Me & my family: a programme for children and parents in families with parental substance use problems – an outcome study

ABSTRACT

Me & my family is a knowledge- and motivation-based programme that includes eight weekly sessions provided by the social services for families in which there are parental substance use problems (SUPs). The aim of this study was to evaluate the programme outcomes in terms of child and parental well-being, family climate, and parental alcohol and drug use. The study involved self-reported ratings from 59 children, aged 5–20 years, and 67 parents, of whom approximately two-thirds completed the programme and provided the self-ratings before, on completing, and three months after the intervention. Findings included a significant increase in family closeness and a significant reduction in family chaos after the intervention, which persisted three months post-intervention. Children also reported a reduction in self-rated conduct problems, whereas parents reported a similar reduction in self-rated symptoms of anxiety and depression. Parents also reported less alcohol and drug consumption post-intervention relative to baseline ratings. The findings suggest that the Me & my family programme supports child and family well-being and could potentially protect children from later detrimental outcomes. The results need to be replicated to evaluate the effectiveness of this programme compared with other programmes within the social services.

KEYWORDS:

• Family intervention
• parental substance use
• child mental health
• parental mental health

Introduction

Many children grow up in families with parental substance (i.e. alcohol and/or other drug) use problems (SUPs). According to recent statistics, approximately 20% (i.e. 430,000) of children aged 0–18 years in Sweden have been exposed to parental SUPs (Ramstedt 2019). Children who grow up in families with parental SUPs are at increased risk of developing a range of problems, such as SUPs, mental health problems, criminality, dysfunctional relationships, poor school grades, and attention and conduct problems at school (Baldwin et al. 2012; Berg et al. 2016; Calabria, Shakeshaft, and Havard 2011; Díaz et al. 2008; Folkhälsomyndigheten 2016; Jääskeläinen et al. 2016; Ramstedt et al. 2022; Ramstedt 2019; Torvik et al. 2011; Velleman, Templeton, and Copello 2005). These children are overrepresented among students with incomplete grades in compulsory school and ineligibility for upper secondary school (Hjern and Manhica, 2013; Lagerberg and Sundelin 2000; Lindgaard 2006; Statens Folkhälsoinstitut 2012; Templeton et al. 2009). They also risk poorer opportunities in working life, poorer health during their upbringing and later in life (Berg et al. 2016; Jääskeläinen et al. 2016; Torvik et al. 2011), and delinquency at a young age (Hoeve et al. 2012; Torvik et al. 2011).

Interventions for these children and for their parents with SUPs are commonly offered by the social services. Interventions may include individual conversation-based support, group activities targeting the children, parental support programmes, family-based treatment, day activities, networking, social initiatives in schools, and field activities (Ramstedt et al. 2022; Rehnman and Andrée Löfholm 2009; Rehnman, Andrée Löfholm, and Wiberg 2009; Socialstyrelsen 2009; Velleman, Templeton, and Copello 2005; Copello, Velleman, and Templeton 2005). According to these refereed studies above, there is agreement that these interventions have mainly targeted either the children directly (e.g. support groups) or parents with problem drinking. Few studies examine interventions offered to the family or the family network. Although such studies are scarce, there are reasons to believe that strengthening the bonds within the family may have beneficial effects on its members. According to social control theory, children in families with strong affective ties are more likely to care about the normative expectations of their parents, which will protect them from delinquency (Hirschi 1969/2002), and attachment theory stresses the need for a continuous parent – child attachment relationship during infancy to prevent long-term detrimental consequences for the child’s emotional character (Bowlby 1973). Hoeve et al. (2012) found in a meta-analysis of 76 empirical studies evidence suggesting that poor attachment increases the risk of delinquent behaviour.

In one of few studies explicitly examining interventions offered to parents with SUPs and their children, Niccols et al. (2012) reported slightly better results for integrated than individual treatment in terms of improvements in child development, growth, and emotional and behavioural functioning. These findings were in line with a previous study by Forrester et al. (2008), who reported positive effects in terms of reduced parental substance use and better child well-being from two different family interventions in the UK. A review of family interventions intended to reduce parental substance use found that studies were often limited by small numbers of participants, weak or absent control groups, and limited follow-up rates over short time periods (Copello, Velleman, and Templeton 2005). Despite this, the researchers argued that family interventions may motivate treatment entry for parents with SUPs and produce positive outcomes for other family members. In a systematic review of US prevention programmes targeting families with parental SUPs, Kumpfer, Alvarado, and Whiteside (2003) highlighted the value of including interactive learning in order to change behaviour, as well as practical help with meals, transport, and childcare.

Despite these preliminary positive outcomes of family interventions, the previous literature seldom assesses how children react to the interventions (Järkestig Berggren and Hanson 2016), and there is insufficient knowledge of how interventions affect the parents’ abilities to support their children. Given the strong evidence of children being negatively affected by parental SUPs, the absence of reported effects of social service interventions, concerning both parental mental health and how the children are affected, is remarkable Raninen and Leifman (2014); Socialstyrelsen (2013); Socialstyrelsen (2009). Given the lack of established and evaluated interventions offered to both parents with SUPs and their children the programme Me & my family was developed, a programme in which children and parents, supported by social workers, participate together in endeavouring to communicate how parental SUPs affect child well-being. Our hypothesis is that participating in the programme might positively affect family bonds and well-being.

The aim of the study

The present study aims to evaluate the outcomes of the knowledge- and motivation-enhancing programme Me & my family, focusing primarily on child well-being but also on family climate, parental alcohol and drug use, and self-rated mental health. Addressing the limitations of previous research, this study also includes children’s self-ratings.

The me & my family programme

The Me & my family programme was developed between 2013 and 2017 to support families with parental SUPs. In a pilot study of ten families, both parents and especially children reported positive outcomes, a finding confirmed by social work professionals’ observations (Alexanderson and Jess 2015). Me & my family can be characterized as a knowledge- and motivation-enhancing programme, based on a biopsychosocial perspective, in which children and parents participate together. This manual-based programme includes eight weekly sessions led by two therapists from the social services, one from the section for child welfare and another from the section for adults with SUPs. The programme includes psycho-educational components, role-playing, and various exercises. The programme aims to create a more open conversational climate within the family, increase family cohesion, and reduce feelings of shame and guilt in the children.

Me & my family sessions

Before the programme starts, a pre-session is held to inform the family about the programme. Each session starts with communication exercises, and goes on to include other playful exercises, ‘mini-lectures’ on the day’s theme, and discussions of various issues. After feedback from the family, the session ends.

The manual offers a number of exercises that go with each theme. The therapists choose the exercise(s) most suitable for the family. A discussion always follows (see Table 1).

Table 1. An overview of the Me & my family programme.

Session	Content
Pre-session	Information about the programme
1	Mini-lectures about the brain and how addiction develops
2	How addiction develops, continued
3	How addiction affects the family and its members
4	How addiction affects the family, continued
5	Communication within the family The importance of fair communication
6	Me, in my home, working for change
7	Security plan: Where and to whom do I turn for help if I need it?
8	Wrap-up
Post-sessions six weeks and three months post-intervention	Follow-up

Examples of exercises:

• Draw a picture or write about how alcohol or drugs have affected you: How do you feel when mum/dad uses alcohol or drugs?

• The wishful balloon: Each family member writes a note about how they want their family to change. The note is put into a balloon that is blown up. The participants catch a balloon, pop it, and read the wish.

• List: What can you do to support, help, and protect your child to improve their well-being? OR List: How do I want my parent(s) to help and support me so I can feel all right and be happy?

Methods

Procedure and participants

The participants were recruited from different sectors within the social services of ten municipalities in Mid Sweden with populations of 11,000–396,000 (mean = 84,900). Each participating municipality was asked to assign contact persons, recruit therapists for designated training, and inform and recruit families to participate. Families were recruited from March 2018 to August 2021. Inclusion criteria were wide, to include as many families as possible. Families with children aged 5–20 years, in which one or more of the adults had SUPs, were invited to participate. The adults did not need to abstain from substance use, but abstinence was an absolute requirement during sessions. Exclusion criteria were presence of an ongoing child custody investigation and/or occurrence of violence in the families, or parents experiencing active psychotic symptoms and/or a period of ongoing intense SUP; families to which these criteria applied were advised to take part in the programme when circumstances were less acute.

For this study, 59 children (females = 32 [54.2%], mean age = 11.17, SD = 3.66, age range = 5–20 years) and 67 parents (females = 34 [50.7%], mean age = 41.24, SD = 8.51, age range = 27–64 years) were included. Of the children, 28 participated with one parent and 31 with both parents. Of the parents, 46 had ongoing SUPs and 49 had undergone previous treatment for their own or a relative’s SUPs. Of the children, 22.4% reported living with both parents, 30.6% with their mother, 14.3% with their father, 14.2% sometimes with their mother and sometimes with their father, 14.3% with others, and 4.1% on their own.

The parental employment rate was quite high: 55 out of 67 parents worked 50% or more of full time working week. Up to 76 out of 126 participants reported having undergone earlier interventions within the social services.

The therapists

The therapists were social workers with academic degrees or were trained alcohol therapists without academic degrees. The therapist training, led by the researchers, lasted two full days with one additional day for follow-up. The programme training included lectures on research design and programme content, as well as exercises. In addition, coaching sessions with the researchers were arranged for each municipality regularly every two months, or more often if requested. Twice-yearly meetings – either on-site or virtual – were organized for all therapists and contact persons. In those meetings, sufficient time was allotted for exchanging practical experiences.

Study design and ethical considerations

The study was initially designed as a randomized controlled trial (RCT) in which participating families were randomly assigned either to start with the intervention or to a three-month waiting list. The families that were not included in our study or randomized to the waiting list received treatment as usual within the social services. However, the aim of using the waiting list proved difficult to realize. On several occasions during the first year, families who were assigned to the waiting list lost their motivation to participate: some relapsed into heavy substance use while waiting, while others became incarcerated or experienced the death of close family members. For ethical reasons, we decided it was important to start the intervention while the motivation to participate was high. We therefore revised the study design and instead measured the outcome by comparing family members with their own baseline measures on completing the intervention and at three, six, and 12 months post-intervention. We did not find that the design with families on a waiting list caused the drop-outs, as the waiting list drop-out occurred before the programme started.

This study reports the outcomes on completing the intervention and at three months post-intervention. The long-term effects at six and 12 months will be presented in an upcoming report, as will interview studies of participating children and adults. The project was approved by the Swedish Ethical Review Authority (Dnr 2018/094; 2019–02212).

Measures

The family climate questionnaire

To assess the family atmosphere and the interaction between family members, we used the Family Climate Questionnaire (Hansson 1989). In this instrument, recommended by the National Board of Health and Welfare in Sweden, all participants aged 11 years and older selected a minimum of 15 out of 85 listed adjectives. The adjectives are grouped into the following categories: closeness, represented by 18 adjectives describing a positive climate of warmth, security, and harmony (e.g. ‘happy’ and ‘loving’); distance, represented by 11 adjectives describing a negative climate (e.g. ‘intolerant’ and ‘aggressive’); spontaneity, represented by six adjectives describing the emotional expressiveness in the family in both positive and negative terms (e.g. ‘lively’ and ‘stressful’); and lastly chaos, represented by six adjectives describing chaotic states (e.g. ‘nervous’ and ‘confused’). The responses were then calculated in accordance with the number of responses relative to the total sum of adjectives and then indexed as a factor index for each individual. The Family Climate Questionnaire has been validated in a Swedish context and shows good conformity with the Child Behavior Checklist, Sense of Coherence Scale, and Family Environment Scale (Johnsson and Söderlind 2004). Pre- and post-intervention ratings, together with the number of adjectives chosen from the Family Climate Questionnaire, were collected from both the parents and children.

The strengths and difficulties questionnaire (SDQ)

SDQ is a 25-item behavioural screening questionnaire for the assessment of child mental health (Goodman 2001). SDQ includes items measuring symptoms of emotional and conduct problems, hyperactivity/inattention, peer problems, and prosocial behaviour on a three-point scale ranging from 0 (‘not true’) to 2 (‘certainly true’), with higher scores indicating more severe problems. Sometimes the three-point scale is reversed, meaning that 0 means ‘certainly true’. SDQ also has an impact supplement targeting the level of distress and social impairment. The Swedish version of SDQ has shown adequate validity and is considered a useful tool for mental health screening in children and adolescents (Malmberg, Rydell, and Smedje 2003). In this study, SDQ scores are based on pre- and post-intervention informant reports from parents of younger children (aged 3–10 years) and on self-report measures from adolescents aged 11–16 years. We applied SDQ as a continuous total difficulties score generated by summing the scores from all the scales except the prosocial scale, with a possible range of 0–40.

The hospital anxiety and depression scale (HADS)

HADS is a 14-item self-report questionnaire for adults, measuring anxiety (seven items) and depressive (seven items) symptoms during the preceding week. Each item is scored on a four-point scale of 0–3, with higher scores indicating the greater occurrence of symptoms. The possible range of summary scores for each of the two subscales is 0–21, with scores of 0–7 representing no signs of anxiety or depression, 8–10 mild levels of anxiety or depression, 11–14 moderate levels, and 15–21 severe levels. The scale has shown good validity in different populations, including Swedish populations (Bjelland et al. 2002; Zigmond and Snaith 1983). Parents’ pre- and post-measures of their own emotional problems were collected; we present the total summed score, with a possible range of 0–42 points.

The alcohol use disorders identification test (AUDIT)

AUDIT is a ten-item self-report questionnaire covering different domains of alcohol consumption, drinking patterns, and alcohol-related problems over the previous 12 months (Saunders et al. 1993). To measure changes in alcohol consumption, we used the first three items (constituting AUDIT-C), measuring the quantity and frequency of drinking. Each item (‘How often do you have a drink containing alcohol?’, ‘How many drinks containing alcohol do you have on a typical day when you are drinking?’, and ‘How often do you have six or more drinks on one occasion?’) is scored on a scale of 0–4. AUDIT is widely used and has been validated with good results (Reinert and Allen 2007). AUDIT-C was completed by the parents pre- and post-treatment and had a possible range of 0–12, with a higher score indicating more alcohol consumption.

The Drug Use Disorders Identification Test (DUDIT)

Drug habits were measured using DUDIT, an instrument parallel to AUDIT for the identification of individuals with drug-related problems (Berman et al. 2005). For follow-up assessment, items 1–4, measuring the quantity and frequency of drug use, were used. Each item (‘How often do you use drugs other than alcohol?’, ‘Do you use more than one type of drug on the same occasion?’, ‘How many times do you take drugs on a typical day when you use drugs?’, and ‘How often are you influenced heavily by drugs?’) is scored on a scale of 0–4. DUDIT was completed by the parents pre- and post-treatment and had a possible range of 0–16, with a higher score indicating more consumption of drugs.

‘Motivation to change’ and ‘Belief in change’ were self-reported using a visual analogue scale of 1–10, 1 means a low belief and 10 a high belief.

Statistical analyses

All analyses were conducted with IBM’s Statistical Package for the Social Sciences (SPSS), version 27. Descriptive statistics were used to describe the sample. Attrition analyses were assessed using independent-samples t-tests and chi-squared tests. Linear mixed models were used to investigate changes in symptom ratings, family climate, and use of alcohol and drugs before (T1), on completing (T2), and three months after (T3) the intervention. The participants’ baseline values were used as random intercepts and time was used as the fixed factor; any changes in the outcome measures from T1 to T2 and from T1 to T3 were compared separately. In the ‘Results’ section (below), the estimated marginal means (EMMs) derived from a model of the outcome measures are presented. As a sensitivity analysis, we recalculated the model for all the subscales of SDQ. All reported results were considered significant at the 5% level.

Results

Participant characteristics

Participant characteristics and attrition analyses are presented in Table 2. Of the total sample, 71% (i.e. 72% of the parents and 69% of the children) completed the intervention. The most commonly reported reason for not completing the intervention was relapsing into addiction for the parent, followed by death in the family and too many ongoing treatment contacts. The attrition analysis showed no differences in the baseline values between those who completed the intervention and those who dropped out. Of those who completed the intervention, 29 (71%) of the children and 45 (94%) of the parents completed the ratings immediately on completing the intervention, and 31 (76%) of the children and 41 (85%) of the parents completed the ratings three months after the intervention.

Table 2. Participant characteristics, and comparison between those completing and those dropping out of the intervention in terms of characteristics and baseline measures.

	Participated n = 41 children; n = 48 parents	Drop-outs n = 18 children; n = 19 parents	Statistics
Gender, children (females)	55.3%	50.0%	χ^2 = 0.14; p = 0.712
Gender, parents (females)	52.1%	47.4%	χ^2 = 0.12; p = 0.728
Age child, mean (SD) parent, mean (SD)	11.15 (3.84) 41.38 (9.08)	11.22 (3.32) 40.89 (7.08)	t = 0.73; p = 0.942 t = 0.21; p = 0.837
School participation, child	100%	90%	χ^2 = 3.08; p = 0.079
Occupation (work, student, parental leave), parent	61.5%	66.7%	χ^2 = 0.10; p = 0.748
Previous treatment for substance use problems	62.4%	65.5%	χ^2 = 0.09; p = 0.760
Family climate ^a Closeness child, mean (SD) parent, mean (SD) Distance child, mean (SD) parent, mean (SD) Spontaneity child, mean (SD) parent, mean (SD) Chaos child, mean (SD) parent, mean (SD)	1.72 (0.80) 1.58 (0.74) 0.44 (0.56) 0.26 (0.48) 1.03 (0.82) 0.97 (0.93) 0.86 (1.16) 1.12 (1.07)	1.30 (0.79) 1.59 (0.85) 0.30 (0.41) 0.20 (0.27) 1.40 (1.03) 1.32 (0.97) 1.29 (0.83) 1.10 (1.43)	t = 1.21; p = 0.238 t = 0.64; p = 0.949 t = 0.58; p = 0.565 t = 0.49; p = 0.623 t = 0.98; p = 0.335 t = 1.34; p = 0.187 t = 0.91; p = 0.370 t = 0.37; p = 0.971
Mental health, child^,b mean (SD)	13.13 (5.00)	13.38 (6.90)	t = 0.14; p = 0.889
Emotional problems, parental^,c mean (SD)	12.25 (6.37)	11.67 (7.87)	t = 0.31; p = 0.761
Alcohol use, parental^,d mean (SD)	3.87 (3.06)	3.67 (2.57)	t = 0.25; p = 0.801
Drug use, parental^,e mean (SD)	3.82 (9.01)	4.17 (9.06)	t = 0.14; p = 0.892
Motivation to change^,f parental, mean (SD)	8.72 (2.75)	8.89 (1.66)	t = 0.24; p = 0.814
Belief in change^,g parental, mean (SD)	8.96 (1.88)	9.02 (1.17)	t = 0.12; p = 0.904

Notes .

^aAssessed using the Family Climate Questionnaire; ^b Assessed using the Strengths and Difficulties Questionnaire; ^c Assessed using the Hospital Anxiety and Depression Scale; ^d Measured using the Alcohol Use Disorders Identification Test; ^e Measured using the Drug Use Disorders Identification Test; ^f Self-reported on a scale of 1–10; ^g Self-reported on a scale of 1–10.

Family climate

Changes in the outcomes over time are presented in Table 3. The children and their parents self-reported more closeness and less chaos in the family on completing the intervention, a change that persisted three months later. Parents self-reported less spontaneity on completing the intervention. No changes over time were found in rated distance within the families.

Table 3. Primary and secondary outcomes with estimated marginal means (EMMs) obtained from the linear mixed model.

Variable	Baseline (T1) EMM (CI)	Post (T2) EMM (CI)	Follow-up (T3) EMM (CI)	T1–T2 B, p-value	T1–T3 B, p-value
Family climate ^a
Closeness
child	1.69 (1.41–1.96)	2.16 (1.86–2.46)	2.19 (1.88. 2.50)	–0.47, p = 0.002	–0.50, p = 0.002
parent	1.58 (1.38–1.78)	2.08 (1.85–2.32)	1.92 (1.68–2.16)	–0.50, p < 0.001	–0.34, p = 0.009
Distance
child	0.40 (0.23–0.57)	0.25 (0.70–0.43)	0.21 (0.27–0.40)	0.15, p = 0.112	0.19, p = 0.053
parent	0.24 (0.15–0.34)	0.17 (0.06–0.28)	0.23 (0.12–0.35)	0.07, p = 0.277	0.01, p = 0.872
Spontaneity
child	1.06 (0.76–1.36)	0.99 (0.68–1.31)	0.83 (0.51–1.15)	0.07, p = 0.551	0.43, p = 0.66
parent	1.07 (0.86–1.27)	0.73 (0.48–0.97)	0.64 (0.38–0.89)	0.34, p = 0.014	0.43, p = 0.003
Chaos
child	0.95 (0.68–1.23)	0.32 (0.01–0.63)	0.35 (0.04–0.67)	0.64, p < 0.001	0.60, p = 0.001
parent	1.11 (0.87–1.36)	0.59 (0.29–0.88)	0.46 (0.16–0.77)	0.53, p = 0.003	0.65, p < 0.001
Mental health, child ^b	12.70 (11.06–14.34)	13.14 (11.26–15.03)	11.22 (9.36–13.07)	−0.45, p = 0.594	1.48, p = 0.084
Emotional problems, parental ^c	12.23 (10.63–13.83)	10.31 (8.49–12.13)	9.23 (7.26–11.20)	1.92, p = 0.026	3.00, p = 0.002
Alcohol use, parental ^d	3.59 (2.90–4.27)	2.05 (1.24–2.86)	2.05 (1.22–2.89)	1.54, p = 0.001	1.54, p = 0.001
Drug use, parental ^e	1.07 (0.51–1.63)	0.32 (−0.37–1.01)	0.06 (−0.63–0.75)	0.75, p = 0.73	1.01, p = 0.017

Notes CI = confidence interval, T1 =before intervention, T2 = on completion of intervention.

T3=three months post-intervention.

Bold indicates p<0.05.

a Assessed using the Family Climate Questionnaire; b Assessed using the Strengths and Difficulties Questionnaire; Table 3 c Assessed using the Hospital Anxiety and Depression Scale; d Measured using the Alcohol Use Disorders Identification Test; e Measured using the Drug Use Disorders Identification Test.

Mental health and parental substance use

Ratings of total behavioural difficulties and emotional problems did not change from before to completion of the intervention in the children (see Table 3). However, the sensitivity analyses, exploring changes in ratings for the different subscales of SDQ, detected reduced conduct problems three months post-intervention compared with baseline ratings (see Table 4). As shown in Table 3, parents reported reduced self-rated symptoms of anxiety and depression on completing the intervention. Parents also reported drinking less on completing and three months after the intervention, and less drug use three months after the intervention.

Table 4. Child mental health outcomes measured using the Strengths and Difficulties Questionnaire with estimated marginal means (EMMs) obtained from the linear mixed model.

Variable	Baseline (T1) EMM (CI)	Post (T2) EMM (CI)	Follow-up (T3) EMM (CI)	T1–T2 B, p-value	T1–T3 B, p-value
Total difficulties	12.70 (11.06–14.34)	13.14 (11.26–15.03)	11.22 (9.36–13.07)	−0.44, p = 0.594	1.47, p = 0.084
Emotional symptoms	3.28 (2.64–3.92)	3.16 (2.41–3.91)	2.56 (1.83–3.30)	0.21, p = 0.741	0.72, p = 0.055
Conduct problems	2.42 (1.92–2.91)	2.16 (1.59–2.72)	1.91 (1.36–2.47)	0.26, p = 0.283	0.51, p = 0.042
Hyperactivity/inattention	4.99 (4.28–5.70)	5.41 (4.59–6.22)	4.83 (4.03–5.63)	−0.42, p = 0.256	0.16, p = 0.670
Peer problems	2.12 (1.65–2.58)	2.49 (1.94–3.04)	2.00 (1.45–2.54)	−0.37, p = 0.178	0.12, p = 0.665
Prosocial behaviour	7.90 (7.32–8.48)	8.28 (7.61–8.96)	7.64 (6.98–8.31)	−0.38, p = 238	0.26, p = 0.434
Impact of problems	1.41 (0.73–2.09)	1.41 (0.57–2.24)	0.55 (−0.22–1.32)	0.00, p = 0.99	0.86, p = 0.063

Notes CI = confidence interval, T1 = before intervention, T2 = on completion of intervention, T3 = three months after intervention. Bold indicates p < 0.05.

Discussion

This study aimed to evaluate the effects of the knowledge- and motivation-based programme Me & my family, provided by the social services for families in which there are parental SUPs. The initial plan to use an RCT design was abandoned at an early stage as it was considered problematic for families to remain on a waiting list. Approximately two-thirds of those who started the programme completed it. Of these, both the children and their parents self-reported more closeness and less chaos in the family post-intervention, and reductions in child conduct problems and in parental symptoms of anxiety and depression were also found. Moreover, parents reported less substance use after the treatment.

As indicated in previous studies, research on families in which there are parental SUPs can be challenging, with small numbers of participants and an absence of control groups and follow-up measures (Copello, Velleman, and Templeton 2005). Accordingly, we soon realized that a prolonged research period was needed to obtain a sufficient number of participants, so the initial plan to randomize families to a waiting list was abandoned. Discussions with the therapists were intense during the first year, and we realized that families from the waiting list had dropped out of the study, constituting an ethical issue. However, families on the waiting list were offered treatment as usual within the social services. The drop-out rate pre-intervention was also higher for those randomized to the waiting list than for those randomized to immediate intervention, which encouraged our decision to abandon the initial plan. In our study, the drop-out rate was about one third, and there were multiple reasons for not completing the programme. Besides increased drop-outs among those randomized to the waiting list/control group for three months, other reasons for dropping out were SUPs and parental mental health problems that were more challenging than was previously known by the social services. There were also suspicions of ongoing violence in some families, which was an exclusion criterion in our study.

Despite the above issues, our research findings show, post-intervention, that families self-reported more closeness and less chaos according to the Family Climate Questionnaire. Poor attachment is linked to delinquency, especially in younger children (Hoeve et al. 2012), and according to social control theory (Hirschi 1969/2002), delinquency is low in families with strong affective bonds. It has been hypothesized that children and adolescents who are strongly attached to their parents are more likely to care about the normative expectations of their parents, in turn protecting them against delinquent impulses. Me & my family aims to strengthen communication processes within the family, making the unspoken communicable and strengthening family bonds by encouraging the family to do things together. These parts of the programme might serve as affective bonds, preventing children and adolescents from engaging in future delinquency. Previous research has emphasized the significance of family interventions being more promising than treating the parent(s) and child(ren) separately (Copello, Velleman, and Templeton 2005; Forrester et al. 2016; Velleman, Templeton, and Copello 2005). We want to highlight the consistency between child and parental ratings of family climate in our study, which indicates that family interventions such as Me & my family do affect the whole family.

In line with previous research on family interventions, parents self-reported higher well-being on completing and three months after the intervention. However, the ratings for child well-being are not comparable to those in previous research due to the lack of ratings for children. Parents self-reported fewer emotional problems according to HADS, and children reported fewer conduct problems according to SDQ. In this study, we expand on previous literature by including the child’s perspective. The findings are important given that children whose parents have SUPs have been shown to exhibit more conduct problems (Ramstedt et al. 2022). The attenuation of conduct problems pre- to post-intervention might further prevent children from not completing their schooling and thus having poorer chances later in life. Although reduced parental substance use was not a primary aim of the programme, the present results indicate an improvement in this respect from before the intervention to completion of the intervention and three months later. This is positive, because previous studies have linked parents’ SUPs to children’s poorer performance at school (Berg et al. 2016; Jääskeläinen et al. 2016; Torvik et al. 2011).

Strengths and limitations

Previous studies have mainly focused on alcohol and drug use, and a significant strength of this study is its use of other relevant outcomes, such as children’s self-reports regarding family climate and emotional problems. Our intervention also explicitly adopts the child’s perspective and assumes children’s participation, which is unique. The Me & my family intervention is among the non-compulsory services offered by the social services. Mean AUDIT-C scores were in the lower range at baseline, possibly because the parents were presenting at a social welfare office that protects children. Furthermore, approximately half of the parents who entered the Me & my family programme had undergone previous treatment for their SUPs. However, we believe that the lowered scores at follow-up mirror real-life reduced substance use.

Delivering the programme presupposes cooperation between the social service units targeting families and children and those targeting adults with SUPs. Beyond these challenges, professionals are also challenged to run a research project within ordinary services. Even with arrangements such as one folder per family with all the required questionnaires in place and regular supervision once a month, the professionals sometimes mixed up or failed to collect the questionnaires. Research in the everyday clinical setting is uncommon within the social services in Sweden, and although they were generally positive towards the research, many therapists were unacquainted with research routines. A further challenge in this study was that of bridging problems, i.e. problems related to the specialization of social services, with professionals belonging to different organizational ‘silos’. The programme challenged this, as professionals from different specialized areas were supposed to cooperate in the programme. These challenges will be discussed further in an upcoming article.

The measures in this study are exclusively self-reported, and these self-reported data are not supplemented with other scales or observations. In the future, self-ratings could be complemented with other more objective measures and observations. Furthermore, a broad age range of children was included in this study, i.e. 5–20 years; hypothetically, the treatment programme may be more beneficial for some age groups than others. We might expect this family-based programme to benefit younger children more than older adolescents who are no longer as dependent on their parents. However, to investigate whether the outcomes are moderated by participant age, we would need a larger study sample. As mentioned above, the initial design in which families were randomized to immediate intervention or a waiting list for three months was abandoned. However, as RCTs are regarded as the ‘gold standard’ method for evaluating the effectiveness of interventions, the conclusions that can be drawn from our study are limited. In the future, we will have to consider study designs that meet the need for control groups as well as the participants’ need for immediate interventions.

Conclusions

Social service efforts for families with SUPs have mostly been carried out as individual interventions targeting adults or as support programmes targeting children. There is an urgent need to develop programmes that consider the family perspective, i.e. strengthening family bonds and upholding the children’s right to express their vulnerable situation in the presence of their parents. The outcomes of the knowledge- and motivation-based programme Me & my family are promising. On completing the intervention and three months later, families reported more closeness and less chaos for both children and parents, improved child well-being rated as fewer conduct problems, less parental anxiety, and reduced parental substance use. Furthermore, this study indicates positive experiences from cooperation between family and child units and adult addiction units within the social services. It is of particular interest to further develop exchange and cooperation across organizational boundaries in a landscape of increasingly specialized social work organizations. The results need to be replicated in order to evaluate the programme’s effectiveness by including an acceptable control intervention. Although our findings support the use of family interventions when parents have SUPs, follow-up measures are needed to ensure that any positive gains from the intervention persist over time. In addition, in-depth interviews are planned in order to explore how the families experienced the intervention, which may further stimulate changes in the procedural manual in order to increase acceptability.

Acknowledgments

We also thank the professionals within the social services and the families for participating in this research project.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The study has had financial support from The Public Health Agency 2018–2020, Dalarna University 2021–2022 and The Alcohol Research Council of the Swedish Alcohol Retailing Monopoly