The extent of the problem

Better care, deinstitutionalisation and the development of community-based mental health services has increased opportunities for those with mental illness to begin and sustain families. Less severe, but more common, mental disorders such as affective or substance misuse disorders may be less likely to be visible to services; this means much parental mental illness may remain undetected. Previously, CAPRI have been studied extensively to calculate the heritability of mental disorders, especially schizophrenia. UK estimates based on those attending adult psychiatric services suggest at least 10% of mothers and 5% of fathers have a mental illness.² However, parents may not seek support from health providers because of concerns about removal of children to care, meaning these figures are likely to be conservative.

We examined the prevalence of CAPRI in the UK, utilising a primary care health register.^{Reference Abel, Hope, Swift, Parisi, Ashcroft and Kosidou1} Currently, 1 in 4 children are exposed to maternal mental illness and this number appears to increase with child age. The cumulative risk is such that by the time a child is aged 16 there is a 53% chance they will have experienced maternal mental illness.^{Reference Abel, Hope, Swift, Parisi, Ashcroft and Kosidou1}

Evidence of need

Our previous population-based research^{Reference Abel, Morgan, Tsuang, Cohen and Jones3} details significant premature mortality from all causes up to the age of 30 years among CAPRI, especially where maternal admission for substance misuse occurred. The reality is that mortality is too rare an outcome to drive new policy and services but little information is available about more common, tractable outcomes like physical health, social/criminal justice or cognitive ability, let alone quality of life and well-being. Most research focuses on mothers, ignoring the father's role in the child-rearing environment and most does not account for the multiple adversities CAPRI experience growing up with parental mental illness.^{Reference Abel, Hope, Swift, Parisi, Ashcroft and Kosidou1,2} We do know that these young people experience significant disruption to their lives from repeated parental admission to hospital and illness. The available evidence indicates that the lack of usual parental supports and disruption of family life impairs educational, social and emotional attainment and is associated with behavioural difficulties.^{Reference Abel, Morgan, Tsuang, Cohen and Jones3}

Information fit for purpose

Although a range of national policies now identifies CAPRI as a priority, there is a clear, urgent need for better information so that policymakers can allocate resources early and intelligently towards those most at risk. Specifically, we require information on the effect of parental mental illness at different times in a child's life, and on how resilience and risk are expressed among CAPRI. Population samples provide for such evaluation, allowing us to examine how multiple, correlated factors influence both good and bad outcomes. For example, CAPRI are well-known to exist in conditions of greater poverty where familial adversity tends to cluster, e.g. premature parental death and/or substance misuse. Although prospective population sampling allows for investigation of multiple resilience/risk factors, many confounding factors within families remain unmeasured and this limits our ability to attribute causal effects directly to parental mental illness. Examples of important ‘missing information’ include quality of parenting or variables on parental maltreatment or neglect, exposure of children to interpersonal violence and/or school bullying. These data are not routinely collected to a high standard in clinical or health survey data so remain unavailable for linkage within whole population data-sets. Administrative, health survey, clinical and cohort data need to be coordinated to allow better quality research with the appropriate detailing of key variables. We agree with the Academy of Medical Sciences⁴ that current administration of data governance is not fit for purpose. CAPRI themselves could be a powerful lobby to challenge and unblock information governance restrictions, meaning better or more routinely collected mental health and social data at, for example, primary care or school entrance. In our view, this is crucial if we are to find reliable potential markers of resilience/risk at the earliest opportunity.

Identifying need

Child and Adolescent Mental Health in Europe calls^{Reference Braddick, Carral, Jenkins and Jané-Llopis5} for better information and evidence to provide adequate services and interventions for children with parental mental illness. In our view, young people should be at the heart of decision-making about service development and the prioritisation of their (primarily non-clinical) needs. Most children with parental mental illness are not receiving any services; if resources flow to the wider family, they do so through the ill parent who may have a social worker and multidisciplinary mental health team.² Our work^{Reference Bee, Bower, Byford, Churchill, Calam and Stallard6} suggests CAPRI themselves want and need basic day-to-day supports, rather than specific interventions for an illness. However, barriers to sharing information between adult mental health and other agencies means early identification of CAPRI need remains challenging. This is a missed opportunity: CAPRI could be readily identifiable through their parents' contact with psychiatric services as well as through schools and primary care. The NHS and PHE could play a key role by engaging with CAPRI and other stakeholders to challenge services to work together and improve routine, non-stigmatised identification of their needs. Standardising communication and information sharing across schools, mental health services and social services supports the remit of a reinvigorated mental health service data-set which might offer solutions to the curation of CAPRI data across settings.

Identifying risk

Although this group faces adversity many demonstrate resilience, so a major challenge is to identify what fosters resilience in some and what places others at highest risk. Asking young people what helps them day to day must be at the heart of our understanding. Clinical risk modelling applied to this group must be informed by CAPRI for better understanding of how resilience and risk maps onto particular life outcomes. In cardiovascular medicine, such algorithms successfully identify resilience and risk to enact preventative interventions. We believe such techniques can be exploited in non-clinical samples of CAPRI. Furthermore, the focus of public mental health should not be limited to service-linked outcomes; little is known about the quality of life for the majority of CAPRI who do not come to the attention of services and who may wish to remain outside traditional service models. Consultation and co-development of outcomes with CAPRI is vital for success.

Service landscape for CAPRI

Globally, only Australia has a well-recognised policy and service platform dedicated to CAPRI. Most countries remain unaware of them and their support needs. Part of the problem is that, once children are identified, lack of information about modifiable factors driving CAPRI outcomes means few dedicated, evidenced, child-centred or cost-effective programmes exist. Recent reviews found contemporary interventions were all parent/family centred and small effect sizes, costly training and time commitments restrict delivery. Most interventions explicitly aim to reduce risk of later psychopathology and fail to consider other child-centred outcomes. Developing an evidence base for cost-effective interventions requires targeting the highest risk children within families and this means having better predictors of resilience and risk over time. It also means identifying young people's own outcome priorities. Initiatives that facilitate CAPRI as citizen researchers are vital to achieve this and they have encouraged young people to co-develop research priorities and child-centred interventions which are both acceptable to them and feasible to deliver within the current UK service contexts.

Future research

At the University of Manchester, we have adopted an interdisciplinary approach to this broad problem with a children and young people's advisory group at its heart. These citizen researchers help us formulate questions in four domains. First, we are creating parallel population data-sets held in Sweden, the UK and Australia to determine prevalence of CAPRI. Second, we are tracking broad, non-psychopathological outcomes (informed by CAPRI) across the life course. This provides the valuable population evidence currently unavailable and allows for the timely identification of CAPRI at highest risk for preventative interventions for more common adverse outcomes. Advisory group members will prioritise and modify how we frame questions about what social adversity and resilience means. Third, we shall map the risk model (informed by our CAPRI advisory group and the epidemiology) onto a clinical cohort of 9- to 18-month-old infants with severe maternal mental illness to stratify them into ‘higher’ and ‘lower’ risk. We shall then test the validity of our risk modelling using functional near infrared spectroscopy (fNIRS). fNIRS can assess the cortical development of speech and language acquisition in at-risk infants from 6 months. We hypothesise that the ‘highest risk’ stratified sample will show significant differences in cortical language processing compared with those in the ‘lower risk’ strata and compared with infants of well mothers. Such early ‘biomarkers’ may allow future routine clinical identification of those at highest risk within this population of vulnerable children and young people, providing valuable evidence for the development of targeted and timely interventions to improve outcome and quality of life. The fourth element of work is a pathway of child-centred supports including digital and face-to-face aspects co-developed by young people.