Interventions to build resilience in children of problem drinkers

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows: To assess the eFects of interventions for building resilience in children or young people living with parents/carers who are problem drinkers. Interventions to build resilience in children of problem drinkers (Protocol) Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. 1 Cochrane Library Trusted evidence. Informed decisions. Better health. Cochrane Database of Systematic Reviews


Description of the condition
Problem drinking by parents or carers is a significant social problem and varying prevalence rates across race and ethnic groups indicate that some children and young people may be at greater risk than others. Problem drinking by parents may be determined by environmental risk and prognostic factors such as cultural attitudes toward drinking and intoxication, the availability and price of alcohol, and stress levels . In the USA, an average of 7.5 million children under the age of 18 years are estimated to live with a parent who had an alcohol use disorder in the past year, equating to 10.5% of all children (SAMHSA 2012). Nine million children and young people in the European Union (EU) are estimated to live with at least one parent addicted to alcohol (Eurocare 2012). In Australia, 13.2% of children are estimated to be at risk of exposure to shortterm risky drinking by at least one adult in households (Dawe 2007) whereas data from the UK (in 2000) suggest that 22% (2.6 million) of children (aged 16 years and under) live with a hazardous drinker, 6% (705,000) with a dependent drinker and 3% (300,000) with a harmful drinker (Manning 2009).

Alcohol use disorders
Alcohol use disorders are typically defined using the Diagnostic and Statistical Manual of Mental Disorders (DSM), 5th edition (DSM-5) (DSM-5) and the International Classification of Diseases and Related Disorders, 10th edition (ICD-10) (WHO 2010), which categorise the severity of an individual's alcohol consumption in relation to a spectrum of distinct drinking patterns. Harmful use refers to a pattern of alcohol use that is causing damage to an individual's physical or mental health (WHO 2010). Alcohol dependence is defined by ICD-10 as "a cluster of behavioural, cognitive and physiological phenomena that develop a er repeated use and that typically include a strong desire to take the drug, di iculties in controlling its use, persisting in its use despite harmful consequences, a higher priority given to drug use than to other activities and obligations, increased tolerance, and sometimes a physical withdrawal state" (WHO 2010).The text revision of the fourth edition of the DSM (DSM-IV-TR) contained seperate diagnoses for alcohol abuse and dependence, which are combined in the recently published . Criteria are provided for alcohol use disorder, accompanied by criteria for intoxication, withdrawal, other alcohol-induced disorders and unspecified alcohol-related disorders. An alcohol use disorder is defined as a problematic pattern of alcohol use leading to clinically significant impairment or distress, as manifested by at least two of the following, occurring within a 12-month period: alcohol is o en taken in larger amounts or over a longer period than was intended; there is a persistent desire or unsuccessful e orts to cut down or control alcohol use; a great deal of time is spent in activities necessary to obtain alcohol, use alcohol or recover from its e ects; craving or a strong desire to use alcohol; recurrent alcohol use resulting in a failure to fulfill major role obligations at work, school or home; continued alcohol use despite having persistent or recurrent social or interpersonal problems caused or exacerbated by the e ects of alcohol; important social, occupational or recreational activities are given up or reduced because of alcohol use; recurrent alcohol use in situations in which it is physically hazardous; alcohol use is continued despite knowledge of having a persistent or recurrent physical or psychological problem that is likely to have been caused or exacerbated by alcohol; tolerance (a need for markedly increased amounts of alcohol to achieve intoxication or desired e ect; or a markedly diminished e ect with continued use of the same amount of alcohol) and withdrawal (the characteristic withdrawal syndrome for alcohol; or alcohol is taken to relieve or avoid withdrawal symptoms). In addition to the above, a diagnostic term which does not appear in either the ICD-10 or the DSM-5 but which o en appears in the extant literature is hazardous (risky) drinking, a pattern of alcohol consumption that increases the risk of harmful consequences for the user or others (Babor 2001).

Problem drinking
When considering the e ects of parental problem drinking on children, it is not appropriate to restrict one's definition solely to the diagnostic categories of the DSM-5 (DSM-5) or the ICD-10 (WHO 2010) for a number of reasons. Neither treatment agency records nor clinical studies accurately reflect the 'hidden' nature of problem drinking (e.g. Chalder 2006;Christensen 2000;Colder 1997;Obot 2004;Orford 1990;Orford 2003;Sher 1991), and studies of clinical samples may overestimate pathology by focusing on more severely impaired individuals (Chassin 1999). Clinical diagnoses do not encompass those who have not acknowledged or who are unaware that their drinking is problematic and are undiagnosed. In addition, parents may be less willing to enter treatment than non-parents due to a fear that their children may be taken into care by social services (Powis 2000). Parents who are not willing to participate in programmes may, nevertheless, endorse the benefits their children receive from a prevention programme (Broning 2012). With this in mind, this review uses the broader definition of 'problem drinking' to include any form of hazardous drinking, harmful drinking, alcohol dependence and alcohol use disorders by carers.

Impact on children
The extant literature on the impact of parental alcohol use on child outcomes reflects the cultural nature of the problem, with the vast majority of studies emerging from the USA, Europe (e.g. UK and Ireland) and Australia. Children depend on their families to meet their physical, psychological and social needs, their economic security and well-being, all of which can be jeopardised by parents misusing substances (NACD 2011). Children living with parental alcohol problems o en: have feelings of guilt, loneliness, anger, worry and uncertainty (Cuijpers 2005, Velleman 1990) or low self-esteem (Rangarajan 2008); experience disharmonious family environments and family instability (Velleman 1990;Velleman 1993a;Velleman 1993b;Velleman 1999); are at risk of higher levels of childhood di iculties (Velleman 1993a; Velleman 1993b); or may have to adopt adult roles (Velleman 1990). Velleman 1992b reported that the negative e ects of having a problem-drinking parent are mediated via family disharmony, and that in the absence of such disharmony the o spring will not be damaged. These children are also at risk of developing a range of negative health outcomes (Passon 2009) due to dysfunctional coping strategies including alcohol and drug problems (Copello 2005;Corte 2007;Schor 1996;Velleman 1992a), and emotional and behavioural problems (Copello 2005) such as depression or anxiety disorders (Velleman 2004;Velleman 2007). Furthermore, it is common for children of alcoholics to have misunderstandings about alcoholism (Emsho 1999).
It is not unusual for families to contain multiple problem drinkers (Percy 2008) and an increased risk of alcohol problems in adulthood has been associated with having had two parents with drinking problems (Orford 1990). Maternal drinking problems frequently occur in the context of paternal drinking problems and maternal drinking may actually be a proxy for "two alcoholic parents" (Keller 2005;Keller 2008). Variations in the role of maternal and paternal alcoholism in predicting outcomes amongst children have been reported to confer di erent risks according to the gender of the o spring (Balsa 2009;Corte 2007). Males have higher rates of psychosocial problems in childhood and adolescence (Werner 1986) as do the o spring of alcoholic mothers (Werner 1986) who tend to report more negative childhood experiences (Velleman 1990). Females generally lean towards withdrawal and social isolation, whereas males more o en display antisocial behaviour (Velleman 2003).
Despite the evidence of increased risks to children living with problem drinkers, a large body of evidence suggests that most are remarkably well-adjusted (Sher 1991) or resilient (Velleman 1999). Resilience has been variously defined as "an interactive concept that is concerned with the combination of serious risk experiences and a relatively positive psychological outcome despite those experiences" (Rutter 2006) and "good adaptation under extenuating circumstances and, from a developmental perspective, meeting age salient developmental tasks in spite of serious threats to development" (Masten 2002). Resilient children are those who display an ability to bounce back from tough times, or have the capacity to overcome challenge or adversity. Resilience is not a fixed trait of individual children; rather, it is best viewed as an interactive concept concerned with the interplay of environmental threats or risk and relatively combined with relatively good outcomes. Velleman 2007 identified a range of factors that provide evidence of resilience in children of substance misusers, including: deliberate planning by the child that their adult life will be di erent; high self-esteem and confidence; selfe icacy; an ability to deal with change; skills and values that lead to good use of personal ability; a good range of problem-solving skills; feeling that there are choices; feeling in control of their own life; and previous experience of success and achievement. Research suggests that having a supportive adult or confidant, either within the immediate family (if only one parent has a drink problem) or in the extended family and beyond, can help to build resilience, encouraging the development of functional coping behaviours. The social support provided by these significant adults (e.g. relatives, teachers) can help to alleviate the risk of developing maladaptive coping strategies (such as drug and alcohol use) and the onset of emotional or mental health problems.

Description of the intervention
This review focuses on interventions designed to build resilience in children living with a problem-drinking parent. Despite an increase in the development and evaluation of services and interventions targeting these children (Templeton 2010), few theory-driven programmes of prevention or intervention have been developed (Cuijpers 2005). Typically, those programmes that have been developed share some common components (see Cuijpers 2005; Emsho 1999). A key component of interventions is to assist young people in developing skills to cope with a parent's drinking. Coping skills may take various forms. They may be emotion focused, problem focused or may prepare young people to actively seek help or social support. Interventions designed to promote emotion-focused coping skills seek to help young people to identify and discuss their feelings (e.g. feelings of sadness or distrust, or worry about their parents) and adaptively manage their feelings using various strategies (e.g. relaxation, playing a game, listening to music). Those focusing on problem-solving skills consist of teaching participants skills on how to deal with alcoholrelated problems (e.g. appraising the situation, determining the problem and the desired solution, trying alternative solutions, knowing how to react when a parent is drunk, dealing with conflict in the home, explaining the situation to friends). Social skills or support-seeking skills can assist a young person with helpseeking behaviour (e.g. talking to a teacher or other professional). Many programmes are group-based. Group-based programmes provide children and young people with opportunities to share common experiences (reducing feelings of isolation and personal responsibility); find social and emotional support (including from a significant adult role model); exchange experiences; and learn problem-solving skills. Most interventions, whether individualor group-based, provide information or education on alcohol use, problem drinking, the consequences of drinking and other terminology (e.g. tolerance, blackouts, withdrawal) so that the child may understand the behaviour of the parent. Selective interventions designed specifically for children of problem drinkers that contain any/all of the above components will be included in the review.

How the intervention might work
Hypotheses as to how these interventions may prevent or attenuate the development of risk factors associated with a parent/carer's problem drinking vary. Social learning theorists suggest a modelling e ect whereby a person learns from the example of another (Bandura 1977). The most e ective models appear to be those individuals who command respect (perhaps because they are well known, or someone who is admired or powerful) but at the same time are not too dissimilar to ourselves. Interventions based on this theory of change, which use positive role models to promote problem solving or the use of social and emotional support, may be beneficial. Bandura 1986 regarded behaviour as an outcome of "reciprocal determinism" (i.e. the mutual interaction of the person's internal characteristics, the environment and behaviour itself). In a similar vein, resilience theory states that the development of resilience is based on three factors: the individual attributes of children or young people, aspects of their families, and the characteristics of their wider social environments (Garmezy 1984;Luthar 2000;Luthar 2003;Rutter 1979, Rutter 1987Rutter 2008;Werner 1982;Werner 1992;Werner 2001). As resilience is the product of an interaction between the individual and their social context, it is potentially open to influence (Velleman 2007). Therefore, the identification of a set of potentially modifiable risk and protective factors could form the basis of programmes designed to build resilience. Studies of resilient children from substance-a ected homes suggest these children are more likely to have social resources outside the home, such as other competent adults or caring relatives (McCabe 1999;). An involvement in a range of stabilising activities, such as school, clubs, sports, religion, also appears to be beneficial in helping a young person to develop a sense of self and self-esteem (Velleman 2007). Resilient children exhibit traits such as high stress resistance, good adaption skills in new situations and high selfe icacy (Velleman 2007). Broning 2012 suggest that children's development is influenced by their own cognitive appraisal of life with a substance-abusing parent as well as by their emotional and behavioural strategies of coping with the di icult situations that arise from parental substance use. In line with resilience theory, programmes that focus on providing access to social resources or stabilising activities outside the home environment, and encourage the development of problem-and emotion-focused coping skills in children may be a promising form of intervention. For example, the 5-Step Method (Copello 2000a, Copello 2000b was developed based on the stress-strain-coping-support model (SSCS) (Orford 1998;Orford 2005;Velleman 2003). An important principle of the model is that living with a highly stressful experience such as the impact of an addiction problem in the family, may lead to psychological and physical symptoms of ill health in family members other than the substance user (Copello 2010). In addition to the methods outlined earlier, this approach may benefit children by reducing the levels of stress they experience, and the provision of relevant information may help the child to understand concepts (such as 'tolerance,' 'blackouts' and 'withdrawal') and the parent's behaviour, helping to reduce self-blame and guilt about parental drinking (Emsho 1999). While these theoretical underpinnings outline how the intervention may work, it is important to note that demographic variables may have an impact on intervention outcomes; for example, 'Gance-Cleveland 2008 reported that girls in the intervention group demonstrated improved resilience measured by increased coping compared to the control group. The authors suggested that adolescent boys and girls are frequently at di erent stages in their development with girls tending to be more socially integrated and mature than boys.

Why it is important to do this review
Practitioners will benefit from the review via the provision of evidence on which interventions are most likely to attenuate the adverse impact of this social problem on the next generation. The review will also be of timely relevance for policy makers. An EU strategy to support Member States in reducing alcoholrelated harm (CEMC 2006) outlined an aim to reduce the harm su ered by children in families with alcohol problems. The high prevalence of problem alcohol use by parents and its adverse consequences for their children (see 'Impact on children'), reinforce the importance of knowing how best to support children whose parents are problem drinkers. These children are at risk of a number of long-term consequences including alcohol and drug problems (Copello 2005; Corte 2007), mental health or behavioural problems (Copello 2005, Velleman 2007), early pregnancies (Werner 1993), di iculties forming relationships as adults (Kelley 2010), academic underachievement (Torvik 2011) and dysfunctional coping strategies (Schor 1996). To date, only a couple of reviews have investigated the e ectiveness of interventions in children of problem drinkers in an attempt to understand how to prevent these long-term consequences. A review by Templeton 2010 focused on psychological interventions for families (including spouses, partners and children) a ected by alcohol misuse. However, a broad review methodology was used allowing for the inclusion of randomised and quasi-randomised study designs, comparative studies (including control or cohort), other observational studies (case series, time series, before and a er studies) and qualitative studies, with the results analysed using a thematic, narrative approach. Furthermore, the review considered a broad range of outcomes (any physical, psychological or other outcome) and age groups (children and adult children of alcoholics). A systematic review by Broning 2012 considered programmes for children aged 0 to 17 years from substancea ected families. In contrast to the current protocol, their review focused on substance-a ected families (as opposed to alcohola ected families only), included studies with a range of designs (randomised controlled trials (RCTs), and controlled or quasiexperimental, descriptive and qualitative studies) and had a number of date and language restrictions (the authors searched for studies over a 15-year period (1994 to 2009) that were published in English or German).

O B J E C T I V E S
To assess the e ects of interventions for building resilience in children or young people living with parents/carers who are problem drinkers.

Types of studies
We will include RCTs, CCTs and prospective controlled observational studies.

Types of participants
Children and adolescents aged 4 years to 18 years and 11 months. The majority of the sample in any included study must fall within this age range and must meet the following inclusion criteria: • have at least one primary carer (biological, step, foster, adoptive parent/carer, grandparent) who currently meets (or in the recent past met) the criteria for problem alcohol use (i.e. hazardous or harmful drinking, alcohol dependence or alcohol abuse). While diagnosis may be based on DSM or ICD criteria, in order to address the 'hidden nature' of the problem, parental drinking may have been diagnosed/assessed in a range of ways, including: by a clinician (e.g. parents who are in contact with alcohol-treatment centres); via parental self-report (e.g. Alcohol use Disorders Identification Test (AUDIT)) (Babor 2001); via child reports of parental drinking (e.g. Children of Alcoholics Screeing Test (CAST)) (Jones 1983; Pilat 1984); or by a third party (e.g. a therapist); • live with the parent/carer or at least have had regular contact with them (visitation) in the case of parents/carers living apart.

Types of interventions
Any face-to-face (either individual-or group-based) intervention, targeted at children (or both carers and children), designed to enhance resilience, irrespective of duration, intensity or frequency, and delivered in any setting (schools, home, community) by either professionals or non-professionals, compared with no intervention, wait-list control or standard care. We will exclude guided self-help interventions, or those delivered via the internet, compact disc or other media.

Types of outcome measures
Primary and secondary outcome measures may be child, parent or teacher self-report measures completed at the end of the intervention and at one year follow up. While a wide range of outcomes can be considered in determining resilience in children living with problem-drinking parents (see Velleman 2007), this review will focus on the following primary and secondary outcomes.

Primary outcomes
• Coping behaviours,* as measured by improvements in the child's emotion-focused strategies, problem-solving strategies and social support-seeking strategies. Examples of measures include the Coping Responses Inventory (Moos 1993; Moos 2004) and the Coping Strategies Inventory (Wills 1985) Secondary outcomes • Social support,* as measured by an improvement in the child's level of feeling supported (from groups or significant individuals) • Knowledge of alcohol use,* as measured by an increase in the child's knowledge of alcohol use (e.g. information and terminology) • Self-esteem, as measured by an improvement in the child's selfesteem scores • Emotional or behavioural problems,* as measured by an improvement in scores on mental health measures (e.g. Strengths and Di iculties Questionnaire (Goodman 1997) and the Children's Depression Inventory (Kovacs 1985) • Use of alcohol or drugs, or both,* as measured by the number of participants who do not initiate/become involved/or do not progress to problematic levels in their use of alcohol or drugs, or both (assessed, e.g., using the Alcohol Use Disorders Identification Test) (Babor 2001) • Self-e icacy • Psychological well being • High risk behaviour • Quality of life • Social functioning We will use outcomes indicated by an asterix (*) to populate the 'Summary of findings' table for the main comparison, where data permit. Where data are insu icient, we will provide a narrative account of the outcomes.

Search methods for identification of studies Electronic searches
We will identify relevant trials from the following sources.

Electronic bibliographic databases:
• We will use no language or date restrictions in the search for trials. We will use a combination of Medical Subject Headings terms and free text terms relating to problematic alcohol use (by parents/carers). We will develop detailed search strategies for each database, accounting for di erences in controlled vocabulary and syntax rules. We show the search strategy for MEDLINE in Appendix 1.

Electronic grey literature databases:
• Dissertation Abstracts; • Index to Theses.
3. Targeted searches of websites: a structured search will also be applied to the following web resources: •

Searching other resources
We will also search conference proceedings likely to contain trials relevant to the review. References from obtained reviews and web resources will be examined to identify related resources. We will contact authors and researchers to source unpublished or incomplete studies. We will conduct a structured handsearching of relevant journals and include those not covered by The Cochrane Library. We will include non-English language studies in all searches and will assess all non-English abstracts for inclusion. When considered likely to meet inclusion criteria, we will translate non-English full studies.

Selection of studies
Two review authors (AMcL, NL) will independently inspect all titles or abstracts, or both, extracted from the searches. We will discard any clearly irrelevant studies. We will resolve any disagreements in the selection of studies at the initial selection stage by discussion. We will retrieve full copies of all potentially relevant studies. Two review authors (AMcL, NL) will independently determine whether studies should be included, excluded or classified as unclear.
We will resolve any di erences in opinion via a reviewer team discussion.

Data extraction and management
Data will be extracted independently by two review authors (AMcL, MMcC) using a standard data extraction form. The form will include the following domains: study design and method, allocation process, participant data, intervention and outcomes.

Library
Trusted evidence. Informed decisions. Better health.

Cochrane Database of Systematic Reviews
We will discuss any disagreements, which if not resolved will be referred to a third reviewer (GM). If necessary, we will seek additional information from the study authors. We will collect information on study design and implementation in a format that will enable us to populate the 'Risk of bias' tables in the completed review.

Assessment of risk of bias in included studies
Two review authors (AMcL, NL) will independently assess each study for risk of bias. We will undertake the 'Risk of bias' assessment for RCTs and CCTs in this review using criteria recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The recommended approach for assessing risk of bias in studies included in a Cochrane Review is a two-part tool, addressing seven specific domains: sequence generation and allocation concealment (selection bias); blinding of participants and providers (performance bias); blinding of outcome assessor (detection bias); incomplete outcome data (attrition bias); selective outcome reporting (reporting bias); and other sources of bias. The first part of the tool involves describing what was reported to have happened in the study. The second part of the tool involves assigning a judgement relating to the risk of bias for that entry, in terms of low, high or unclear risk. To make these judgements we will use the criteria indicated by the Cochrane Handbook for Systematic Reviews of Interventions adapted to the addiction field. See Appendix 2 for details.
We will address the domains of sequence generation and allocation concealment (avoidance of selection bias) in the tool using a single entry for each study.
We will consider the blinding of participants, personnel and outcome assessors (avoidance of performance bias and detection bias) separately for objective outcomes (e.g. dropouts) and for subjective outcomes (e.g. social functioning as integration at school or at work, family relationships).
The criteria drawn from the Newcastle-Ottawa Scale (NOS) will be used to assess observational studies. The 'Risk of bias' tables will be operationalised to be used for the assessment of RCTs, CCTs and prospective observational studies, according to the criteria recommended by the Cochrane Drugs and Alcohol Review Group. See Appendix 2 for details.

Measures of treatment e ect
We will calculate unadjusted treatment e ects using Review Manager (RevMan) 2012 where possible.

Dichotomous outcome data
Dichotomous outcomes will be analysed by calculating the relative risk (RR) for each trial, with the uncertainty in each result expressed using 95% confidence intervals (CIs).

Continuous outcome data
Continuous outcomes will be analysed by calculating mean di erences (MDs) if all studies use the same measurement scale, or standardised mean di erences (SMDs) if studies use di erent measurement scales, each with 95% CIs.

Unit of analysis issues
Cluster randomised trials are possible in this area of research, as allocation to the intervention group may occur by school or community. It is recognised that cluster designs are susceptible to unit-of-analysis error and that P values may be artificially small (Higgins 2008). We anticipate that investigators will have controlled for a clustering e ect when presenting their results.
Where the clustering e ect has not been controlled for, we will request participant data to calculate an estimate of the intracluster correlation coe icient (ICC). If participant data are not available we will search for external estimates of the ICC from similar studies.

Dealing with missing data
We will contact authors to follow up missing outcome data, missing summary data or missing study level characteristics for subgroup analyses. If standard deviations are missing from continuous data, we will scan studies for any other statistics (CIs, standard errors, T values, P values, F values) that allow for its calculation. We will describe missing data and all forms of attrition for each included study in the 'Risk of bias' table, and discuss the extent to which missing data could impact on the conclusions of the review. Missing data will be treated according to whether data is 'missing at random' or 'not missing at random'. In relation to the former, the main option will be to analyse the available data and ignore the missing data.
It is possible that missing data may not be missing at random. For example, if a participant did not experience any positive outcomes from the intervention they may be more likely to drop out of the intervention programme, or to fail to return and complete all necessary follow-up assessments. In the event that data are not missing at random, we will use replacement values to impute the missing data. When imputing missing dichotomous data, we will assume that missing data are negative (e.g. the participant demonstrated high risk behaviour). When imputing missing continuous data, we will use a 'last observation carried forward' approach.
Some relevant studies may fail to provide summary data (e.g. standard deviations). Where this occurs we will, if possible, obtain these data using calculations provided in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

Assessment of heterogeneity
We will examine clinical heterogeneity by inspecting each included study for variability in the participants, interventions or outcomes described. If unexpected variability arises, we will discuss this in full in the review. Methodological heterogeneity will be examined by inspecting each included study for variability in the study design and risk of bias. If any unexpected variability arises, we will discuss this in full in the review. We will assess statistical heterogeneity using the Chi 2 test and its P value, by visual inspection of the forest plots and the I 2 statistic. A P value of the test lower than 0.10 or an I 2 statistic of at least 50% will indicate significant statistical heterogeneity.

Assessment of reporting biases
We will use funnel plots (plots of the e ect estimate from each study against the sample size or e ect standard error) to indicate possible publication bias. We acknowledge that asymmetry in the plot could be due to publication bias, but this can also reflect a real relationship between trial size and e ect. We will use tests for funnel plot symmetry only when at least 10 studies are included in the meta-analysis, as a smaller number of studies will render the power of the tests too low to distinguish chance from real asymmetry.

Data synthesis
Regression analyses will be run using Stata (StataCorp 2013) and imported into Review Manager (RevMan) 2012. The outcome measures from the individual trials will be combined through metaanalysis where possible. We plan to synthesise results from studies using an RCT design in a meta-analysis where the interventions are similar with regard to: i) method of delivery (individually/groups); ii) setting (school-, home-or community-based); and iii) intensity, frequency and duration of the programme. We will perform a random-e ects meta-analysis using an inverse variance weighting method using Review Manager (RevMan) 2012 as we expect a certain level of heterogeneity among the included studies. We will perform both fixed-e ect and random-e ects analyses as part of our sensitivity analyses. If some primary studies report an outcome as a dichotomous measure and others use a continuous measure of the same construct, we will convert results for the former from an odds ratio to a SMD, provided that we can assume that the underlying continuous measure has approximately a normal or logistic distribution (otherwise we will carry out two separate analyses). If meta-analysis is not appropriate, we will report results from individual studies. We will provide narrative summaries of the non-randomised studies (NRS).

Subgroup analysis and investigation of heterogeneity
We will explore heterogeneous results by conducting the following subgroup analyses: • participant stage of development (i.e. children (aged 4 years to 12 years and 11 months) and adolescents (aged 13 years to 18 years and 11 months); • gender of child/young person; • gender of parent/carer; • child living with parent/carer OR child living away from parent/ carer/currently in care; • severity of parental drinking.
We will carry out subgroup analyses only if 10 or more studies are retrieved during the data collection process, as it is unlikely that the investigation of heterogeneity will produce useful findings unless there is a substantial number of studies (Higgins 2008).

Sensitivity analysis
If more than 10 studies are included in the analysis we will perform sensitivity analyses. To incorporate quality assessment in the review process we will first plot intervention e ect estimates stratified for risk of bias for each relevant domain. If di erences in results are present among studies at di erent risk of bias, we will then perform a sensitivity analysis, excluding from the analysis studies with a high risk of bias. We will perform subgroup analyses for studies with low and unclear risks of bias.
We will also perform both fixed-e ect and random-e ects analyses as part of our sensitivity analyses.
We will also perform sensitivity analyses by re-analysing the data without imputing data for the missing participants, and by reanalysing the data imputing data for the missing participants assuming they experienced a positive outcome.

Velleman 1992b
Velleman R. Intergenerational e ects-a review of environmentally oriented studies concerning the relationship between parental alcohol problems and family disharmony in the genesis of alcohol and other problems. II: The intergenerational e ects of family disharmony. The International Journal of the Addictions 1992b;27(4):367-89.

Velleman 1993a
Velleman R, Orford J. The importance of family discord in explaining childhood problems in the children of problem drinkers. Addiction Research and Theory 1993a;1:39-57.

Velleman 1993b
Velleman R, Orford J. The adult adjustment of o spring of parents with drinking problems. British Journal of Psychiatry 1993b;162:503-16.

Velleman 1999
Velleman R, Orford J. Risk and Resilience: Adults who were the children of problem drinkers. Oxon: Routledge, 1999.

Velleman 2003
Velleman R, Templeton L. Alcohol, drugs and the family: results from a long running research programme within the UK. European Addiction Research 2003;9:103-12. Cochrane Database of Systematic Reviews 4. Blinding of participants and providers (performance bias) Subjective outcomes

Low risk
Blinding of participants and providers, and unlikely that the blinding could have been broken

High risk
No blinding or incomplete blinding, and the outcome is likely to be influenced by lack of blinding Blinding of key study participants and personnel attempted, but likely that the blinding could have been broken, and the outcome is likely to be influenced by lack of blinding Unclear risk Insufficient information to permit judgement of low or high risk 5. Blinding of outcome assessor (detection bias) Objective outcomes Low risk No blinding of outcome assessment, but the review authors judge that the outcome measurement is not likely to be influenced by lack of blinding Blinding of outcome assessment ensured, and unlikely that the blinding could have been broken

Record linkage
6. Blinding of outcome assessor (detection bias)

Subjective outcomes
Low risk No blinding of outcome assessment, but the review authors judge that the outcome measurement is not likely to be influenced by lack of blinding Blinding of outcome assessment ensured, and unlikely that the blinding could have been broken High risk No blinding of outcome assessment, and the outcome measurement is likely to be influenced by lack of blinding Blinding of outcome assessment, but likely that the blinding could have been broken, and the outcome measurement is likely to be influenced by lack of blinding Unclear risk Insufficient information to permit judgement of low or high risk 7. Incomplete outcome data (attrition bias) For all outcomes except retention in treatment or dropouts

Low risk No missing outcome data
Reasons for missing outcome data unlikely to be related to true outcome (for survival data, censoring unlikely to be introducing bias) Missing outcome data balanced in numbers across intervention groups, with similar reasons for missing data across groups For dichotomous outcome data, the proportion of missing outcomes compared with observed event risk not enough to have a clinically relevant impact on the intervention effect estimate For continuous outcome data, plausible effect size (difference in means or standardised difference in means) among missing outcomes not enough to have a clinically relevant impact on observed effect size Missing data have been imputed using appropriate methods All randomised participants are reported/analysed in the group they were allocated to by randomisation, irrespective of non-compliance and cointerventions (intention to treat) High risk Reason for missing outcome data likely to be related to true outcome, with either imbalance in numbers or reasons for missing data across intervention groups (Continued)