur as part of a broader societal debate around the role of fathers in relation to social exclusion, environmental factors, and the role of the wider family and social networks. Thus, there is a need for further work and understanding, both on the role that fathers play in increasing risk to children, both directly via their negative behaviours and indirectly via both their negative behaviours towards the child's mother, and any lack of acceptance of responsibility for their role as a parent. However, it also needs to explore the more positive or protective role that they may play if they do accept responsibility as a parent for protecting their children or acting in ways that promote their children's resilience. There is little literature that has explored fathering in much depth; even fewer have explored fathers' voices on this subject. McMahon & Rounsaville (2002) suggest a number of areas that need investigation, including the ways in which substance misuse contributes to a 'compromise of fathering', the ways in which this compromise of fathering contributes to psychological distress in these men, and the ways in which intervention might be used to minimise the harm associated with paternal substance misuse. Some work is underway in Scotland (Whittaker, 2005) to investigate this issue in more depth but more work is clearly needed.

Children's Views

An identified gap throughout the literature is work investigating or reporting children's views about a whole range of issues related to parental substance misuse. This needs to incorporate qualitative work in particular as well as the views of young children. Two key pieces of work provide recent reviews of the literature (Gorin, 2004; Kroll & Taylor, 2003). Other work of note is by McKeganey and colleagues in Glasgow ( e.g. McKeganey, McIntosh & MacDonald, 2003). Due to a lack of detail in the majority of the other abstracts reviewed, this summary is heavily informed by these pieces. A number of key, overlapping, themes dominate the findings in relation to children's views. These include secrecy, isolation, emotions, conflict & disharmony, roles and coping. For the most part, these themes operate in two ways: within the family, that is between parents and children; and externally, between the child and the outside world, but the focus in the literature remains largely on the impact of substance misuse on children. The views expressed mirror the wider literature describing the impact of parental substance misuse on children so will not be detailed here. However, it is important to recognise the implications for service provision. Although some qualitative Scotland based work (Barnard, 2005) has been done to explore the impact of substance misuse on siblings, this focused on families where another child was the substance misuser rather than a parent. Further work is needed to explore the potentially differential impact and needs of siblings affected by parental substance misuse.

A key finding with particularly relevant implications for service provision relates to barriers to help-seeking. Reference is commonly made to children's reluctance to speak to people outside the family about the problems they are facing within it. A number of related reasons emerge, including loyalty, fear (of nothing being done), the reactions of others, shame and stigma. Interestingly, and somewhat conversely, children have reported feeling aggrieved that people have not tried harder to break down this barrier and uncover the truth (Kroll & Taylor, 2003).

However, regardless of the barriers, children's needs exist. Hay, Gannon & McKeganey (2005) have done the most recent prevalence related work in Scotland (focusing on Glasgow). Their conclusion states, "We know very little about those children - in particular we do not know who are caring for them when the mother is not caring for them. We know very little of the needs of these children or indeed what proportion have had their needs assessed or have remained hidden from services" (p28).

Because children are concerned about talking to people about what is going on at home, confidentiality is paramount. Children have frequently highlighted the importance of establishing trust when discussing their needs (Gorin, 2004). Gorin also describes the 'personal qualities' of helpers as being important for children - so, for example, confiding in someone who will listen, and someone who is kind and consistently valued by children (see also Bancroft et al, 2004; Liverpool DAAT, 2001). Whilst this may seem a simple need to fulfil, a number of studies have reported the fact that children often feel professionals in particular fail to listen or understand, and appear to talk in a different language. Additionally, professionals do not always speak directly to children (Gorin, 2004).

Christensen (1997) reports that children taking part in her study felt, 'the best place to get attention and help was…the treatment institution where the parents get help' (p24). However, a common finding (Gorin, 2004) is that children talk about needing confidential support, for example from helplines. Both Gorin (2004) and Kroll (2003) also report the desire some children have to meet others who have had, or are having, similar experiences. Hill, Laybourn & Brown (1996) report on a Scottish study that talked to children, young people and professionals about impact and need - "Many wished they could meet with others in the same position, so they could feel less isolated and learn from each other" (p159).

Support also comes from protecting some sense of normality. Children describe the need to get out of the house and engage in childhood activities (Gorin, 2004; Liverpool DAAT, 2001). They also describe calling on parents, other family members and friends for informal support (Bancroft et al, 2004; Gorin, 2004; Liverpool DAAT, 2001). It is important to bear in mind the efficacy of more informal sources and types of support. Gance-Cleveland (2004) reports the efficacy of a school-based support group for adolescents. The young people reported benefiting from experiential knowledge gained at the groups, and said the groups "enhanced self-knowledge and led to self-care and self-healing" (p379). The author concluded that the opportunity to share experiences and learn from others 'empowered' the young people to make changes to the dysfunctional patterns in their lives.

Identified needs in a study by Hill, Laybourn & Brown (1996) are for a range of services to be available including group work, counselling, family mediation and education. More help needs to be available at a generic level, which triggers referrals to specific services targeted at children. Hill, Laybourn & Brown (1996) highlight five things to bear in mind when thinking about help for children: that parental substance misuse is a widespread problem, that problems arise mainly from excessive problem drinking, that there are diverse problems and therefore diverse needs, that children need to ask for help, and the importance of informal sources of support, for example from non-misusing parents, siblings and other adults needs to be recognised.

Conclusions and recommendations

Whilst the focus of many of the studies exploring children's voices about parental substance misuse is on their negative experiences, it is by no means taken for granted that parental substance misuse per se has a negative impact on children. Finding out what children think and want enables researchers to unpick the complexity of the relationship between parental substance misuse and unfavourable outcomes for children. As discussed above these studies have revealed that it is associated factors, such as parental conflict and family disharmony or worry about drinking or drug taking, that most significantly affect children. The implication of this is that interventions focusing purely on parents' use may not be the most effective. Children need support in dealing with their often confused feelings and emotions towards their parents and their families, they need strategies to help them cope with the various consequences of their parent's substance misuse. This support needs to continue in its own right, regardless of where the parent may be in their treatment, and needs to be provided in an environment where children can feel safe to talk, but not forced to. How to initiate the provision of support is perhaps a more problematic issue. What is clear from the findings discussed above is that professionals need to be open to even the subtlest signs of parental substance misuse, and respond with sensitivity and patience.

As we, and others ( e.g. Gorin, 2004; Kroll, 2003) have found, there is a significant lack of research that has directly explored children's views and experiences of parental substance misuse, and therefore our understanding of impact, resilience factors, service needs, existing service provision, and intervention and treatment often lacks an essential and informative angle. A clear recommendation must therefore be the commissioning of more work that will directly explore the views and experiences of children living with parental substance misuse, and this work must include attempts to talk to children of parents who are and who are not receiving help for their alcohol or drug problem, the voices of children whose parents misuse alcohol and the voices of children whose parents misuse drugs, and the voices of children from different cultural backgrounds (Gorin, 2004; Kroll, 2003). The majority of studies to date have reported the views and experiences of children who have been in contact with services - a factor which might influence the conclusions drawn. Finally, such work should take account of gender, age and developmental stage (Cleaver, Unell & Aldgate, 1999).

Mental Health

The area of mental health in relation to parental substance misuse is a complex one. There were, however, few clear messages to emerge from the abstracts that were reviewed for this section, demonstrating a clear need for further work in this area. The key reason for this lack of clarity appears to be the number of possible relationships between mental health problems, parental substance misuse and the impact on children. The following list gives some examples of this variety:

1. Parental substance misuse and co-existing mental health problems, and the impact on parenting.
2. Parental substance misuse and its impact on children's mental health.
3. Parental substance misuse and its impact on children's mental health and co-existing substance misuse.
4. Parental substance misuse and co-existing mental health problems, and their impact on children's mental health.
5. Parental substance misuse and co-existing mental health problems, and their impact on children's substance use and co-existing mental health problems.
6. Prenatal substance use and its impact on children's mental health.
7. Prenatal/perinatal substance misuse and co-existing mental health problems.
8. Parental substance misuse and co-existing mental health problems, and their impact on children's substance use.
9. Parental mental health problems and its impact on children's substance use.

The above encompasses children who are infants/babies, younger children, adolescents and adult children. Any work in this area should take age, gender and developmental stages into account. Linked to this, 'mental health' could be interpreted widely to include both psychological development and psychiatric problems.

A number of the studies examined the extent to which parental substance use was harmful to the mental health or psychological development of the children in the family. The findings appear to be equivocal, with some evidence showing a clear negative impact on children (Cuijpers et al., 1999; Johnson 1995; Mathew et al., 1993; Obot and Anthony, 2004; Martin et al., 2000; Caudill et al., 1994; Williams and Corrigan, 1992; Carmichael-Olson et al., 2001; Beckwith et al., 1999). One study reviewed showed ethnic differences (Marse, 2002) and given the lack of research addressing ethnic differences this would be an area for further research.

Many studies, however, showed that parental alcohol and drug use does not predict psychiatric problems in offspring (Lehnert, 1998), nor that alcohol or drug use is solely responsible for children developing mental health problems. Any problems emerged from a number of individual and environmental variables (Glaun & Brown, 1999; Martin et al., 2000; Ellis et al., 1997; Luthar et al., 2003, Reder and Duncan, 2000; Lyman, 1997).

Several authors touched on service provision issues. Cornelius et al., (2001) found that parental substance problems and mental health problems acted as a barrier to accessing mental health services when their children needed them. Grella (1996, 1997) drew attention to the fact that services do not meet the needs of pregnant dually diagnosed women and Schwab et al., (1991) found that children's needs were not always met where there were "dual disordered" parents. Only one report reviewed addressed the practice and policy challenges of providing services to meet the needs of young children and parents with substance and/or mental health problems (Knitzer, 2000).

There were a number of studies that highlighted gender differences, either in terms of the children or the parent. Johnson (1995) found "female offspring" more likely to experience depression "regardless of the parental disorder", i.e. mental health or substance use problems, and that male children suffered more drug abuse problems. This was supported by other studies that tended to find female adult children suffering psychiatric problems and male adult children suffering alcohol and drug problems (Matthew et al., 1993). Conversely, Lynskey et al., (1994) found no gender differences. In relation to the gender of the parent, for example, Cuijpers et al., (1999) found the father's problem drinking was more closely related to the children's psychiatric problems, in particular the development of substance use problems. Luthar et al. (1993) found maternal mental health problems associated with mental ill health among children and the father's alcohol problems "showing associations" with black children's substance problems.

While this section focused on parental substance use and associated mental health problems, it was significant that domestic violence and childhood abuse were issues that arose in many of the studies. Alcohol-related domestic violence by the father had a negative impact on the children's mental health (Malpique et al., 1998). Chaffin et al., (1996) found that parental substance abuse and psychiatric problems increased risk factors for physical abuse and neglect, while physical and sexual abuse were features of adult children with mental health problems (Lehnert, 1998). In addition, Killeen et al., (1995) found pregnant dually diagnosed women had significant histories of childhood abuse. Moss et al. (1995) looked at the impact of father-son relationship in the context of paternal mental health and substance problems. While these sons were more aggressive this was not as a result of substance use or psychiatric problems of the father, rather it was associated with the father's other personality variables including paternal aggression and low self-esteem. Das Eiden and Leonard (2000, Das Eiden et al., 1999) also found that the mental health problems associated with father's drinking had a negative impact on parental attitudes to children and particularly the father's interaction with the child.

Conclusions and recommendations

This is a complex area, and for reliable conclusions to be made there would need to be a further review of the literature. This should take into account the various possible relationships between substance use and mental health problems for both adult and child. From this brief review of the abstracts, it is evident that there is a range of ways the two issues can relate and this has resulted in no clear messages about the impact on children and the family. However, the findings show how such levels of complexity can have negative effects on the children but that these are often mediated or exacerbated by environmental factors. This again suggests that holistic approaches to intervention are needed that address the many variables affecting the family rather than focusing on just the mental health or substance misuse alone.

Interventions / Service Provision

The response to substance misuse remains largely focused on the individuals who are misusing alcohol or drugs. There is growing evidence, however, that services and interventions which have been found to be helpful pay attention to a number of factors including: the family and social context, engagement, support, communication, therapeutic and educational support, and being needs responsive (for example, sessions out of hours, child care, transport and home visits). There is evidence of the effectiveness of a range of ways of working with families affected by substance misuse though further work is needed in this area (Copello, Velleman & Templeton, 2005; Barnard & McKeganey, 2004). Services and interventions tend to be American, not always rigorously evaluated, use relatively small sample sizes, do not always differentiate between alcohol and drugs and are often resource intensive. There is therefore a need for further work and discussion in this area, with a particular focus on how responses could be adapted and transferred to the UK (and to Scotland).

Engagement, of both substance misusers and their families (including children) is paramount. A recent national evaluation of pump-priming drug prevention projects for vulnerable young people (University of Glasgow and the Department of Health, 2004) produced two fact sheets, one on drug prevention with parents and carers of drugs users and the other on drug prevention with children of drug using parents. Important issues when working with parents and carers include: mothers are more likely to access support; there needs to be a clear referral process for getting help, home visits are useful, support groups must accept that attendance will fluctuate thus emphasising the need for flexibility, and staff training and good links with others are important. Finally, GPs are often the first point of contact for many family members but GPs often do not feel well enough equipped to be able to respond. Drug prevention with children of drug using parents must consider the need for specialist skills / knowledge, have clear protocols for the work, both within the organisation but also with other agencies, understand the fear of social services held by many children and parents, offer help with transport where possible, and generally consider less structured work ( i.e. work responsively and proactively) with clear boundaries and reassurance for parents on issues of safety and confidentiality.

Interventions or services with children and families can be broadly summarised into four categories:

1. Those working with whole families (usually at least one parent and at least one child);
2. Those focused on working with children;
3. Those focused on adult family members; and
4. Those more focused on complex situations, such as women who are pregnant, children who also live with domestic violence or who are from black and minority ethnic groups.

Interventions that are family focused

Many evaluated programmes come from the USA, and emphasise that the family is a substantial resource for healing and recovery. With a particular focus on work with severely fractured families, most of these programmes attempt to keep children connected with their own families, even when circumstances prevent them from living together. Where family preservation is not possible, it is seen to be of critical importance to utilise strategies to keep families involved in the treatment process, to reinforce contact between parents and children. Many studies argue for the development of family preservation programmes and family-oriented substance misuse treatment programmes. Others ( e.g. Velleman, Templeton & Copello, 2005; Copello, Velleman & Templeton, 2005) have reviewed the benefits of responding in a family focused way. Overall, however, it appears that there is a lack of focus on the children who attend such family services or programmes, on their experiences and outcomes.

One of the most widely known, and well evaluated, programmes is the Strengthening Families Programme ( SFP). A systematic review of primary prevention programmes for alcohol misuse in young people (Foxcroft et al., 2003), concluded that this was the only programme that could demonstrate effectiveness, maintained longer-term. SFP is a community based, primarily drug and alcohol prevention programme, which combines family and child work, focused on factors of risk and protection, in a series of parallel and combined sessions over a number of weeks. Evidence of effectiveness has been demonstrated across a range of groups and settings. SFP is currently being adapted and replicated in North England (see Velleman, Templeton & Copello, 2005, particularly pages 101-102, for more detail and references).

Another approach that has been tested is the Focus on Families programme (see the Resilience section for more detail and references). One important factor to be highlighted here is the benefits of combining clinical, therapeutic work, with home visits.

Interventions with children

This is a key area for further attention. Most of the work that has been done has focused on prevention initiatives (the work of Cuijpers is useful here), or on American work that is biased towards 'children of alcoholics' or 'adult children of alcoholics'. Barnard (2001) reviewed interventions for drug dependent parents and their children, concluding that working in families where there are younger children brings higher rates of success, and that home-based interventions, peer support, work through schools, community based schemes and play based schemes all have potential. She also highlights the potential for engagement and continued work with parents and children identified through methadone maintenance clinics.

The abstracts reviewed support the general conclusion that a range of interventions is beneficial, including school-based programs, play therapy, social support development and group therapy. Emshoff & Price (1999) suggest that information, coping skills (emotion focused and problem-solving) and support (social and emotional) are key components for working with children of parents with alcohol or drug problems. Emshoff & Jacobus (2001) report that play therapy can be both a relief for children and can also reduce risk. A main element of the experience of having a parent with an alcohol or drug problem, reported in the general literature, is 'loss of childhood', demonstrated by children who miss out on opportunities to play, either because of a lack of positive parenting or because the child has to look after parents or siblings (parentification). Banwell, Denton & Bammer (2002) [summarised in Copello, Velleman & Templeton, 2005], summarise six challenges to be overcome when working with children: 1) getting the balance right between intervention and trust; 2) location; 3) staff support; 4) multi-agency collaboration; 5) funding (including for evaluation); and 6) the need for flexibility.

Recent developments in working with children affected by parental substance misuse in the UK have paid much more attention to the needs of the child, and of working with the whole family where possible, but where the child is at the centre. Linked to this is the increase in thinking and practical work that considers the importance for, and development of, resilience (Velleman & Templeton, 2005 provides a useful summary and references). Two examples in England and Wales are the Family Alcohol Service and Option 2. The Family Alcohol Service, a joint initiative between the NSPCC and the London-wide Alcohol Recovery Project, is a multi-disciplinary team that combines alcohol and family work. The service works centrally with the children but also with whomever else in the family wishes to engage (including the misuser). A report of the evaluation of the pilot year of this service discusses the project's success, but also some of the key challenges to have emerged when working in this way (Velleman et al., 2003). Option 2 is a short-term but intensive programme of work that engages with a family at a point of crisis, usually when a child is at serious risk of being removed from the family (see Hamer, 2005 for a detailed description of the Option 2 way of working). Both Option 2 and the Family Alcohol Service have the development and maintenance of resilience as a theoretical foundation to their practice. The authors of this current review are aware of a few other services in England that are similarly grounded in this family focused way of thinking and working. In the light of the principles of Every Child Matters (2003), with its focus on improving opportunities and outcomes for all children, and the contradictory lack of focus on children and the family within some substance misuse (particularly alcohol) policy, these developments are important and significant.

Baker & Cunningham (2004) list ten strategies to consider when working on parenting issues. They were developed for children living with violence, but could be more widely considered for parental substance misuse.

1. Positive role modelling
2. Clear expectations
3. Praise good behaviour
4. Focus on behaviour not qualities of child
5. Explanation for requests
6. Avoid emotional reactions and yelling
7. Givens and choices
8. Reasonable expectations
9. Boundaries around adult matters
10. Spending time with the children

Interventions with parents / adult family members

Copello, Velleman & Templeton's review of family interventions (2005), summarise (adult) family focused interventions as: 1) acting as a mechanism for the entry and engagement of substance misusers into the treatment system; 2) working jointly with substance misusers and (usually, adult) family members; and 3) responding to the needs of family members in their own right. All three areas demonstrate evidence of effectiveness though the authors (and others, e.g. Barnard & McKeganey, 2004) argue that further work is needed.

Interventions that act as a mechanism for the entry and engagement of substance misusers into the treatment system have been shown to be effective. The most well known examples are the Australian Pressures to Change approach (Barber & Crisp, 1995), and the American programmes of CRAFT (Community Reinforcement and Family Training [Smith & Meyers, 2004] and ARISE (A Relational Intervention Sequence for Engagement [Garrett et al., 1998]).

Joint work with substance misusers and (usually adult) family members is based on the demonstration that, " attention to the person's social context and support system is prominent among several of the most supported approaches" (Miller & Wilbourne, 2002 p276). Miller & Wilbourne's major review (2002) demonstrated that several of the most effective treatments for alcohol problems included attention to social context. Behavioural, couples and marital work, social skills training and the community reinforcement approach are all good examples. A growing area of interest is in network approaches. Most recently tested is Social Behaviour and Network Therapy ( SBNT; Copello et al., 2005; Copello et al., 2002), developed and tested as one of two treatments in the UK Alcohol Treatment Trial, an amalgamation of key elements of several other, socially grounded, treatments. A key finding from UKATT was that SBNT demonstrated effectiveness on a par with its comparison treatment of Motivational Enhancement Therapy, and that led to improvements in drinking behaviour and associated problems ( UKATT Research Team, 2005).

Other recent work has focused specifically on responding to the needs of family members in their own right, by testing a brief intervention treatment package, to be delivered by primary health care professionals to family members (Copello et al., 2000). This is a coherent package that can respond to family members' identified needs in the primary care environment; it could be rolled out across Scotland. Further work has tested a self-help version of the intervention, and has also tested feasibility in specialist drug and alcohol teams (Templeton, Zohhadi & Velleman, 2004); and this work further demonstrates the potential for further consideration.

An additional and important area to mention is that of self-help support groups and interventions. Al-Anon (alcohol) and Adfam (drugs) are the most well known of these. Whilst rigorous evaluation of such anonymous organisations is difficult, membership numbers and anecdotal evidence are testimony of their reach, significance and benefits

As with the studies relating to children, the evidence therefore suggests that a range of interventions is beneficial to adult family members (including parents), and a range of positive outcomes can be demonstrated. Several studies report a lack of hypothesised differences between treatments, something that has also been noted in other comparative, trial-based studies of interventions with alcohol misusers and/or their families ( e.g. Project MATCH, a major alcohol treatment trial in the USA [Project MATCH Research Group, 1997]; UKATT, the UK Alcohol Treatment Trial [ UKATT Research Team, 2005], and a brief intervention delivered in the primary care setting in England [Copello et al., 2000]). Figlie et al. (2002) note that, "there is not a consensus about the type of treatment to be used" (p327) which, when combined with the point just made, suggests that factors other than the nature of the treatment or intervention are equally important for engagement, retention, positive change and maintenance of change. This needs further exploration with regard to children and family members.

Interventions that respond to complexity

There are several areas that will be considered here, first of which is working with pregnant women or mothers with babies / young children. There is a lot of work in this area, most of it again from the USA, and referring mainly to working with women who are, with or without children being part of the intervention, often involving a study within a residential treatment facility (usually USA studies), thus bringing implications of resources. Whilst acknowledging the potential for, and sometimes the benefits of, interventions (usually undertaken within these studies within residential facilities) for mothers / pregnant women, there is a lack of research that focuses on the short- and long-term outcomes for babies and children. Parenting is clearly a key issue for consideration here, both in terms of the low levels of parenting skills reported by the women in these studies and hence in others who use such treatment facilities, and also in terms of parenting skills training and development needing to be a focus of the help that women, and their children, receive. There are some examples of residential facilities that cater for the needs of mothers and children in the UK ( e.g. the Aberlour Child Care Trust in Scotland and Phoenix House in England) but further work is needed. This work must also consider the needs of, and role of, other members of the family and network, for example, fathers, siblings and grandparents.

Second, are the particular needs of children affected by parental substance misuse who also live with domestic violence. As already highlighted in this report, this is an area where further work is needed at all levels. This review has repeatedly highlighted the presence of domestic violence where there is substance misuse. Service response and family interventions need to think carefully about how family work can be made safe where domestic violence exists. Current evidence suggests that many network-based therapies have failed to screen for, or consider, domestic violence prior to starting family intervention, placing women and children at potentially greater risk. Any such change will involve training staff in responding to domestic violence and ensuring referral processes screen for domestic abuse. The London based Stella Project has been specifically established to provide services with training and policy guidance on the overlap between substance misuse and domestic violence. There are no known examples of services that cater specifically for children and families affected by both substance misuse and domestic violence, though some separate services (substance misuse or domestic violence) are developing joint-working protocols and providing mutual support to improve their service delivery, for example, the Family Alcohol Service in London and the Nia Project in London (formerly Hackney Women's Aid).

Finally, it is important to recognise the needs of children from black and minority ethnic groups living with substance misuse. The STARS Initiative in Nottingham, England, is one service that has considered the particular needs of these groups of children, though primarily in the area of child protection placements (Mayer, 2004). There is a need for further work in this area, to understand whether children from different backgrounds and cultures have different experiences and needs, and to develop culturally sensitive services that can maintain engagement.

Conclusions

"…children have only rarely been the direct focus of intervention, with the assumption that they will benefit from the support offered to their parents…..[require] interventions that are equally cognizant of children's perspectives and needs…..Interventions have also overlooked the significance of the extended family…..[the] same point applies to fathers or partners with drug problems" (Barnard & McKeganey, 2004 p557).

• There is a clear and definable need for further work in this area, particularly with other family members, especially with fathers and with the children themselves. Copello, Velleman & Templeton (2005) say that future work in this area needs to focus on: "….1) pragmatic trials that are more representative of routine clinical settings; 2) cost-effectiveness analyses…..3) explore treatment process; and 4) make use of qualitative methods" (p369).
• Given the amount of work in this area that comes from the USA, with its focus on the medical model, abstinence and associated terminology, discussion is needed on the transferability and adaptability of interventions and services to the UK / Scotland.
• Barnard & McKeganey (2004) say that those services and interventions that exist, and which have had their benefits demonstrated, tend to be localised. One of the conclusions of Murray's (2003) assessment of the prevalence of children of substance misusers was that, "there are clear gaps in the delivery of local services that must be addressed to comply with national guidelines" (p4). The expansion and generalisability of such services, for example across the statutory sector, is a challenge for the future, and one to be considered as a recommendation from this review.
• Indications are that some services / programmes exist, but that solid research and evaluation, particularly focusing on children and their experiences / outcomes is lacking.
• In developing services for children and families, there are related training needs in working with children, in responding to complexity and in working with other agencies in a safe, ethical and helpful way. There is longstanding evidence from other areas of research that staff will only undertake new and potentially challenging work if they are both adequately trained and adequately supported in these new roles.
• Continued consideration should be given to issues of resilience and how factors likely to increase resilience could be potentially included as elements of intervention. Whilst the general resilience literature is sizeable, work on its practical application is particularly lacking (Newman, 2004), and this recommendation can be applied to children affected by parental substance misuse.