Extent of the problem
Better care means that people of reproductive age who suffer from a mental illness are more likely to start families and become parents.1 Consequently, the number of children and adolescents living with experience of significant parental mental illness may be increasing. Population data from the UK seem to bear this out. Abel et al.2 report a significant increase in the prevalence of children and adolescents living with serious parental mental illness (CAPRI) in the last 30 years. In 2007, the number of children in the UK exposed to a mother with a mental illness diagnosed in primary care was 22.2%. By 2017, this figure had increased to 25.1%. Over the same period, the proportion of children exposed to maternal non-affective psychotic disorder such as schizophrenia increased from 0.16% to 0.20%, and the proportion living with maternal affective psychotic disorders increased by 50% (from 0.25% to 0.37%). Currently, by the time a child reaches the age of 16 years, there is a 53% chance that their mother will have experienced a mental illness that has come to services’ attention.2
These findings suggest that parental mental illness is an increasingly common factor in the lives of children in the UK (note that these figures exclude paternal mental disorder burden). Therefore, given the scope and magnitude of the problem, and with similar estimates reported elsewhere (e.g. Australia,3 the USA,4 Canada5), the need for better information and more resources to support these young people has never been greater.
Improving the lives of CAPRI has become an urgent public health priority.6,7 This vulnerable group is likely to experience significant difficulties on a daily basis and throughout their lives,8–10 as a result of which they are likely to suffer a poorer quality of life and utilise services more than children living with ‘healthy’ parents.11,12 Recent epidemiological data13 show that CAPRI average one more general practitioner (GP) or nurse consultation per year than their peers, and the excess health utilisation cost to the NHS has been estimated at £652M per year. These combined facts present a strong public health argument for better early and preventative care.
Mechanisms of risk
As a group, CAPRI have, overall, been reported to show poorer outcomes than their peers across a range of domains.14 Mechanisms of this effect vary depending on the outcome, but include social influences of multiple deprivation and life stressors,14 lack of parental support15 and parental difficulties in combining the management of their own mental health with caregiving for their children.16,17 Such environmental effects interact with genetics,18 as well as with in utero19 and obstetric events;20 consistent reports find an association between prenatal maternal conditions, including psychopathology, and offspring neurodevelopmental abnormalities.21 Direct effects of parental mental illness may vary and may be less detrimental than the social adversities associated with mental illness, such as poverty, multiple deprivations, living in a single-parent family and repeated parent–child separations.22,23 Greater exposure to family discord and parental hostility is also reported to increase risk;24 one study25 reports that children of depressed mothers with antisocial personalities fare much worse than children of mothers with depression alone. Other important factors, such as parentification, neglect, maltreatment and domestic violence, compound unmet developmental needs.6,26
Although it remains unclear how age interacts with CAPRI risks, timing of parental mental illness and associated adversity across a child’s life is likely to be important; early exposure may mean that children are exposed for longer, whereas exposure during adolescence may influence critical developmental periods. Evidence suggests that vulnerability exists during pregnancy,27 infancy,28 throughout childhood and adolescence24,29 and into adulthood.30,31 This means that CAPRI may be exposed to a clustering of multiple adverse influences over time22,26,32 and, in many cases, effects are additive or interactive with independent direct effects of parental mental illness.25,33,34 But the well-being of these young people should not be seen as corresponding to ‘the ebb and flow of parental mental health’, as child difficulties may persist long after parental symptoms abate,24 although risk heightens when the parent experiences hospitalisation or an acute phase of illness.35 This especially affects children who do not have another parent or relative to look after them when this occurs.3,35,36
Most research on the relationship between parental mental illness and adverse child outcomes revolves around the mentally ill mother,37 but evidence on the independent effects of paternal psychopathology is increasing.38,39 For example, in their early years, children with depressive fathers are twice as likely as unaffected children to develop behavioural and emotional problems.40 In adolescence, the effect of having a father with depression either parallels41 or outstrips the effect of having a mother with the same condition, the latter being more influential earlier on in terms of shaping child outcomes.42
It should be noted that, even with this evidence, we cannot assume that all the same factors act as promoters or protectors in relation to all outcomes in all individuals. Rather, it is likely that many individuals show resilience across a range of circumstances and across a range of outcomes, whereas some show little resilience in few and not in all circumstances (for further discussion see Rutter,43,44 and Cicchetti and Curtis45).
Implications for CAPRI lives
With these caveats in mind, evidence suggests that CAPRI do less well across a range of life outcomes than their peers.7 These outcomes include poorer physical health46–49 and more behavioural and emotional difficulties,9,10 which also contribute to poorer educational outcomes.50–52 Entering adulthood, there is greater susceptibility to socioeconomic difficulties, alcohol/substance misuse and premature death.31,53 Health care and mental health service utilisation is significantly higher among CAPRI.11,12 Whether this always reflects greater ill health or more need for support by parents is unclear.
Living with parental mental illness also places a demonstrable strain on quality of life. CAPRI frequently grow up in environments of high family conflict,54 stress,55 and maltreatment and neglect,6,26 all of which amplify risks for impaired social functioning.56 Several studies demonstrate that CAPRI experience more severe interpersonal difficulties than their peers, including rejection, victimisation and not being liked.30,31,57 This vulnerable group is reported to show higher levels of internalisation of problems (manifesting as, for example, anxiety, crying, withdrawal or quietness) that make socialisation difficult and can lead to stigmatisation or bullying by other children.56,58 In addition, CAPRI suffer socially because of parents’ potential inability (or unwillingness) to support their child’s engagement in social activities. This may deprive children of opportunities not only to make friends, but also to take part in school/afterschool activities or sports teams that would improve their interpersonal skills.56,59,60 CAPRI may assimilate maladaptive social parental behaviours, which influence interactions with their peers outside the home.61 Higher rates of autism62 and other neurodevelopmental disorders63 may also contribute to the social problems that they experience. Social difficulties experienced by CAPRI are neither specific nor exclusive to their parental mental illness exposure and are similarly associated with parental physical illness, parental incarceration and premature parental death.24 Notwithstanding, the particular stigma of having a parent with mental illness is specific and may lead children and adolescents to fear developing a mental illness themselves.
But risk and resilience cannot be viewed as constant general traits among children exposed to parental mental illness. Resilience following adversity may be found for one outcome but not every outcome, or indeed any outcome of relevance to children themselves.43,44 Thus, our approach to improving health-related quality of life (HRQoL) is based on the following premises: first, that we must ask CAPRI themselves what their needs are in relation to HRQoL; second, all CAPRI (including those not expressing risks such as mental illness or behavioural problem as detailed in literature) will benefit from such a CAPRI-informed intervention to improve HRQoL; and, third, such an intervention requires a public health approach that is scalable to all in the risk subset (i.e. all children with parental mental illness).
In summary, most CAPRI will not develop a mental illness; many will perform well in school and sustain relationships,8,15 and not all children living in the same affected family will be influenced in the same way.64 However, multiple factors contribute to variance in risk versus resilience for a range of outcomes in individual children;65 for example, the availability of alternative familial, social, health and cognitive resources appears to benefit children.66 The severity and duration of parental mental illness are consistently strong predictors of outcomes; children of parents with enduring and severe disorders exhibit the highest risks expression.10,24 Children are less likely to experience problems if they have access to a healthy parent, one who acts as a ‘buffer’ between the child and the behavioural and emotional difficulties of the ill parent.67 Mental health literacy, social connectedness and self-efficacy also equip children with internal resources to manage difficulties associated with parental mental illness on a day-to-day basis.66
Thus, the evidence paints a picture of CAPRI as vulnerable, but with a heterogeneous set of needs that vary widely between individuals, between circumstances and over time, which means that some children require significantly more help than others. Services must understand such heterogeneity and variation in the planning of their response to unmet needs. For more complex interventions to be effective and cost-efficient, such understanding and targeting is essential.18 No interventions currently target quality of life (QoL) in CAPRI and virtually none has been child, as opposed to parent or family, centred.68 Currently, poorer intervention efficacy/cost-effectiveness is attributed to limited resources, budgetary restraints and skill limitations among practitioners, whereas inadequate screening and assessment processes are likely to contribute significantly.26 Importantly for future planning, lack of multiagency collaboration between adult mental health services, Child and Adolescent Mental Health Services (CAMHS), social care and child protection services means that support lacks co-ordination and leadership.17,68
Although many parents want to consult with professionals about their children,69 some are reluctant to do so because they fear that they may be judged or lose their children to care.70 Such factors are likely to mean that CAPRI remain hidden because, unless they come forward themselves, they may not be identified as having a need by services whose focus remains the ill parent/parents.17 We agree that a ‘fundamental paradigm shift is required at all levels of service development, delivery and policy’71 lest the needs of these children remain unmet. However, unlike these writers, our perspective is specifically focused on prevention for non-clinical unmet need and aligned with enhancing daily HRQoL across childhood for all CAPRI and not simply those at highest risk.
Background and rationale for the Young SMILES programme
The European Union’s (EU) Child and Adolescent Mental Health in Enlarged Europe initiative urged a change in political systems, legislative systems and health and social care systems to acknowledge and attend to the needs of CAPRI.72 This initiative also called for better information on CAPRI and how to target vulnerable groups before their life trajectories are disrupted.72 However, in spite of adequate policy guidance50,73 and a strong case for early intervention to support these young people,74 little reliable or child-centred evidence for intervention was available to take the process forward.68,75 Therefore, the National Institute for Health Research (NIHR) Health Technology Assessment (HTA) programme posted call 14/29, which posed the following question to researchers:
Is it possible to develop a community-based intervention to enhance the health-related quality of life (HRQoL) of children who live with a primary carer with serious mental illness, and would it be suitable for a future trial?
This report examines whether or not the Young Simplifying Mental Illness plus Life Enhancement Skills (SMILES) programme achieved its primary aim to answer that question.
The 14/29 call for interventions aimed at enhancing HRQoL among CAPRI derived, in part, from our HTA-funded systematic review.68 This had shown that there is little or no community-based provision for CAPRI, let alone any high-quality evidence or child-centred approaches to HRQoL. The likelihood of these young people developing mental illness, although greater than that of their unexposed peers, is outweighed by the likely influence on QoL of daily hardship and multiple deprivation.7 Thus, the rationale for creating and piloting a novel intervention to improve CAPRI HRQoL was driven by this understanding alongside the knowledge of the significant and growing numbers of exposed and vulnerable young people in the UK today.7 In the UK and the wider EU, we were aware of only two models aimed at supporting CAPRI directly, as opposed to supporting their parents.76 Our task, therefore, was to create, with stakeholders, a viable approach to a significant public health problem with a view to improving resilience and reducing long-term effects of poor HRQoL in children and adolescents with serious parental mental illness.
The 2014 NIHR HTA programme call drew broadly on our own HTA systematic review of interventions for CAPRI.68 This review had been widely informed by three focus groups with 19 different stakeholders: eight representatives from the children’s charities Barnardo’s, Young Minds, the National Children’s Bureau, the National Society for the Prevention of Cruelty to Children (NSPCC) and the Fairbridge Trust; five independent parents (four mothers and one father); and six CAPRI.6,68,76 Key elements specified in the original HTA call were that the intervention should be community based and delivered to children across the age range living with severe parental mental illness with the express intention of improving HRQoL.
Quality of life in CAPRI
As suggested above, the HRQoL model may reflect the needs and priorities of CAPRI better than other health-related outcomes, such as mental and behavioural disorders. It is increasingly recognised that ‘well-being’ refers to more than absence of disease, with QoL being increasingly seen as a valid clinical outcome for services.77 With recent UK50,73 and European72 policies highlighting the need for clinicians to consider young people’s perceptions of their own life experience within the context of their personal goals, expectations and priorities, a HRQoL model is anchored in the lived realities of children, rather than being service driven with disease-specific outcome measures.6 Bee et al.68 highlight five core life domains in QoL measures for children: (1) physical health, (2) emotional health, (3) social function, (4) material well-being and (5) environmental well-being. However, evidence is needed for a HRQoL that specifically captures the experiences of children living with a mentally ill parent.74
Evidence for interventions to improve health-related quality of life in CAPRI
Existing interventions concerned with parental mental illness usually target the affected parent.68 They include various modes of service delivery, including individual support78 and peer support groups,78 online courses79 and psychoeducational programmes.80 The aim of these interventions is to enhance protective behaviours in the parent, usually the mother.68 Recently, there has been a shift away from parent-based interventions to interventions centring on the needs and preferences of the family.68,76 Numerous interventions with a family-centric model of delivery report positive outcomes for child well-being.9,58,81,82 Although encouraging, such evidence should be treated with circumspection as studies reporting the largest effect sizes are invariably poorest in quality.75 High-quality studies are still in the minority, but also consistently report modest effect sizes. Furthermore, only few studies contain any analyses of longitudinal effects, making it difficult to monitor effectiveness over time.75,83
Reliable data on the efficacy of child-centred interventions are conspicuously lacking.68 Most data derive from small, biased samples reporting effects too small to enable them to be considered cost-effective.75 Efficacy might be underestimated or go undetected by aggregating effects produced by interventions across at-risk and resilient children18 (i.e. high- and low-risk individuals within the risk subset). Furthermore, most follow-up in studies is limited to < 1 year from baseline; monitoring effects over extended periods allows ascertainment of whether or not an intervention is acting to prevent difficulties in children who are at risk.84 Potential effect moderators need to be explored; adequately powered samples might then identify which subgroups will be more responsive or sensitive to an intervention. Such moderators may include a child’s sex, age, parental diagnosis, and socioeconomic status.84,85
Similar methodological problems apply to interventions aiming to improve the HRQoL in CAPRI. Bee et al.68 demonstrated a lack of reliable evidence to support the effectiveness of any existing interventions. Many included studies were conducted over 20 years ago, mainly outside UK or European settings, which limits their generalisability to the NHS. Furthermore, most studies focused on parents with mild to moderate (postnatal) depression; only three considered parental serious mental illness (SMI). Interventions were not child centred and did not consider young people’s QoL; all focused exclusively on parental outcomes.68
The review concluded that further work is needed to develop and evaluate child-centred interventions that improve HRQoL in these young people.68 However, the challenges implicit in this recommendation are numerous and complex. One such challenge is deciding when, where and how best to intervene, while demonstrating clinical effectiveness, cost-effectiveness and the potential for non-stigmatising and non-threatening delivery of specific interventions for CAPRI. In our view, interventions should place emphasis on mental health promotion and early intervention (i.e. public health approaches), while putting young people’s preferences and priorities at the centre of service delivery.6,17,68,76 Such preventative and early interventions should use models that are strength based and focus on resilience and, as such, better reflect CAPRI’s short- and long-term goals. Delivery should take place within community settings, such as schools, community centres, the home and other venues embedded within the child’s routine.76
Developing the model for Young SMILES
We considered the two existing models of working with families with parental mental illness, KidsTime and Family SMILES, not only because of their suitability for adaptation, but also to take account of expert reviewer and HTA Board comments from the original grant proposal, and because of the 3-year time scale of the proposed call. We knew that both of these group-based interventions ran successfully with CAPRI families in community settings but that neither had been evaluated and that they were not designed explicitly to enhance CAPRI HRQoL. Consultation with stakeholders and NSPCC collaborators provided us with permission to adapt Family SMILES to focus on a more child-centred approach and to work alongside the NSPCC to develop a non-NHS platform of delivery through the third sector. This was important for us because we recognised that the need for a new service fell into a preventative rather than direct health care domain and, therefore, potentially sat outside a CAMHS or NHS remit. This meant that the feasibility study needed to consider willingness to pay for, arguably, a public mental health approach in an increasingly large group of children and adolescents, most of whom, at least from a mental health perspective, would remain well. We were asked to test the suitability of the new intervention for a future trial. This meant that the intervention we developed needed to be manualised, staff needed to be trained, participants needed to be randomised with a control group and, where possible, assessors needed to be blinded. With an eye to future cost-effectiveness and sustainability in community settings, which we anticipated would not be traditional NHS settings, it was agreed that the established national network of family centres offered by our NSPCC collaborators represented an ideal setting to fulfil the brief of the 14/29 call.
We saw the need to design an intervention that responded to the changing shape of service provision and stood a chance of sustainability and scalability in economically stringent environments. To this end, we determined that delivering an intervention in different community settings and to different age groups would best be achieved by partnering with established third-sector community providers and with NHS Improving Access to Psychological Therapies (IAPT), which was beginning to broaden its remit to work with children and families.
Chapter 2 provides an overview of the policy landscape behind a drive towards better HRQoL for CAPRI and, in this context, sets out the aims and objectives of the final Young SMILES programme against which we report.