Background: Quality of life for children and adolescents living with serious parental mental illness can be impaired, but evidence-based interventions to improve it are scarce.

Objective: Co-production of a child-centred intervention [called Young Simplifying Mental Illness plus Life Enhancement Skills (SMILES)] to improve the health-related quality of life of children and adolescents living with serious parental mental illness, and evaluating its acceptability and feasibility for delivery in NHS and community settings.

Design: Qualitative and co-production methods informed the development of the intervention (Phase I). A feasibility randomised controlled trial was designed to compare Young SMILES with treatment as usual (Phase II). Semistructured qualitative interviews were used to explore acceptability among children and adolescents living with their parents, who had serious mental illness, and their parents. A mixture of semistructured qualitative interviews and focus group research was used to examine feasibility among Young SMILES facilitators and referrers/non-referrers.

Setting: Randomisation was conducted after baseline measures were collected by the study co-ordinator, ensuring that the blinding of the statistician and research team was maintained to reduce detection bias.

Participants: Phase I: 14 children and adolescents living with serious parental mental illness, seven parents and 31 practitioners from social, educational and health-related sectors. Phase II: 40 children and adolescents living with serious parental mental illness, 33 parents, five referrers/non-referrers and 16 Young SMILES facilitators.

Intervention: Young SMILES was delivered at two sites: (1) Warrington, supported by the National Society for the Prevention of Cruelty to Children (NSPCC), and (2) Newcastle, supported by the NHS and Barnardo's. An eight-session weekly group programme was delivered, with four to six children and adolescents living with serious parental mental illness per age-appropriate group (6-11 and 12-16 years). At week 4, a five-session parallel weekly programme was offered to the parents/carers. Sessions lasted 2 hours each and focused on improving mental health literacy, child-parent communication and children's problem-solving skills.

Main outcome measures: Phase ll children and parents completed questionnaires at randomisation and then again at 4 and 6 months post randomisation. Quality of life was self-reported by children and proxy-reported by parents using the Paediatric Quality of Life questionnaire and KIDSCREEN. Semistructured interviews with parents (n = 14) and children (n = 17) who participated in the Young SMILES groups gathered information about their motivation to sign up to the study, their experiences of participating in the group sessions, and their perceived changes in themselves and their family members following intervention. Further interviews with individual referrers (n = 5) gathered information about challenges to recruitment and randomisation. Two focus groups (n = 16) with practitioners who facilitated the intervention explored their views of the format and content of the Young SMILES manual and their suggestions for changes.

Results: A total of 35 families were recruited: 20 were randomly allocated to Young SMILES group and 15 to treatment as usual. Of those, 28 families [15/20 (75%) in the intervention group and 13/15 (87%) in the control group] gave follow-up data at the primary end point (4 months post baseline). Participating children had high adherence to the intervention and high completion rates of the questionnaires. Children and adolescents living with their parents, who had serious mental illness, and their parents were mainly very positive and enthusiastic about Young SMILES, both of whom invoked the benefits of peer support and insight into parental difficulties. Although facilitators regarded Young SMILES as a meaningful and distinctive intervention having great potential, referrers identified several barriers to referring families to the study. One harm was reported by a parent, which was dealt with by the research team and the NSPCC in accordance with the standard operating procedures.

Limitations: The findings from our feasibility study are not sufficient to recommend a fully powered trial of Young SMILES in the near future. Although it was feasible to randomise children and adolescents living with serious parental mental illness of different ages to standardised, time-limited groups in both NHS and non-NHS settings, an intervention like Young SMILES is unlikely to address underlying core components of the vulnerability that children and adolescents living with serious parental mental illness express as a population over time.

Conclusions: Young SMILES was widely valued as unique in filling a recognised gap in need. Outcome measures in future studies of interventions for children and adolescents living with serious parental mental illness are more likely to capture change in individual risk factors for reduced quality of life by considering their unmet need, rather than on an aggregate construct of health-related quality of life overall, which may not reflect these young people's needs.

Future work: A public health approach to intervention might be best. Most children and adolescents living with serious parental mental illness remain well most of the time, so, although their absolute risks are low across outcomes (and most will remain resilient most of the time), consistent population estimates find their relative risk to be high compared with unexposed children. A public health approach to intervention needs to be both tailored to the particular needs of children and adolescents living with serious parental mental illness and agile to these needs so that it can respond to fluctuations over time.

Trial registration: Current Controlled Trials ISRCTN36865046.

Funding: This project was funded by the National Institute of Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 24, No. 59. See the NIHR Journals Library website for further project information.