The Parents under Pressure parenting programme for families with fathers receiving treatment for opioid dependence: the PuP4Dads feasibility study

Parenting programmes

Parenting programmes targeted at the general population have been successful in improving parents’ psychological functioning, including depression, anxiety, stress, anger, guilt, confidence and satisfaction with the partner relationship. 23 However, parenting studies rarely include fathers, and those that do rarely report outcomes for fathers. 24 Evidence suggests that fathers are reluctant to attend parenting programmes,25 which may relate to their perceptions regarding gendered parenting roles and mother-focused interventions. 26 Tully et al. 27 also found that lower levels of practitioner competencies, training and experience in engaging fathers in parenting programmes resulted in lower levels of father involvement. One concern is that there is some evidence that fathers benefit less from mainstream parenting programmes than mothers. 28,29 A meta-analysis of parenting intervention studies that included fathers compared with those that did not found significantly more positive changes in children’s behaviour, but fathers reported fewer desirable changes in parenting behaviours and beliefs than mothers did. 29 The authors concluded that fathers should be encouraged to engage in parenting programmes to maximise benefits for children, but that further research is needed on recruitment and retention of fathers, adjustments required to address the parenting needs of both fathers and mothers, and how mothers and fathers work together, or not, to implement the active components of the intervention.

Encouragingly, a small body of researchers have implemented and evaluated parenting programmes for marginalised fathers, such as young and low-income fathers, those engaged in intimate partner violence and those subject to child protection orders. 30–34 Findings indicate improvements in psychological functioning30,33 and father–child relationships. 31,34 However, poor engagement with fathers remains problematic. 32,35 In addition, it is not known how fathers affected by other forms of disadvantage and co-occurring problems might respond to parenting programmes, for example fathers who are drug dependent. Parents who use drugs are highly stigmatised and marginalised, and are often reluctant to engage with group-based parenting programmes. 36 Consequently, there has been growing recognition that targeted and more nuanced interventions are required for marginalised families, including fathers,23 especially when the father poses a risk of child maltreatment. 37

Parenting programmes for parents with a substance use disorder

A review of parenting programmes for those with problem drug use found that therapeutic interventions can be helpful in reducing drug use and improving parenting. 38 However, the review found that programmes commonly target parenting without taking into account the broader needs of substance-using parents. Although it is difficult to disentangle the negative impact of prenatal exposure and the postnatal environment that the child is raised in, it is clear that children need a family environment in which they feel safe and loved and are nurtured by reliable and affectionate carers. The multiple difficulties facing parents who use drugs can interfere with or lessen their ability to provide a child with this safe, loving and nurturing environment.

An innovative US intervention for fathers with problem drug use – ‘Fathers for Change’ (F4C) – was compared with ‘Dads “n” Kids’, which combines psychoeducational parenting with practical support for housing and welfare. 39 F4C addresses drug use, parenting skills and proximal psychological factors associated with intimate partner violence. Both intervention groups reduced affect dysregulation, anger and intimate partner violence. The reduction in affect dysregulation was significantly larger among participants receiving F4C, and participants were also less likely to use drugs (relapse) following the intervention. The authors suggest that F4C could be delivered in the community with fathers and their children (see also Stover40). However, the pilot randomised trial focused on fathers only, and those in abstinence-based residential drug treatment (a 12-step programme), with the majority of fathers mandated by the criminal justice system in return for reductions in their sentences.

The Parents under Pressure (PuP) programme,41 which was developed by members of the research team, is a parenting intervention that is specifically designed to address the needs of parents who use drugs. The PuP programme is underpinned by an integrated theoretical framework that draws on attachment theory,42 the concept of emotional availability,43 neurobiological models of stress and trauma44 and a socioecological developmental model of human development. 45 Each of these theoretical perspectives identifies the quality of the parent–child relationship as crucial for the healthy development of a child. Furthermore, the personal resourcefulness of parents (i.e. capacity to regulate their own emotional state, problem solve and engage social support) and the demands of the real world (e.g. housing and legal problems, financial strain and food insecurity) can have an negative impact on the parent–child relationship. Therefore, the PuP programme is based on the premise that responding to the needs of parents who use drugs requires a framework that can address problems across multiple domains of family functioning.

In the UK, the PuP programme has been subjected to a randomised controlled trial (RCT) and economic evaluation. 46 Parents were recruited from community-based drugs services in the UK and 115 primary caregivers (mostly mothers) were randomly allocated to the PuP programme or treatment as usual. The intervention was delivered by accredited PuP practitioners in the family home. Those participants receiving PuP demonstrated significantly reduced child abuse potential (primary outcome) compared with controls, and the intervention was cost-effective. 46 However, the small number of fathers in this UK study precluded an analysis of how the intervention might have affected them and their children.

Neger and Prinz38 recommended that more robust evaluations were required to assess efficacy and feasibility of these interventions. Evidence from the PuP programme and F4C studies present a compelling argument to involve fathers who use drugs in programmes that aim to improve fathering and father–child relationships within the context of family life. Greater emphasis should be given to reducing aggression towards children and partners and improving affect regulation. Given the context and target population [i.e. fathers mandated to attend residential abstinence-based drug treatment in F4C (USA) and mothers experiencing problem substance use and living in the community in PuP (UK)], difficulty with recruitment and retention46 and initial challenges embedding the intervention into existing services,39 it is efficacious to assess the feasibility of implementing an intervention for fathers into a new setting, such as the UK.

Rationale for the PuP4Dads feasibility study

We chose to focus on PuP in the current study because of our existing research collaboration with the originators of that programme. The proposed research also takes the vital next step of targeting men in parenting intervention research. It provides one of the first attempts to improve family functioning by addressing affect regulation as a key driver of parenting and couple-related behaviours among drug-dependent men receiving community-based opioid substitution therapy (OST). In addition, to the best of our knowledge, it is the first attempt to include opioid-dependent fathers, their partners/mothers, their children and their significant others in a ‘whole-family’ approach that is in line with the governmental policy priorities for this population in the UK. 14,16 In the UK, there are very few residential drug treatment units that can accommodate fathers and their children,47 and so demonstrating successful implementation within community-based settings in the UK is essential for future scale-up. As PuP has already been trialled in the UK, the research and intervention are highly relevant to the UK public health agenda, with implications for improving the quality of caregiving in families with complex needs, reducing child abuse and situational family violence, improving children’s developmental outcomes, and improving the trajectories of adults and families affected by drug use.

In addition, interventions need to include a focus on the broader ecological context of families’ lives, help families connect with a wider social environment and provide safety and some security around accommodation and financial issues. 38,48 Although relatively small in number, there is growing evidence that promotes such approaches when delivered across community settings, such as OST settings,49,50 and within the context of family drug courts. 51,52 Notably, these programmes address multiple domains in families’ lives and incorporate home visiting and a case management approach to address wider contextual factors.

Following the Medical Research Council guidance on the development and evaluation of complex interventions,53 there are five key reasons why a feasibility study was required in this instance. First, our proposed intervention is novel in that it involves an extremely high-risk group of fathers and is positioned in an area of research that is considerably under-researched. Second, given the well-known problems in engaging with parents who use drugs, we need to understand how these problems might affect the flow of participants in a research study (i.e. from recruitment to post-intervention follow-up). Third, we need to better understand the extent to which the intervention works as intended, including the facilitators of and barriers to adoption and implementation. Fourth, there is the need to establish a theoretical understanding of the mechanisms that may lead to improvement and thereby identify the most appropriate mediators, moderators and outcome variables and measures for a future evaluation. Finally, given the small number of evaluations, and drawing from guidance regarding the best use of resources,53 we need to consider the resource and methodological implications of conducting a larger evaluation, including the feasibility of an economic analysis.

Aim

Our aim was to implement and test the feasibility and acceptability of the PuP programme for opioid-dependent fathers and their families and to determine whether or not a future pilot RCT and full-scale evaluation, including an economic evaluation, could be conducted.

Research questions

Objectives

• To determine whether or not a pilot RCT and full evaluation, including an economic evaluation, could be undertaken on the PuP programme with drug-dependent fathers and their families.

• To assess the recruitment and retention of drug-dependent fathers, as well as feasibility and acceptability of the intervention among fathers, mothers, practitioners, referrers and key services.

• To assess the fidelity and reach of intervention delivery by PuP practitioners, including barriers to and facilitators of successful implementation.

• To refine and test the logic model and theoretical basis of the intervention.

• To enhance understanding of the parenting needs of drug-dependent fathers and what programme components work best with fathers.

• To determine key trial design parameters for a possible future large-scale trial, including recruitment and retention rates and strategies, outcome measures, intracluster correlation and sample size.

• To determine the key components of a future cost-effectiveness analysis and tested data collection methods.

• To establish whether or not pre-set progression criteria are met and a larger-scale trial is warranted.

  • If yes, to design the protocol, including identification of required structures, resources and partnerships.

Protocol changes and project extension

Research pathway diagram: PuP4Dads study

A visual representation of the research pathway is outlined in Figure 1.

FIGURE 1.

The PuP4Dads research pathway diagram. R&D, research and development.

Setting

The study was conducted in two non-NHS community-based services for adults, children and families affected by parental substance use in the Lothian region in Scotland:

1. PREPARE, which is a local-authority social work-led specialist multidisciplinary service for pregnant women, substance users and their families

2. CIRCLE, which is a Scottish charity and early years family support service for disadvantaged children and families.

The Lothian region in Scotland includes one health board (NHS Lothian) and four local authorities (The City of Edinburgh Council, East Lothian Council, Midlothian Council and West Lothian Council). This study took place in the three local authority areas in Lothian where PREPARE and CIRCLE provide services [i.e. City of Edinburgh (PREPARE and CIRCLE), East Lothian (CIRCLE) and West Lothian (CIRCLE)].

The City of Edinburgh is a largely urban area with a population of 513,210 (as of December 2020), with 11.6% of the population living in the most-deprived category [Scottish Index of Multiple Deprivation (SIMD) category 1] and 43.5% of the population living in the least-deprived category (SIMD category 5). 57 East Lothian has a population of 104,840 that is distributed fairly equally across SIMD categories 1 to 5, with approximately half of the population living in urban (21.6%) or accessible rural (23%) areas. West Lothian has a population of 181,310 that is distributed fairly equally across SIMD categories 1 to 5, with the majority (82.5%) of the population living in accessible small towns. 57 Although Edinburgh and many surrounding small towns in East Lothian and West Lothian are considered affluent, approximately 22% of all children in Edinburgh live in poverty. 58 There are several areas of extreme deprivation, comprising large social housing ‘estates’ with a high prevalence of drug use. 59

Across Scotland, health and social care is now ‘integrated’ with social care, primary and community health-care services and some acute services managed by integration joint boards. Integration joint boards are made up of representatives from councils, NHS health boards, third-sector representatives, service users and carers. These integration joint boards, through their chief officer, have responsibility for the planning, resourcing and the operational oversight of a wide range of health and social care services. The delivery of adult services are the responsibility of Health and Social Care Partnerships. Within these partnerships, there are various services for adults with drug and alcohol problems and in each local authority there is an Alcohol and Drug Partnership (ADP) that is made up of local authority, NHS, third-sector, Police Scotland and prison representatives that’s primary aim is to co-ordinate the design, delivery and evaluation of drug and alcohol services.

Although drug and alcohol services are not ‘integrated’ with children’s services, three local authority areas in Lothian commission CIRCLE (i.e. the third-sector organisation involved in this study) to provide services for families affected by parental substance use. 60 Uniquely, only the City of Edinburgh Health and Social Care Partnership provides a specialist service for pregnant women with problem substance use via PREPARE (i.e. the other service involved in this study). 61

PREPARE is a local government (City of Edinburgh Council)-led service that was established in 2006. It consists of an early intervention integrated multidisciplinary team that works with pregnant women and their partners who have significant substance use problems. The service aims to reduce substance use and related harms to mother and baby. The women and men who are allocated to PREPARE have multiple and complex needs related to poverty, substance use, poor mental health and domestic abuse. PREPARE receives approximately 50 referrals a year and normally has a caseload of up to 30 women plus their partners at any one time. Of PREPARE’s open cases, 90% will be involved in the child protection system, whereas the other 10% will be supported using the Getting It Right For Every Child (GIRFEC) model62 and Children Affected by Parental Substance Use (CAPSU) guidelines. 15 PREPARE comprises nine (full-time equivalent) staff, including a social work team leader, a senior early years officer, early years officers, community mental health (addiction) nurses, a health visitor and a midwife. The team is managed under social work children and families services and is supported by a named consultant obstetrician and consultant psychiatrist in addictions. At the start of the study, PREPARE also hosted a dedicated ‘fathers’ worker’ who was jointly funded and managed by both PREPARE and CIRCLE. The fathers’ worker is now based at CIRCLE.

CIRCLE, a Scottish charity founded in 2006, provides a range of family support services that aim to improve the lives of disadvantaged children and families. CIRCLE works in partnership with other organisations to provide whole-family support and delivers commissioned services specifically for CAPSU in the City of Edinburgh, East Lothian and West Lothian. These three services support approximately 240 families per year. In total, there are 14 (full-time equivalent) staff across the projects (Edinburgh, n = 7; West Lothian, n = 3.5; East Lothian, n = 2.5). These staff are assisted by a young person’s worker and a father’s worker in Edinburgh, and a pregnancy worker in East Lothian. Staff are qualified in social work, community development, health and social care and nursing. Funding for CIRCLE services and staff comes from a range of trusts and donations, as well as the ADPs in Edinburgh and West Lothian and a National Lottery Grant in East Lothian.

Parents under Pressure intervention theory of change

The PuP programme aims to enhance parents’ capacity to provide a safe and nurturing environment and sensitive and responsive caregiving for children. Sensitive and responsive caregiving, in combination with knowledge about appropriate parenting behaviours, can lead to improvement in child developmental outcome. However, to provide sensitive and responsive caregiving (including managing difficult behaviours and better limit setting), it is essential that parents are able to understand and manage their own emotions. Impulsivity and poor affect regulation are key features of substance use and can be viewed as a contributor to and a consequence of substance use. 64 Before parents, and in particular fathers/male caregivers, who have engaged in hostile and reactive behaviour patterns in the context of family life13 are able to respond sensitively to their children and partners, they need to be able to manage their own dysregulated affect.

Therefore, the PuP programme extends beyond instruction in traditional behavioural parenting strategies, such as managing non-compliance, better limit setting and rewarding good behaviour, to a focus on helping to develop a calmer, less reactive family environment in which both parents and children learn how to improve emotion regulation. There is extensive evidence supporting the direct relationship between the quality of caregiving, parenting knowledge and child outcome. 19,23 However, this is also mediated by parental emotion regulation: parental emotion dysregulation is associated with poor coping skills, poorer parental sensitivity and greater likelihood of harsh and abusive parenting. These, in turn, increase child abuse potential. For the families involved in the PuP programme, parental emotion dysregulation is a significant problem and is, therefore, directly addressed during programme delivery. This can be represented as a conceptual model in which the relationship between sensitive and responsive parenting (i.e. quality of caregiving), parenting skills (i.e. knowing what to do) and child outcome is influenced by the parent’s capacity to manage their emotions, which, in turn, influences child abuse potential (Figure 2).

FIGURE 2.

Conceptual representation of the PuP integrated theoretical framework.

For the purposes of the PuP4Dads feasibility study, we proposed a logic model for the intervention that is specifically in relation to fathers with a substance use disorder (see Report Supplementary Material 1).

Parents under Pressure intervention material

To deliver the PuP programme, an individually tailored family support plan is developed in collaboration with the parent/carers. The underlying integrated theoretical framework (see Report Supplementary Material 2) for parents and practitioners provides an assessment framework in which to identify strengths and challenges across six domains of family functioning: (1) child outcome; (2) quality of the caregiving relationship; (3) parenting values, knowledge and skills; (4) parental emotion regulation; (5) connections to culture, community and family; and (6) the wider ecological context. Additional resources documenting exercises, discussion points and activities are contained in the PuP parent workbook and the PuP online toolkit, which can be used flexibly to help parents and children enhance family and child well-being. The parent workbook consists of 12 discrete modules [Table 1; see also the NIHR Journals Library URL: www.journalslibrary.nihr.ac.uk/programmes/phr/158201/#/ (accessed 6 September 2021)]. 42 The order and sequencing of the family support plan is tailored to each family. Therefore, the order and focus of the modules delivered differs across families.

The PuP therapist manual provides the theoretical overview underpinning the programme and was provided to PuP practitioners as part of the training process. Materials are copyrighted to Griffith University (Nathan, QLD, Australia). Practitioners undergoing training and accredited practitioners in the PuP programme have access to the PuP resources, including online exercises from the parent workbook, access to standardised measures, scoring and an interpretation report. These measures (e.g. the Depression, Anxiety and Stress Scale) reflect the domains of family functioning contained in the integrated theoretical framework. Each practitioner has their own username and password used to access the PuP online toolkit.

Parents under Pressure intervention procedures

The PuP programme is individualised for each family. An assessment model allows for an individualised case plan to be developed, which is guided by a model of case conceptualisation. Immediate priority areas and goals for change are identified by the practitioner and parent, and are worked towards collaboratively. Modules are then selected and ordered according to the needs of the family and the immediate presenting issues. This approach allows for flexibility, as immediate problems may include, for example, potential homelessness and high risk of relapse to drug use, which need to be addressed to introduce both stability in the family environment and engage high-risk families. Integral to these processes is engagement of the parent(s) in the process of developing confidence in their parenting capacity, nurturance of their children and skills to identify and manage their emotions. These three ‘threads’ runs through all PuP sessions regardless of the task at hand or even the location of the session. Practitioners are encouraged to engage with the three ‘threads’ at all times (e.g. supporting a parent during a visit to the housing association provides a therapeutic opportunity to comment on and support each of the three threads).

The quality of the parent–child relationship is intrinsically linked with the capacity of parents to provide nurturing, sensitive and responsive caregiving,50 and the capacity of parents to regulate their own emotional state in the face of parenting challenges can fundamentally impact this capacity. 65,66 Therefore, many of the PuP treatment modules focus on improving parents’ emotional state and fostering a positive parent–child relationship. For example, the ‘connecting with your child to help them feel loved and safe’ and ‘supporting your child to develop self-regulation’ modules focus on helping parents to develop a range of appropriate and non-punitive child management techniques, strategies for ‘mindful play’, skills for understanding their children’s cognitive and emotional states, and mindfulness techniques to promote sensitive caregiving in stressful parenting contexts (e.g. tantrums or prolonged infant crying). Being in the ‘right state of mind’ to manage difficult parenting situations, helping parents to develop coping skills and mindfulness strategies to reduce dysregulated affect aims to reduce coercive hostile parenting behaviours, make caregiving more nurturing and child focused, and enable a reduction in situational aggression between partners. 66

With regard to parental emotion regulation, the PuP parent workbook contains several treatment modules that aim to reduce dysregulation and psychopathology through the use of mindfulness exercises (e.g. the ‘how to manage emotions when under pressure: increasing mindful awareness’ module) and urge-surfing techniques for substance use issues (e.g. the ‘managing substance use problems’ module). In addition, this study provided an opportunity for fathers and their partners to develop communication skills and to co-regulate by identifying high-risk situations for situational verbal and physical aggression, drawing from the ‘relationships’ module. This component of PuP was initially undertaken with fathers alone and then extended to couples’ sessions. Although relationship work needs to be undertaken with great sensitivity and awareness of safety issues for both partners, the work of Stover40 indicates that this approach is acceptable for both partners. PuP includes exercises on ‘communication in intimate relationships’, and these were combined with modules on managing emotions to address interpersonal aggression between partners and, potentially, towards children.

Finally, self-regulatory skills are developed with children through combined sessions with the caregivers and child/children. These self-regulatory skills, again, draw from mindfulness constructs, with a growing body of evidence supporting the relationship between mindfulness and adaptive emotion regulation,67 particularly for young children with difficulties with emotion regulation. These skills are appropriate for children aged 3–5 years. As parents become more emotionally regulated, they are able to provide more sensitive caregiving. This, in turn, is associated with the development of emotion regulation in young children. 68 Therefore, the parent workbook supports the parent by allowing for documentation of their personal journey through the programme.

Parents under Pressure intervention delivery

The PuP programme is designed to be delivered face to face by the PuP practitioner in families’ homes. When there are concerns regarding practitioner safety, it can be delivered in community-based clinical settings. It has also been delivered in residential therapeutic communities and prison settings.

The PuP programme was delivered to fathers, resident or non-resident, with or without partners/mothers. Depending on the individual needs and circumstances of the parents, some modules were delivered by the PuP practitioner to the couple, whereas other modules were delivered individually to one or both parents. If the parents were not cohabiting, then the programme was delivered individually to each parent or just to the father if the mother did not wish to participate. Participation of the partner/mother in the programme was not dependent on whether or not they enrolled in the study. One mother chose to engage in the programme with the father, but declined to enrol in the study.

The time allocated to the family support plan varied and was dependent on the family’s needs and family support remit of the agency delivering the PuP programme. Appointments were generally held on a weekly or fortnightly basis for 1–2 hours. The PuP programme has been delivered over 646,49 and 12 months. 69

Parents under Pressure intervention providers

To use the PuP programme, the intervention provider is required to have undertaken an 8-month training course, which comprised an initial 2-day training in the overall model and programme followed by implementation support provided at 2, 4 and 8 months (see Parents under Pressure training programme). (Note that COVID-19 restrictions resulted in a 2020 online adaption of the PuP training.) Completion of this training leads to accreditation as a PuP practitioner. Practitioners are not required to have a professional qualification (e.g. social work or psychology) to become an accredited PuP practitioner. However, commitment to working with complex-needs families using intensive case management and an approach that encompasses therapeutic family support is essential. The intervention in this study was delivered by accredited PuP practitioners from two organisations (CIRCLE and PREPARE) who were experienced family support workers from a variety of backgrounds, including social work, early years practice and community education.

Practitioner training and accreditation in Parents under Pressure

Parents under Pressure supervision and implementation support

Expert staff supervision for practitioners in the study was provided initially by an accredited PuP supervisor and trainer. To ensure sustainability, the PuP programme training process includes the identification of senior practitioners or team managers who are able to take on additional training as an accredited PuP supervisor. Identified individuals undertake an additional 20 hours of training and support, working through a series of guided exercises and tasks. The purpose of this training programme is to provide an advanced understanding of the underlying conceptual framework of the PuP programme. A series of readings are provided and practitioners are asked to a complete set of exercises to practise the skills and demonstrate competency. This includes additional training in case conceptualisation, an understanding of the presenting issues for children with a range of developmental disorders, advanced readings and practice in mindfulness, and a comparison of the PuP programme with another parenting programme. Training is also provided in implementation support, including helping trainees to provide constructive feedback on case study requirements for accreditation as a PuP practitioner.

Reflections on our experience of patient and public involvement and engagement for this study

Engaging parents who are drug dependent in patient and public involvement and engagement (PPIE) presents a number of challenges, as they are a stigmatised and marginalised group of parents who often lack the confidence to participate in groups or committee meetings, and the families have child care responsibilities, but often do not have child care or the spare time to engage in PPIE activities. Engagement often needs to be on a one-to-one basis and in settings that are comfortable and non-threatening (e.g. home visits, coffee shops or over the telephone). Families are often engaged with multiple services and can be working in jobs with irregular hours, and so achieving contact within usual ‘office hours’ can be challenging. Therefore, extreme flexibility is required by the RFs. In addition, parents need to be reasonably ‘stable’ drug users (not intoxicated) to be able to contribute to discussions and to maintain contact with the study team. Even when they are stable and motivated to be involved in research, this population often lack the resources to contribute (e.g. time, effort, equipment and money, including sufficient credit to return missed calls and enough data to attend video-conferencing meetings). Therefore, considerable research resource is needed to engage this population in the research process in a meaningful and consistent way. Despite these difficulties, the involvement of parents in this study who were active drug users (prescribed OST and involved in drug treatment services) was really valuable, as they were able to provide up-to-date accounts of some of the challenges that fathers who use drugs face.

Participants: fathers and families

The target population for this study were families living in Lothian with children (aged 0–8 years) affected by paternal drug use, namely families with fathers/male caregivers diagnosed with opioid dependence and currently prescribed OST. We anticipated that most fathers who enrolled in the study would reflect the population of men normally engaged in OST (i.e. the majority were aged between 18 and 55 years, on a low income or unemployed and in receipt of welfare benefits, living in social housing in areas of deprivation and had a history of polydrug use and criminal justice involvement, including imprisonment).

Participants: Parents under Pressure practitioners, supervisors and service managers

Participants: referrers

Analytic framework

Feasibility for a main study was assessed in a number of ways. Quantitative and qualitative data, as well as project monitoring data, were used to assess pre-set progression criteria (see Appendix 5, Table 10) and all the study RQs and objectives outlined in the study protocol (see Chapter 4). In addition, we used the ADePT framework,54 a tool used to systematically identify and appraise issues and solutions in feasibility studies, to inform our decision-making on whether or not a future main study was feasible and justified.

To assess the degree to which our progression criteria were met, and to answer the specific feasibility RQs, the following quantitative and qualitative analyses were undertaken.

Partners/mothers enrolled in the study with father participants

Of the 25 fathers who were consented into the study, 19 were identified as living, or co-parenting with partners/mothers. Of these partners/mothers, 17 (89%) agreed to take part and were consented into the study. One mother did not take part because the father wanted to undertake the programme alone and the other mother did not want to take part in the study, but did want to engage with the PuP programme.

Fathers

Of the 25 fathers enrolled into the study, baseline measures were obtained for 23 (92%). The characteristics of the fathers who provided baseline data are summarised in Table 3. Fathers were aged between 27 and 52 years (mean age 39 years), the majority were white British and all were prescribed OST, as this was an inclusion criterion for the study. Fathers were prescribed methadone (n = 18) or buprenorphine (n = 5) and five fathers were also prescribed diazepam for benzodiazepine dependence. Fathers reported a history of problematic drug use for 4–30 years (mean 21 years). Most fathers had a current (female) partner (n = 17), but four were single (one because the mother of his child had died of a drug overdose) and two were co-parenting with the mothers of their children, but living apart. Most fathers were living in social housing or rented accommodation in areas of deprivation. Six fathers were homeless and living in temporary accommodation.

Fathers left school aged between 11 and 17 years (mean age 15 years). Data gathered on educational attainment indicated that 21 (84%) fathers had < 10 years of formal education, with six leaving school with no qualifications. Most fathers (n = 17) had a history of incarceration, and five reported current legal issues at baseline (for a range of criminal offences). Twenty-one fathers (84%) reported government benefits as their main source of income, one reported full-time employment and data on income were missing for one. Ten (43%) fathers reported current financial debts.

Mothers

Of the 17 mothers who were consented into the study, 17 provided sociodemographic data. Participating mothers were aged between 22 and 41 years (mean age 34 years). Of the 17 mothers who provided baseline sociodemographics, 14 (82%) were prescribed methadone, one was prescribed dihydrocodeine and two were non-drug users (discordant couples). The mothers receiving OST reported a history of problematic drug use for 9–20 years (mean 14 years). Mothers reported having left school aged between 12 and 17 years (mean age 15 years), with 13 mothers leaving when they were aged ≤ 16 years and four mothers leaving with no qualifications. Six mothers reported current criminal justice issues and six had previously spent time in prison. Thirteen (76%) mothers reported government benefits as their main source of income, one reported full-time employment and data on income were missing for two. Two (12%) mothers reported current financial debts.

Length of relationship between fathers and the mothers in the study ranged from 8 months to 16 years (median 7 years, IQR 7.0 years). The majority of fathers and their partners reported physical and/or mental health comorbidities (commonly pain, anxiety, depression, trauma/history of childhood abuse, domestic abuse, etc.) and low levels of wider family and social support (see Appendix 7, Table 14).

Children

Based on data reported at baseline, the 25 fathers who enrolled had a total of 51 children (aged 0–16 years). Fathers had between zero and eight children each (median 2 children, IQR 2.0 children; note that three fathers were expecting their first child). Fathers reported living with very few of their biological children [11/51 (22%)]: fifteen fathers had no children at home, six fathers had one child, one father had two children and one father had three children.

At baseline, the families reported a total of 47 children in the formal state care system (including children of both fathers and mothers). Thirty-eight children were living with kinship carers, four were in foster care and three were in residential care, and two had been adopted. Five fathers also reported having grown-up (adult) children living away from home. All fathers recruited to the study had regular contact with at least one child aged 0–8 years (referred to as the ‘index child’ in the remainder of the report).

The varied, complex and evolving nature of the circumstances of families in this study is illustrated in infographics of the families from baseline to FU2 (see Report Supplementary Material 3). Complex blended family structures, movement of children between places of care, formation and break-up of relationships, homelessness, unplanned house moves and incarceration were all observed in this sample of families in their time in the study.

Practitioners

Eight staff from PREPARE and CIRCLE were trained in 2016 in preparation for delivering the intervention as part of the study. However, only four practitioners were accredited at the start of the study in April 2017 (reasons for this included staff being on maternity leave or long-term sick leave, staff moving agencies, staff being promoted to managerial post and accreditation was derailed because of commissioning cycle). Both agencies committed to a further cohort of staff training in the PuP programme in 2017 and the intervention was eventually delivered by eight accredited PuP practitioners across the two organisations. To support implementation, both agencies opted to train one experienced PuP practitioner to PuP supervisor level, and this development helped to further embed the model in routine practice. Implementation support was provided to practitioners during formally organised sessions ran by the independent trainer/supervisor initially and then collaboratively with the PuP-trained ‘in-house’ supervisors. The formal PuP development training sessions were conducted approximately every 2 months across the study period. Of the eight practitioners who delivered the programme to participants, six were female and two were male.

Service managers

All service and team managers (n = 7) in CIRCLE and PREPARE were invited to take part in a qualitative interview during the research study and all agreed to take part. Interviews took place between 1 October 2019 and 28 April 2020. Managers came from a range of backgrounds, including social work, community education and management, and had been in their roles for between 3 months and 12 years (one senior manager who was instrumental in the study set-up was interviewed 3 months after leaving the organisation). Five managers were female and two were male.

Referrers

Referrers were contacted via their managers or directly by e-mail and telephone and so data on the total numbers invited and the proportion who took part are not available. However, none of the organisations or managers refused to assist with recruitment. A total of 28 professionals took part in four focus groups held between 12 March and 26 June 2019. Professionals based in East Lothian and all areas of the City of Edinburgh took part. A final focus group was planned for West Lothian, but was cancelled because it coincided with the start of the COVID-19 pandemic and, therefore, no data from referrers from West Lothian are available. However, key stakeholders from West Lothian were invited to attend the online expert event in October 2020.

Participation of a range of referrers was achieved, including community mental health nurses (n = 13), social workers (n = 4), family support workers (n = 3), health visitors (n = 3), third-sector drug recovery co-ordinators (n = 2), GPs (n = 1), clinical psychologists (n = 1) and midwives (n = 1). The age ranges of participants were 25–34 years (n = 8), 35–44 years (n = 7), 45–54 years (n = 9) and 55–64 years (n = 2). The mean length of experience working with drug-using parents was 9 years (range 1–20 years).

Fathers

The 20 fathers who were accepted into the PuP programme had a median engagement with their PuP practitioner of 182 days (i.e. 26 weeks) [range 7–344 days (i.e. 1–49 weeks)]. The median number of sessions was 13 (range 1–34 sessions, IQR 13.25 sessions). Seventy per cent (14/20) of fathers attended six or more sessions and were classed as ‘engagers’ (in line with classification in UK PuP RCT46). Non-attendance at scheduled sessions ranged from zero to 17 sessions (median 4 sessions, IQR 5.0 sessions). As a proportion of the total number of sessions offered, fathers’ attendance ranged from 11% to 100%, with a mean of 71%.

Mothers

Fourteen mothers accepted for the PuP programme had a median engagement with their PuP practitioner of 211 days (i.e. 30 weeks) [range 0–386 days (i.e. 1–55 weeks)]. The median number of sessions was 14.5 (range 1–29 sessions, IQR 19.5 sessions). Seventy-one per cent (10/14) of mothers attended six or more sessions and were classed as ‘engagers’. Non-attendance at scheduled sessions ranged from zero to 17 sessions (median 4.5 sessions, IQR 6.25 sessions). As a proportion of the total number of sessions offered, mothers’ attendance ranged from 0% to 100%, with a mean of 71%. As can be seen in Figure 6, within families, rates of attendance at sessions were similar for fathers and mothers.

FIGURE 6.

Fathers’ and mothers’ attendance as a proportion of PuP sessions scheduled.

Delivery of the Parents under Pressure programme to couples enrolled in the study

Fourteen mothers were concurrently enrolled in the study alongside the father. The number of PuP sessions delivered by practitioners ranged from 2 to 43 per couple (median 15.5, IQR 26.7). Practitioners (n = 8) delivered 248 sessions to couples [140 (56%) sessions were joint sessions, with the remaining 108 (44%) sessions being individual sessions for the father (n = 52) or for the mother (n = 56)].

For individual families, the frequency of joint sessions (with both parents present) varied from 0% to 100% (mean 60%). For sessions with fathers, the frequency of sessions that were joint ranged from 0% to 100% (median 76%). For sessions with mothers, the frequency of sessions that were joint ranged from 0% to 100% (median 85%).

Overall, there were no differences in the level of engagement in the programme between fathers and mothers.

Non-engagers and dropouts

Six fathers (and four mothers) attended five or fewer PuP sessions (classed as ‘non-engagers’). The following reasons were provided for non-attendance:

• One father was working shift work and was too tired (father and mother completed five sessions each).

• One family stopped engaging because of problems with their older child (father completed one session and mother completed none).

• One family stopped engaging when they experienced urgent housing issues (father and mother completed two sessions each).

• Two fathers stopped attending without giving particular reasons (one father attended one session but the mother did not attend and the other father, who did not have a partner, attended one session).

• One father did not attend as he felt that he and his family were doing well and did not need the PuP programme (the father completed one session and his partner attended three sessions, although she was not enrolled in the study).

Fidelity

The PuP programme is individually tailored to each family and so there is variability in the delivery of components (i.e. not all families needed all 12 modules). The following data illustrate parent reports of the PuP programme components that they received. Excluding the six non-engaging families (i.e. those who received five or fewer sessions of the PuP programme), fathers reported receiving a range of 9–20 components (median 15 components) of the PuP fidelity measure and partners reported receiving a range of 5–19 components (median 15 components) of the PuP fidelity measure (with a maximum possible score of 20). As shown in Figure 7, the PuP programme components most frequently endorsed by fathers as having been received were setting targets for change, having shared goals and discussing their child’s best qualities, whereas using relaxation techniques, using mindfulness and using wise mind were reported less frequently. For partners (mothers), setting targets for change, discussing their child’s best qualities and having a good understanding of child’s developmental stage were the components most frequently endorsed as having been received, and using wise mind was reported less frequently (Figure 8). For both fathers and mothers, formal components of mindfulness were less frequently reported, but, notably, one of the key elements of mindful parenting – ‘being in the moment with your child’ –was one of the most frequently reported. This suggests that, although formal mindfulness exercises, such as mindful eating and mindful walking, were endorsed less, the key to engaging with children by being fully present was endorsed by both fathers and mothers.

FIGURE 7.

Father-reported intervention components received.

FIGURE 8.

Partner (mother)-reported intervention components received.

Retention in the study

Treatment Outcomes Profile

EuroQol-5 Dimensions, five-level version

Brief Child Abuse Potential Inventory

Parenting Sense of Competence Scale

Difficulties in Emotion Regulation Scale

Revised Conflict Tactics Scale

Child well-being measures (Brief Infant Toddler Social and Emotional Assessment and Strengths and Difficulties Questionnaire)

Parent–child relationship (Paternal Antenatal Attachment Scale, Maternal Antenatal Attachment Scale and Emotional Availability Scale)

Barriers to and facilitators of uptake and retention

Referrers and practitioners reported that fathers who use drugs can be a very challenging population to identify. Their fatherhood status is not always known, even if they are attending drug treatment services. They were portrayed as a ‘hidden’ population who could remain ‘under the radar’, especially if their partner is not a drug user. Some practitioners also portrayed opioid-dependent fathers as being largely disinterested in parenting and parenting programmes:

[Fathers] don’t really want to engage with that type of programme [parenting programmes]; they feel it’s more geared towards the mums than themselves . . . I find it quite difficult to kind of engage with the dads . . . I’m working with the mum and the children, not really the dad.

Referrer, focus group 3

Even if fathers were known about, professionals suggested that approaching opioid-dependent fathers and talking to them about parenting can be difficult to do. The ‘stigma’ of being a parent who uses drugs and their fear of social services was not conducive to talking about getting involved in parenting programmes:

I was really glad that you said that [PuP is] strengths based, because I feel that . . . and it’s not about criticising social work in any way, but I feel like the interaction between our [drug service] patients and social work can be really negative and very stigmatising.

Referrer, focus group 2

Professionals needed to feel as though they could ‘sell it’ to fathers. In addition, professionals needed to have a good rapport with fathers before they would ‘raise the subject’. It seemed important for referrers that the intervention was supportive and empowering and was not likely to result in a negative experience, as many parents who use drugs have had previous encounters with children’s services that had been unhelpful:

I suppose from the very beginning, maybe selling it to dads and getting them . . . because the focus is on them and they’re not used to that maybe, you know, why this would benefit them . . . family support has really been, just all a whole load of discussions about what drugs they take, primarily . . . so no sense that they’ve, you know, been involved in setting their own goals and being an active agent in the whole process.

Practitioner 4

The focus groups with referrers and the interviews with practitioners and parents all highlighted gendered practice as a barrier to father-inclusive practice. Parenting and family support services traditionally focused on mothers and mothering. Fathers also assumed that most services were for mothers and children, and not for them. Fathers were often perceived and treated as optional, rather than essential, to the parenting and child welfare agenda:

[Practitioners] are very comfortable working with mothers, you know, that’s the norm . . . and even if unconsciously they don’t want to exclude fathers, they would . . . they’d find themselves kind of doing that anyway.

Practitioner 4

Participants suggested that uptake and retention in the programme were more likely if there was an organisational culture that supported a ‘whole-family’ and father-inclusive approach, and where practitioners understood the intervention and could emphasise the supportive nature of the programme (as opposed to psychoeducation-type parenting programmes that ‘tell parents what to do’). To overcome practitioners’ lack of knowledge about the programme and to build confidence in ‘selling it’ to fathers, there were suggestions for programme information materials that could include direct quotes from fathers with experience of the programme (e.g. on leaflets, as a short video, as a podcast or as a blog/short story). Professionals could then use these materials to demonstrate to fathers how the programme could be helpful for them, their children and family.

Optimal level of recruitment and consent

There were 44 fathers referred into the study, of whom 38 (86%) met the eligibility criteria. Of these fathers, 25 (66%) consented to take part in the study (i.e. 57% of referral numbers). For this population, this is a very reasonable conversion rate. In most cases, if the research team managed to speak to the father and meet them in person, then they were willing to take part and so were the mothers. However, one father was incarcerated during the intervention phase and four other fathers had criminal court cases and/or permanency proceedings instigated on account of serious levels of violence that made them ineligible to proceed to allocation for the PuP intervention after consenting into the study. Only one father withdrew of his own volition before allocation. These situations are perhaps more likely to occur with fathers than mothers in parenting research involving substance-using parents. If one subtracts these five fathers from the total number eligible, then the percentage to allocation is 52% (or 45% of all referrals). In a recent study in Glasgow involving the recruitment of ‘hard-to-engage’ mothers into THRIVE (Trial for Healthy Relationship Initiatives in the Very Early years), MacLachlan et al. 91 reported that 50% of women referred were recruited to the trial. This suggests that a realistic rate of referral to consent in a main study would be around 50% (i.e. double the number of participants would be required to ensure that enough patients were eligible, consented and allocated to the intervention).

One way of ensuring that a larger number of eligible fathers are referred into the study is to check that the study inclusion criteria is compatible with the inclusion criteria for the service that will be delivering the intervention. In our study, these ‘thresholds’ varied somewhat between areas and across services, with some fathers and families accepted into the programme where there were permanency proceeding in place, and others not. Although the decisions regarding the inclusion or not of fathers in the intervention were made for very good reasons, these referrals were quite challenging to manage. For example, when these fathers were included in the study, supervised contact arrangements for the father and child to meet with the PuP practitioner were not always sanctioned by the corporate parents (social services) and only two researcher meetings could be arranged to conduct the EAS video-observational measure during contact visits. Therefore, the practicalities of delivering the intervention with fathers who have very little contact with their children, and where the practitioner cannot work with the parent and child together, limits the delivery of the programme as intended, although, arguably, the delivery of the programme could change the course of proceedings. With one family enrolled in the study, this is exactly what happened and their child was returned to the parents’ home. Therefore, agreeing eligibility to join the study with the service delivering the intervention may be the most pragmatic way to manage consent with a population whose circumstances are complex and difficult to predict.

Optimal level of retention

Retention in the study was also very respectable for this population. Twenty-three fathers completed baseline measures and the first qualitative interview, 16 completed FU1 and a second qualitative interview and 13 completed FU2. In terms of our pre-set progression criteria, we proposed a minimum of 10 fathers to complete FU2 measures plus one qualitative interview.

Retention of those who were allocated to the intervention (n = 20) was also reasonable. Thirteen of 20 (65%) fathers were retained in the study for a period of up to 18 months. By applying all the learning from this study about strategies that can improve retention with this population (see Chapter 5), it would be reasonable to expect that we could reach a 70% retention rate, provided all fathers who consent are allocated to the intervention and contact between the researcher and the family is frequent (e.g. monthly).

In terms of retention in the intervention, 14 of 20 (70%) fathers completed six or more sessions. The median number of sessions was 13, the median length of engagement was 26 weeks and mean attendance was 71%. Again, these are high retention and attendance rates for this population. However, there was a great deal of variation in participant’s timelines through the study, from referral to consent, to baseline, to intervention start date and to follow-up time points. Allowance for these variations and measures to mitigate them would be important considerations in a larger trial. Time from baseline to PuP intervention start dates varied from 1 day to 10 weeks. Three of the four families (FA05, FA15 and FA19) that waited longest for a PuP practitioner to be allocated subsequently had very short engagement with the intervention, suggesting that timely engagement with a practitioner may be an important factor. Capacity issues were a factor in the delays and so ensuring adequate capacity and sustainability of intervention delivery in a main study will be important issues to address.

A number of families did not complete the PuP programme and reasons for this were not always available. However, for at least four families (FA15, FA17, FA19 and FA24), life events and crises had an impact on their capacity to fully engage with the practitioner at an early stage. Repeat non-attendance at appointments led to families being discharged from the service in all these cases. When interviewed at follow-up, one family (FA17 and MO17) expressed disappointment at the speed of this and a desire to re-engage, which involved re-referral and a lengthy delay. Planning mechanisms to accommodate short-term disengagement might be worth considering if the service can manage this in terms of their waiting lists, role and remit, and commissioning targets.

Attendance data

Results show that attendance data can be reliably collected from practitioners who deliver the programme, and with relative ease. We requested attendance data from practitioners by creating an attendance sheet (one for each father and mother participant with ID codes) that asked for the date of every PuP session scheduled and whether or not it was attended. Sheets were sent and returned by e-mail and practitioners referred to their case notes and diaries to complete the sheets. This approach resulted in no missing data (assuming that case records and practitioner diaries were accurate).

Completed attendance sheets allowed for the calculation of the total number of sessions delivered to the cohort (and to fathers, mothers and couples), as well as father and mother attendance rates, the number of sessions attended and the length of engagement in the programme. Attendance sheets also provided a useful insight into variability in the way the programme was delivered to different families, with some couples attending all sessions together and some families receiving a large number of individual-based sessions (i.e. father- and mother-only sessions) as well as couple sessions. We were able to conclude from these attendance data that fathers who enrolled in this study engaged in the programme in similar ways, and to much the same extent, as mothers.

Based on these findings, it would be feasible to obtain attendance data in a main study. These data would be required to further examine ‘dose’ response. The dose parameters could potentially be extended, for example, to include not only the frequency, duration and number of PuP sessions, but also the number and type of modules completed/not completed as per individual family case formulation. These additional data could enable a more nuanced analysis of dose response in relation to identified family needs and goals, and the degree of fidelity could be evaluated across the intervention dose parameters. This could be supplemented, for example, by asking participants about their views on the acceptability, feasibility and effectiveness of a programme in which the dose is tailored to a family’s needs or delivered more uniformly to accommodate service capacity. 92

NHS prescription data

Twenty-three out of 25 (92%) fathers consented to their OST prescription data being obtained from NHS records (two fathers did not consent). Two fathers withdrew from the study and so we requested data on 21 fathers and their partners (if opioid dependent, n = 13). Data were requested from NHS Lothian drug treatment services and participants’ GP prescriber if their drug treatment was managed in primary care. Our approach for obtaining the prescription data was to request this by e-mail to the secondary care addiction service lead pharmacist and by letter to GP prescribers.

Our approach was moderately successful. Approval to obtain data from the addiction service prescription database was granted by the lead pharmacist soon after requesting it in October 2020. However, pressures on the service because of COVID-19 meant that the data were not extracted and made available to the research team until 30 December (after several prompts). There were no refusals from GPs to provide data, but returns were incomplete. It is acknowledged that not all data could be obtained from GPs, as participants frequently changed their GP if they changed address and participants also moved between GPs and the addiction service for their OST prescription (e.g. if they needed to be titrated to a higher dose, the GP would refer back to the addiction service). Owing, in part, to this movement, and to the long time period of the request, addiction service prescription data were more reliable and complete than data from GPs.

Child protection data

Social Work Scotland data were requested from two local authority children and families services. We requested child protection data on two children (from two families) from East Lothian Council in early January 2020 (this request was sent to the chief social work officer, as instructed by Social Work Scotland). Our request was treated as a freedom of information request and processed as such. It was then passed to the data protection officer for the council. After some correspondence back and forth, we received approval on 17 February 2020. It was then passed back to the chief social work officer to extract the data and return to the research team. However, the COVID-19 pandemic took effect in March 2020 and we were unable to pursue the return of the requested data.

We also requested data on 10 children (from eight families) from the City of Edinburgh Council in January 2020. We were then asked to complete a research study ethics application to conduct research in the council. However, COVID-19 delayed the completion and processing of this request for many months. We decided to pursue this application again in October 2020 to test whether or not ethics approval could be obtained, in principle, as the majority of families in the study were located in this local authority. We were given approval in December 2020 and the request was then sent to a social work manager to extract the data and return to the research team. This resulted in a full data set for these children.

In summary, the feasibility of obtaining child protection data from social services records on index children in this study was successful in terms of obtaining informed consent from study participants (only two families refused) and receiving approval to access and extract the data from responsible local authorities. However, COVID-19 created unavoidable obstacles to the data being provided within the timescale of this study. In a future larger evaluation of PuP4Dads, assuming that it is not conducted within the context of a pandemic, it would be reasonable to conclude that data of this nature could be obtained, although retrieving these data may take some months.

Cost data

Only one father returned the service use survey sheets to the researcher (a full set). Parents were open to the idea of self-completing a service use questionnaire, which included questions on health care, social work, criminal justice and drug treatment utilisation, to test the feasibility of a conducting a future health economic assessment in a main study. Participants were asked to self-complete the questionnaire after each PuP session and return the questionnaire to the researcher at follow-up. Participants were given a supply of forms at baseline. Reasons for not completing the forms included ‘forgot to fill them in’, ‘had lost or misplaced the forms’ or ‘had moved house and lost them in the move’.

Practitioners in this study refused to administer the questionnaires and/or collect them, citing concerns over their responsibilities in terms of child protection if they were party to any of the information contained in the service use reports (e.g. if the parent recorded that they had taken the child to accident and emergency, but had not disclosed this information to the practitioner, then they would have to pass this information on to social services if the child was on the child protection register). After a number of parents failed to return the forms, the research team asked the practitioners if they would be willing to remind parents to complete the forms at each PuP session visit. Practitioners agreed to do this. However, this resulted in no further returns. This highlights the critical importance of researcher support for data collection with this population.

Acceptability of the measures

The majority of fathers and mothers reported that the study measures were acceptable, with no refusals to answer the questions asked of them. One father withdrew from the study after he completed all the baseline measures because he found the interview ‘intrusive’. He specifically objected to questions contained in the sociodemographic sheet (e.g. regarding his and his family’s income). He did not voice any objections to the validated measures themselves. One family declined to take part in the video-observational measure (i.e. the EAS), voicing concerns over confidentiality, and one mother declined consent for video-recording for the EAS measure and obtaining data from social work about child protection (this mother had experienced having a child removed from her care previously). A few other parents were reluctant to consent to researcher access to NHS and social services data, but were reassured when they were told that detailed child protection reports would not be obtained and their wider health records would not be available to the research team. A few parents required reassurance about confidentiality before they were willing to disclose illicit drug use (via the TOP measure). There were no objections to requests for criminal justice/offending data (contained in sociodemographic data and received via the TOP measure and service use survey). Some parents found the CTS2 long and repetitive and the language in the CTS2 challenging because of the explicit content and non-culturally relevant references (e.g. ‘rape at gun point’), although this did not stop parents answering the questions.

Some PuP practitioners voiced initial concerns over the CTS2 after the researcher told them it was ‘intrusive’ and ‘sexually explicit’. Practitioners felt that this may negatively affect their relationship with the family and, for the first 6 months of the study, these practitioners were reluctant to refer families. The PuP practitioners were reassured after the research team indicated that the fathers/mothers themselves voiced no strong objections to the CTS2 questionnaire and that completion rates were high.

Suitability of measures

Approximately half the measures used in this study were found to be suitable for the study population at every time point and were easily understood and appropriate for the circumstances of the parent. These included the BCAPI, the DERS, the TOP and the EQ-5D-5L. Other measures were found not to be not suitable (i.e. the service use questionnaire) or only partially suitable (i.e. the EAS, the SDQ, the BITSEA, the PAAS and MAAS, the PSCS and the CTS2).

Although the EAS (i.e. the video-observational measure of parent and child interaction) was acceptable to most parents (with 22 of 23 fathers consenting to video-recording), ultimately it proved infeasible to undertake for most families at most time points. Some participants were expectant parents when enrolled in the study and so the baseline EAS measure could not be obtained. Some parents did not have custody of their index child at FU1 and some did not have custody at FU2, and so, again, the EAS measure could not be obtained. In two of these cases, the corporate parent (social services) consented to the video measure being taken during a supervised contact visit between the father and child, but this took considerable time and effort to arrange. Nevertheless, it demonstrated that administering this measure when children are accommodated is technically feasible. However, the main obstacle to collecting video-recordings in this study was obtaining and complying with NHS R&D requirements for data management and security. This included purchasing an NHS laptop and having this set up by the NHS information technology department, negotiating to use the laptop to record video files on the device (this required Caldicott Guardian approval) and negotiating permission to store the data on NHS servers. This whole process took over 12 months before the first video measure could be undertaken. In addition, the NHS laptop (and hence all the videos) could be used only by the principal investigator (AW), who was an NHS employee. This meant that the principal investigator needed to attend all home visits (at baseline and at 6 and 12 months) where it was possible to obtain a video-recording. This required considerable flexibility because of the high rates of cancellations/rearranging appointments and so the feasibility of successfully obtaining the data was extremely low (hence the limited numbers of completed EAS measures). In addition, the EAS measure requires coding by trained EAS coders (to be undertaken by Griffith University) and this was not possible because of COVID-19.

The child development/well-being measures (i.e. the SDQ and the BITSEA) were partially suitable because the age range of children in this study (i.e. 0–8 years), the longitudinal nature of data collection (over three time points and spanning a minimum of 12 months) and the age limits for the use of the measures meant that infants aged < 12 months at any time point had no measures taken, some who were aged between 12 and 36 months completed the BITSEA only, and some completed the BITSEA and then the SDQ. In addition, the BITSEA contains several questions that assumes that the parent has custody of the child, as they relate to everyday child care/contact, and so this measure was inappropriate for several participants whose children were accommodated (e.g. in kinship or foster care). Therefore, very few index children in the study had the same measure administered at all three time points. In a larger evaluation, child development and well-being could be consistently measured across multiple time points using the SDQ, but only for children who are aged > 24 months at baseline. Arguably, a measure of child development/well-being for infants aged 0–24 months who are accommodated for substantive periods of time or hospitalised as neonates on account of neonatal abstinence syndrome (both of which can affect early development) may not reflect parental care provided by the study participants. Likewise, the antenatal attachment scales (i.e. the PAAS and MAAS) were easy to administer, but were administered only once at baseline (before the baby was born), as the EAS was the preferred measure after the baby was born.

Other measures were found to be unsuitable for some study participant because of their circumstances. The CTS2 (domestic violence) questionnaire was only partially suitable, as some fathers were not in a relationship at one or more of the data collection time points and so this measure was not applicable to them. Likewise, the PSCS was suitable for only those participants who were already parents. First-time expectant parents were unable to complete this scale. However, in a larger evaluation with multiple data collection time points, this measure could be suitable for expectant parents, as only the baseline measure would be missing data.

Burden on participants

The number of questionnaires administered (i.e. nine at baseline, 10 at FU1 and nine at FU2) and their length and complexity (i.e. the large number of questions per questionnaire and time taken to explain some questions) meant that administration of the measures was burdensome for many parents, with the length of time to complete each interview being approximately 2–3 hours. Many questionnaires broached highly sensitive topics, including illicit drug use, domestic violence, child abuse, criminal justice issues, parenting attitudes and values, and so could not be administered ‘cold’. The researcher needed to build rapport with the parents before launching into the data collection and so sociodemographic data were collected beforehand. Most parents had complex social circumstances and multiple health problems to report. This meant that the researcher needed to arrange two home visits with couples for each interview time point.

Administration of measures

Data were not to be obtained by asking the parents to complete the measures unaided or over the telephone, via the internet or in a clinic setting. The reasons for this were that some parents in this population had poor literacy and numeracy skills, some questions needed to be explained to parents (or parents asked for clarity on their meaning), some parents had a limited attention span/poor concentration and need to be prompted, and most parents would not return data if requested to do so. Clinic appointments are also not suitable, as attendance is often poor and parents do not have access to reliable child care (or they have limited child care) and so attending long appointments is not feasible. Administration by telephone or via the internet was not feasible because of the duration of administration, poor mobile telephone connections or unreliable internet access (i.e. no money for data), and many parents did not have internet access in the home or were not digitally literate. Last, many of the topics are sensitive in nature and the data are not appropriate to obtain remotely, as privacy and confidentiality could not be assured (i.e. other people present in the home). It should be noted that in line with policies on lone working and home visiting, two staff must attend the first home visit for safety reasons and to undertake a fieldwork risk assessment to establish whether or not it is safe to return alone. In addition, where the study involved both the father and mother enrolled in the programme, the researcher often had to visit the home on two separate days to interview each parent individually (to maintain confidentiality and also because one parent had to look after a child while the other completed the questionnaires/recorded interviews).

Feasibility of collecting data

Overall, most measures administered in this study were well tolerated by the participants, and completeness of data within and across measures were high. However, data collection was extremely time-consuming and resource intensive. Nevertheless, the methods we employed to overcome barriers to data collection were largely successful and we identified several strategies that could be employed in a future evaluation to improve completion and retention rates. Most notably, more frequent researcher contact with participants and fewer measures administered per contact. In addition, more rigorous methods of contact tracing participants who become unobtainable could be employed, along with telephone calls to complete short questionnaires (e.g. the ED-5D-5L) or small numbers of missing data (e.g. sociodemographics). Some repeat measures in a future study may be suitable for video calls (e.g. the BCAPI) if headsets could be provided to participants to offer privacy and mobile data costs could be paid for, although, as stated previously, digital literacy is likely to remain problematic in this population.

Parental child abuse risk

Evidence of statistically significant improvement in child abuse risk, using the BCAPI measure, was seen for 4 out of 14 (29%) fathers at FU1 and this was maintained at FU2. One of eight (12.5%) mothers at FU1 and four of six (67%) mothers at FU2 showed evidence of statistically significant improvement using the BCAPI. When the results of PuP programme ‘engagers’ were examined (i.e. those who had attended six or more PuP sessions), the pattern was similar.

Parental emotion regulation

Evidence of statistically significant improvement in emotion regulation, using the DERS measure, was seen for 7 of 15 (47%) fathers overall at FU1 and for 4 of 11 (36%) fathers at FU2. One of 11 (9%) mothers at FU1 and two of nine (22%) mothers at FU2 showed evidence of statistically significant improvement in emotion regulation. When the results of PuP programme ‘engagers’ were examined (i.e. those who attended six or more PuP sessions), the pattern was similar, but a larger proportion of fathers who engaged in the PuP programme maintained the improvement in emotion regulation at FU2 than the overall sample (40% vs. 27%, respectively).

Profile of change outcome measures for fathers and mothers is presented in Table 7.

Qualitative data on the impact of PuP4Dads

Fathers reported multiple benefits as a result of engaging in the PuP programme, which, to a large extent, were good and accurate reflections of some of the core components of the programme and the goals of the intervention, namely improved relationships with their children, a better understanding of their children’s needs, improved parental emotion regulation, increased parental self-efficacy, better couple relationships and co-parenting, increased social support, improved goal-setting and multiagency working and a greater understanding of the effects of their substance use on their children. Woven into many of the accounts from fathers were references to the therapeutic alliance with their PuP practitioner and the importance of trust, respect and collaborative working. The selected excerpts below illustrate some of these points.

The following quotation illustrates better understanding of children’s behaviour and developmental needs:

Oh, a lot better. More, a better understanding of his way of thinking . . . I never used to, it’s like, I was brought up with my grandparents, as I said, brought up where children should be seen and not heard. So, that’s kind of the deal I had at first with my other kids, and with [my boy] . . . and everything[s] changing. I’ve kind of, you know, it’s more about him . . . I can understand where he’s coming from as a kid, and not knowing things . . . I do get why, sometimes, they react the way they react . . . and not understanding . . . and security and safety, and how important that is, to feel secure.

Father 4

The following quotation illustrates a helpful approach to goal-setting:

Yes, instead of getting muddled up, likes the drug workers telling you ‘right, they want you to do this’ . . . and social worker this, this and this. Then you’re muddled up. You don’t know what you’re doing. If you’ve just got one set of goals to do instead of three or four different sets of goals, [if] you get muddled up and you forget to do some of it . . . it goes badly against you.

Father 11

The following quotation illustrates parents connecting with their children:

Just getting me to connect more with my daughter, ’cause I felt disconnected because of my anxiety, and not going out and that, not being able to play. Like, not a lot of dad stuff, going to school, go to the park.

Father 6

The following quotation illustrates parents building parental self-efficacy:

Just being able to have one-to-one time with someone who was a professional in that field and getting feedback from them about myself was . . . I found that really, really good; most of their feedback was . . . in fact just about all of their feedback was really positive.

Father 12

The following quotation illustrates parents connecting with their family and building social support:

I’ve started talking to my own family more, you know, these are just things that I was speaking to with [the PuP practitioner], yeah . . . I’m getting on a lot better with my mother than I did. I see my dad a little bit more than I did, you know. And we’re going to be looking at getting in contact with my other two children . . . that’s something we’re going to look at in the future, you know.

Father 13

The following quotation demonstrates emotion regulation:

Just like if he’s misbehaving, just . . . what’s it called, the pause button or something, so you’re not jumping right in and shouting or giving him a row. Just like stopping and thinking is it really that bad what he had done.

Father 6

The following quotation demonstrates therapeutic alliance and trust:

Honesty and trust, like, aye, being able to approach them and if things go wrong . . . then you know, that it’s just not going to fall apart . . . and they’ll work with you, you know, and make things better if things go wrong, and I think that’s the key thing there.

Father 9

The following quotation illustrates the impact of substance use on children:

[PuP has] given me a lot more insight into how my kids were, kind of, thinking, you know. Obviously, as I’ve said to you before, it’s the kids that suffer. I can sit and I can wax lyrical all day long about, like, woe is me, and how bad I feel; but really . . . learning a lot more about how it [substance use] affected them, that was really, really super helpful.

Father 23

The following quotation demonstrates a ‘whole-family’ approach:

I think, [names PuP practitioner], I think, he works more with the whole family, whereas other services they’re just working with [names his child] or like with our drug use.

Father 6

Some fathers directly attributed their engagement in the programme to getting their child home, having more contact with their child or being ‘given a chance’ to retain custody of their children:

They’ve even said it recently, so you are . . . what’s the right word, like a rare example of a child, a baby going into care, and how quickly you managed to turn it all around, and get her back home, and get yourselves straightened up.

Father 12

I think [PuP] should be made compulsory, especially if the social worker’s thinking of removing the children, I think it might be, like, giving the parents more of a chance, and having more things in place to help parents, rather than just stepping in and taking the bairn [child] away straightaway.

Father 2

Mothers reiterated many of the same positive aspects of the programme, but also made the point that the programme had helped them as mothers and as a couple because the programme engaged with the father and involved the father in co-parenting:

We did it at the same time, so we were communicating the whole way through, but yeah, I think we both got a lot out of it.

Mother 10

We are all doing [PuP] together, it’s something you can get out of together. Because everything is kind of separate at the moment, like [my partner]’s no really been involved in anything, it’s kind of all sort of been on me, and I found that a bit frustrating, and putting me proper under pressure sometimes, because I’m thinking ‘well why is everything end at me!’ and he’s kind of like being left in the background . . . everything’s kind of changing [now], he is involved and people are involving him in things and it’s kind of like, turned round a bit . . . [and] he’s involved.

Mother 13

In summary, the profile of change in individual fathers and families in this study, analysed using quantitative and qualitative data, provides an indication of change that is clinically relevant and important, and is feasible to measure and examine in a main study. Combined with our assessment of the delivery of the intervention, we revised the logic model for the study to account for maternal factors and change in relation to intervention outcomes (see Research question 11: which study design would best suit a future evaluation, including an economic evaluation?).

Progression criteria results

Pre-set progression criteria were assessed using project monitoring data, quantitative data results and the analysis of qualitative data. The results are summarised in Table 8.

Expert event (public engagement)

Our study protocol included a plan to organise an expert event with a wider group of stakeholders approximately 3 months before the end of the study. In this section, we describe the event and report the findings.

Decision-making using ADePT

The ADePT55 framework was used to guide the decision-making on the feasibility of a main study and whether or not it was justified. The ADePT framework involves the identification and appraisal of problems and solutions to a series of 14 methodological questions in feasibility studies. It allows the research team to consider problems and solutions related to the intervention, the context and the study design to be systematically appraised. Our analytic summary using the ADePT framework is presented in Appendix 16, Table 23.

When we considered problems and solutions to the main methodological issues in the study, we concluded that the results of the study indicate that a main study is feasible. The intervention itself was highly acceptable and suitable for the study population, it was delivered successfully within non-NHS settings and is a good fit for the current policy environment. Recruitment (including eligibility and consent) and retention were very challenging, although the results were more than acceptable for this study population and successful strategies were identified to improve rates. Participant engagement and attendance rates were very encouraging for this population and problems with allocation can be addressed with some revisions to eligibility criteria. Outcome measures were acceptable and well tolerated by the population and completion rates were high (using intensive data collection methods). Most measures were suitable and appropriate for most families. We identified a primary outcome measure (i.e. the BCAPI) and were able to calculate a sample size for a main trial. Interventions costs can be calculated and the feasibility of an economic evaluation for a main study was considered (see The feasibility of health economic evaluation in a main study). We considered the logistics for a multicentre study and tested these in a series of public engagement events, including scale-up in Scotland, England and Ireland. Last, all the components of the protocol, with some adjustments, worked well together, and much of the learning from the feasibility study could be incorporated in a future larger evaluation of the intervention.

In the light of our feasibility assessment and the success of involving the fathers’ partners (mothers) in the study, we revised the logic model for the intervention to include the following: mother’s affect regulation, mother’s problem-solving, couple relationship and co-parenting, increase in mother’s capacity to manage stressful situations and reduced child abuse potential as a short-term outcome (see Report Supplementary Material 4).

When we considered the question of whether or not a main study is justified, as well as being feasible, three key issues and decisions were identified: (1) whether or not the clinical need for the intervention justifies a main study, (2) whether or not there had been any significant advances in knowledge since the beginning of the study that would make a main study redundant and (3) whether or not the high costs to conduct a main study would be justifiable. The clinical need for the intervention is compelling and there is good evidence that the needs of this population are largely unmet and the adverse consequences are serious. Although the size of the study population is relatively small, the very high costs associated with not treating this population have been clearly demonstrated and there is growing evidence that targeted interventions for this population can be cost-effective. To the best of our knowledge, there has been little advancement in this field since the study began and only a handful of research teams worldwide are developing interventions specifically for this population (see Parenting programmes for parents with a substance use disorder). Therefore, a main study would be novel and innovative research. Last, the justification for funding a main study (and one that is likely to be high cost) is a decision that involves weighing up the evidence on the clinical need for the intervention and the costs of not treating this population, and deciding whether or not the investment could result in significant learning and short- and long-term benefits to the affected population of children and families who are the subject of this study. We would argue that it is, given that this population of families are among the most marginalised and disadvantaged in society.

The feasibility of health economic evaluation in a main study

Although qualitative interviews suggested no aversion to questionnaires on criminal justice and social service use among participants, only one participant completed and returned the self-report surveys (albeit almost in full) when asked to complete them in their own time. In the UK RCT of the PuP programme,46 in which similar questionnaires were administered by researchers, completion rates were much higher. Nevertheless, this suggests a willingness to provide such data, but an inability to manage the task at hand (parents forgot to complete and lost forms).

Permissions for access to routine NHS and child protection data were obtained, with few exceptions, and there were no refusals from NHS or social services to provide these data. However, the completeness and usability of these data could be only partially assessed, as the COVID-19 pandemic affected the ability to pursue missing data. Assuming that these data are, at least, partially usable, they have the potential to reduce patient burden and enable some elements of long-term follow-up, as participants were largely willing to consent, which could be taken at recruitment to study. In principle, consent would also be viable for in-person researcher access to GP and addiction service records to extract prescribing data and for other resource use, although the challenges around patients moving between GPs and addiction services, and changing address over the course of the study could make this challenging for any items where a central tracking identifier cannot be used. Therefore, it is likely that some resource use will need to be measured and/or validated by researcher-administered self-report. For example, the prescription data reported by parents in this study appeared to be reliably accurate.

On the health outcomes side of the cost-effectiveness equation, the EQ-5D-5L was acceptable and well completed by participants, and showed a small increase in index score (see Appendix 7, Table 14). Barlow et al. 46 observed a similar pattern. Although this finding is non-significant and underpowered, examination of the subscales driving this showed an improvement in the ‘anxiety and depression’ and ‘mobility’ scales (with the former being plausible and the latter arguably not). Consultation with the lead economist of the UK RCT of PuP46 (Stavros Petrou, University of Oxford, 2020, personal communication) confirmed a similar improvement in ‘anxiety and depression’, but not in mobility, implying that at least the mobility improvement may be an artefact of underpowering. Nevertheless, improvements in index scores in either study were not to scale with improvements in BCAPI or DERS (see Research question 9: is the profile of change in fathers, mothers and children clinically significant?), suggesting that the EQ-5D-5L may be underestimating improvements in health-related quality of life and potentially undervaluing the PuP programme in cost-efficiency calculations.

The matter is further complicated by outcomes, such as child abuse and parental drug use, being multifaceted and having long-term implications that are challenging to quantify. Many of these outcomes may be brought forward by successful interventions, minimising long-term cost and harm. However, for research with a limited follow-up period, a cost–consequence approach may be more appropriate for primary health economic analysis in such circumstances,96 with EQ-5D-5L being used as an opportunistic secondary measure alongside other consequences to most.

In summary, the RQ for the economic component of a main study would be:

• What are the economic costs and consequences associated with the implementation of PuP4Dads, in terms of health care, social care, child protection and criminal justice costs and outcomes?

Key research objectives and questions

Setting

Study sites would ideally include agencies that are already delivering the PuP programme, or an equivalent home-visiting family support service, to families affected by parental substance use and/or to families with complex needs. Sites in England, Scotland and elsewhere in the UK or Europe, where there is a high prevalence of opioid use and well-developed OST community-based treatment programmes, would be ideally placed to take part in a main study. PuP programme staff training can now be delivered online over an 8-month period and there is built-in implementation support that accompanies new adopters of the programme, and this would be essential for agencies that collaborate in a main study as a recruitment and implementation site.

Target population

For a main study, the target population would be opioid-dependent fathers and their families, as in the feasibility study. However, revised inclusion criteria would be required to maximise successful allocation to the intervention and, subsequently, retention in the study. In addition, participants in a main study would include PuP practitioners, supervisors, agency managers and the wider multiagency team around the child and family to build a clearer assessment of how well the intervention fits with current multiagency working and the wider policy and practice environment across sites.

Methods

Given the necessity for a quasi-experimental design, it would be important to estimate intervention effects within an appropriate causal inference framework to mitigate potential bias. We will use structural equation modelling to model both observed and latent variables, while accounting for within-participant variation, effect modifiers and covariance within a single model. We will incorporate repeated outcome measures (collected quarterly over 18 months) and investigate the direct effect of the PuP programme, potential mediators (e.g. parental emotion regulation and quality of caregiving) and dose–response (e.g. attendance rates and number of PuP modules). In addition to the observed measures of the primary and secondary outcomes, the models will include observed measures of parental mental health and drug use, child behaviour, staff/organisation indicators and demographic measures. Latent variables will comprise social support, abuse risk, co-parenting, therapeutic alliance and socioeconomic status.

Measures

The primary outcome measure will be the BCAPI to measure parental child abuse risk. Using data from the feasibility study, a clinically significant difference of 9 points on the BCAPI was determined using a RCI approach. 97 The feasibility study also estimated a standard deviation of 4.7 points for the change in BCAPI score from baseline to follow-up. Given these parameter estimates, and our structural equation modelling framework with an anticipated 15 observed variables and four latent variables, a minimum sample size of 116 participants will be required. This sample will provide 90% power to detect important differences over time (assuming a two-sided significance of 5%) and to ensure propriety, given the planned model structure. 98 Allowing for an anticipated 70% retention rate gives a total sample size required of 165 participants.

Other measures will include parental psychopathology, including transdiagnostic symptoms (e.g. emotion dysregulation due to substance use, trauma, depression and personality disorder), and quality of the caregiving environment (e.g. parent–child attachment and relationships, parental stress and the home environment), although as a latent variable to account for out-of-home placement effects. This could potentially be explored using a combination of social work child protection records and qualitative data as part of the realist evaluation, which could investigate complexity of home environments through the development of kinship maps for children.

Process evaluation informed by a realist evaluation approach99

The process evaluation will explore family, as well as organisational and policy, context to explore the key questions of what works, for whom, under what circumstances and why? The study will take the innovative step of seeking to interpret measured outcomes for a subset of intervention participants within the contexts in which they are achieved to explain the mechanisms of action that impede or facilitate success. Drawing on the QuaSIC (Qualitative approach to exploring SynergistIC interactions) method,100 we will explore the synergistic and catalytic impact of the intervention through conducting in-depth case studies with participants as the unit of analysis. This will explore experiences and views of the intervention, implementation and organisation (e.g. by including interviews with the diverse mix of staff involved in family support of individual participants) to explore unanticipated outcomes. This could be undertaken with a subset of five case studies, each with a purposive sample of six services (totalling 30 case studies). The purpose sample would ensure a diverse service mix to explore variations in mechanisms of action. Context–mechanism–outcome analysis would seek to situate outcomes within their variable contexts, thereby seeking a more nuanced understanding of how best to roll out the PuP programme at scale if the outcomes data suggests positive effects. Reporting of the results would adhere to RAMESES (Realist And Meta-narrative Evidence Syntheses: Evolving Standards) II criteria for reporting the results of realist evaluations. 101

Health economics component

In the light of our feasibility scoping, a cost–consequence analysis will be conducted to profile the health/social care, addiction, criminal justice, and child protection services, alongside key participant and child outcomes, such as parental child abuse risk, parental emotion regulation, parental drug use, child out-of-home placements and quality of caregiving. A detailed activity-based costing exercise102 will be undertaken to quantify the usual per-component cost and total cost of the delivering the PuP programme, and participants will be consented to allow data on each service type to be collected by researchers with access to case notes (where feasible), topped up with researcher-administered self-report for items unavailable and to validate data in the light of changes of address. Service use will be combined with standard UK price weights,103,104 applied where available, to estimate costs. We will additionally search the literature for cost-impact estimates of less tangible or longer-term issues, such as child protection and criminal justice end points. Rates of service use and cost for health/social care, addiction, criminal justice and child protection services will be compared between baseline and follow-up.

Potential to include a matched cohort identified via administrative data

The study will also explore whether or not routinely collected administrative data could identify a ‘virtual’ control group of drug-dependent fathers for the PuP programme-exposed participants. Administrative data sets do not routinely record fatherhood/male caregiving, but the potential for record linkage between paternal data (e.g. OST prescription data) and child records would be explored. If it were possible, it could result in a biased sample of biological, resident fathers who hold legal parental rights for children, as opposed to those who do not meet these criteria. Nevertheless, the feasibility of identifying an untreated comparator group for our study would be a worthwhile exercise.

2017

• 12 April.

• 10 May.

• 15 June.

• 6 July.

• 3 August.

• 13 September.

• 5 October.

• 8 November.

• 7 December.

2018

• 27 February.

• 22 March.

• 19 April.

• 28 June.

• 26 July.

• 20 September.

• 29 November.

2019

• 24 January.

• 27 February.

• 11 April.

• 23 May.

• 25 June.

• 22 August.

• 10 October.

2020

• 7 January.

• 3 March.

• 23 June.

• 13 July.

• 27 July.

• 13 August.

• 9 September.

• 30 September.

• 14 October.

• 11 November.

• 3 December.

• 15 December.

Chairperson of the Study Research Group

• Professor Anne Whittaker, University of Stirling (chief investigator).

Research fellows

• Dr Kirsty Killick, University of Stirling (RF until 31 August 2018).

• Dr Barbara Farquharson, University of Stirling (RF from 1 December 2018).

• Dr Heather Strachan, University of Stirling (RF from 1 January 2020).

Co-investigators

• Professor Lawrie Elliott, Glasgow Caledonian University.

• Professor Julie Taylor, University of Birmingham.

• Professor Sharon Dawe, Griffith University.

• Professor Paul Harnett, Griffith University.

• Andrew Stoddart, University of Edinburgh.

• Dr Peter Littlewood, NHS Lothian.

• Professor Roy Robertson, University of Edinburgh.

Service collaborators

• Michelle Kirkpatrick, Manager, PREPARE.

• Natalie Duffill, PuP practitioner and PuP supervisor, PREPARE.

• Rhona Hunter, Chief Executive Officer, CIRCLE (end date 30 March 2019).

• Mark Kennedy, Chief Executive Officer, CIRCLE (start date 20 April 2019).

• Lindsay Fraser, PuP practitioner and PuP supervisor, CIRCLE.

2017

• 13 June.

• 30 November.

2018

• 16 May.

2019

• 10 January.

• 27 June.

2020

• 9 January.

• 2 July.

• 22 November.

Note that a November 2018 meeting was deferred while the study was paused.

Chairperson of the Study Steering Committee

• Mr David Liddell, Chief Executive Officer, Scottish Drugs Forum.

Members

• Professor Christopher Weir, Professor of Medical Statistics and Clinical Trials, Edinburgh Clinical Trial Unit, University of Edinburgh.

• Professor Thomas McMahon, Professor of Psychiatry and The Child Study Center, Yale School of Medicine.

• Dr Tessa Parkes, Senior Lecturer, University of Stirling.

• Ms Anne Neilson, Director of Public Protection, NHS Lothian.

• Dr Alison Munro, Senior RF, University of Dundee.

• Dr Claire McIntosh, Consultant Psychiatrist in Addictions, NHS Forth Valley.

• Ms Gwynne Rayns, Senior Social Worker, City of York Council.

• Mr Chris Cuthbert, Director of Learning and Development, A Better Start (previously National Society for the Prevention of Cruelty to Children).

• Professor Brigid Daniels, Dean of the School of Arts, Social Sciences and Management, Queen Margaret University (start date 9 January 2020).

• Dr Catherine Nixon, Research Officer, Scottish Children’s Reporter Administration, University of Glasgow (start date 9 January 2020).

• Parent representatives (one mother and one father).

In attendance

• Professor Anne Whittaker (chief investigator), University of Stirling.

• RFs and other co-investigators.

Chairperson of the Data Monitoring and Ethics Committee

• Dr Christine Jones, Lecturer, School of Social Work and Social Policy, University of Strathclyde.

Members

• Dr Vicky Ryan, Statistician, Institute of Health and Society, Newcastle University.

• Professor Bernie Carter, Professor of Children’s Nursing, Faculty of Health and Social Care, Edgehill University.

• Professor Tony Long, Professor of Child and Family Health, School of Nursing, Midwifery, Social Work and Social Sciences, University of Salford.

• Professor Graham Connelly, Senior Lecturer, School of Social Work, University of Strathclyde.

In attendance

• Professor Anne Whittaker (chief investigator) and RFs.

Category: feasibility/deliverability/adoption/implementation

Category: acceptability of the Parents under Pressure programme/suitability of the programme for father

Category: impact of the Parents under Pressure programme