BackgroundAfrican-Caribbean people in the UK experience the highest incidence of schizophrenia and the greatest inequity in mental health care. There is an urgent need to improve their access to evidence-based care and outcomes. Family intervention (FI) is a National Institute for Health and Care Excellence-approved psychosocial intervention. Although clinically effective and cost-effective for schizophrenia, it is rarely offered. Evidence for any research into FI is lacking for ethnic minority people generally and for African-Caribbean people specifically.Aims(1) To assess the feasibility of delivering a novel, culturally appropriate psychosocial intervention within a ‘high-risk’ population to improve engagement and access to evidence-based care. (2) To test the feasibility and acceptability of delivering FI via ‘proxy families’.DesignA mixed-methods, feasibility cohort study, incorporating focus groups and an expert consensus conference.SettingTwo mental health trusts in north-west England.ParticipantsWe recruited a convenience sample of 31 African-Caribbean service users. Twenty-six family units [service users, relatives/family support members (FSMs) or both] commenced therapy. Half of the service users (n = 13, 50%), who did not have access to their biological families, participated by working with FSMs.InterventionsAn extant FI model was culturally adapted with key stakeholders using a literature-derived framework [Culturally adapted Family Intervention (CaFI)]. Ten CaFI sessions were offered to each service user and associated family.Main outcome measuresRecruitment (number approached vs. number consented), attendance (number of sessions attended), attrition (number of dropouts at each time point), retention (proportion of participants who completed therapy sessions), and completeness of outcome measurement.ResultsOf 74 eligible service users, 31 (42%) consented to take part in the feasibility trial. The majority (n = 21, 67.7%) were recruited from community settings, seven (22.6%) were recruited from rehabilitation settings and three (9.7%) were recruited from acute wards. Twenty-four family units (92%) completed all 10 therapy sessions. The proportion who completed treatment was 77.42% (24/31). The mean number of sessions attended was 7.90 (standard deviation 3.96 sessions) out of 10. It proved feasible to collect a range of outcome data at baseline, post intervention and at the 3-month follow-up. The rating of sessions and the qualitative findings indicated that CaFI was acceptable to service users, families, FSMs and health-care professionals.LimitationsThe lack of a control group and the limited sample size mean that there is insufficient power to assess efficacy. The findings are not generalisable beyond this population.ConclusionsIt proved feasible to culturally adapt and test FI with a sample of African-Caribbean service users and their families. Our study yielded high rates of recruitment, attendance, retention and data completion. We delivered CaFI via FSMs in the absence of biological families. This novel aspect of the study has implications for other groups who do not have access to their biological families. We also demonstrated the feasibility of collecting a range of outcomes to inform future trials and confirmed CaFI’s acceptability to key stakeholders. These are important findings. If CaFI can be delivered to the group of service users with the most serious and persistent disparities in schizophrenia care, it has the potential to be modified for and delivered to other underserved groups.Future workA fully powered, multicentre trial, comparing CaFI with usual care, is planned.Trial registrationCurrent Controlled Trials ISRCTN94393315.FundingThis project was funded by the National Institute for Health Research (NIHR) Health Services and Delivery Research programme and will be published in full inHealth Services and Delivery Research; Vol. 6, No. 32. See the NIHR Journals Library website for further project information.