BackgroundThe children of parents with severe personality difficulties have greater risk of significant mental health problems. Existing care is poorly co-ordinated, with limited effectiveness. A specialised parenting intervention may improve child and parenting outcomes, reduce family morbidity and lower the service costs.ObjectivesTo develop a specialised parenting intervention for parents affected by severe personality difficulties who have children with mental health problems and to conduct a feasibility trial.DesignA pragmatic, mixed-methods design to develop and pilot a specialised parenting intervention, Helping Families Programme-Modified, and to conduct a randomised feasibility trial with process evaluation. Initial cost-effectiveness was assessed using UK NHS/Personal Social Services and societal perspectives, generating quality-adjusted life-years. Researchers collecting quantitative data were masked to participant allocation.SettingTwo NHS mental health trusts and concomitant children’s social care services.ParticipantsParents who met the following criteria: (1) the primary caregiver of the index child, (2) aged 18–65 years, (3) have severe personality difficulties, (4) proficient in English and (5) capable of providing informed consent. Index children who met the following criteria: (1) aged 3–11 years, (2) living with index parent and (3) have significant emotional/behavioural difficulties. Exclusion criteria were (1) having coexisting psychosis, (2) participating in another parenting intervention, (3) receiving inpatient care, (4) having insufficient language/cognitive abilities, (5) having child developmental disorder, (6) care proceedings and (7) index child not residing with index parent.InterventionThe Helping Families Programme-Modified – a 16-session intervention using structured, goal-orientated strategies and collaborative therapeutic methods to improve parenting, and child and parent functioning. Usual care – standard care augmented by a single psychoeducational session.Main outcome measuresTrial feasibility – rates of recruitment, eligibility, allocation, retention, data completion and experience. Intervention acceptability – rates of acceptance, completion, alliance (Working Alliance Inventory-Short Revised) and experience. Outcomes – child (assessed via Concerns About My Child, Eyberg Child Behaviour Inventory, Child Behaviour Checklist-Internalising Scale), parenting (assessed via the Arnold-O’Leary Parenting Scale, Kansas Parental Satisfaction Scale), parent (assessed via the Symptom Checklist-27), and health economics (assessed via the Client Service Receipt Inventory, EuroQol-5 Dimensions).ResultsThe findings broadly supported trial feasibility using non-diagnostic screening criteria. Parents were mainly referred from one site (75.0%). Site and participant factors delayed recruitment. An estimate of eligible parents was not obtained. Of the 86 parents referred, 60 (69.7%) completed screening and 48 of these (80.0%) were recruited. Participants experienced significant disadvantage and multiple morbidity. The Helping Families Programme-Modified uptake (87.5%) was higher than usual-care uptake (62.5%). Trial retention (66.7%, 95% confidence interval 51.6% to 79.6%) exceeded the a priori rate. Process findings highlighted the impact of random allocation and the negative effects on retention. The Helping Families Programme-Modified was acceptable, with duration of delivery longer than planned, whereas the usual-care condition was less acceptable. At initial follow-up, effects on child and parenting outcomes were detected across both arms, with a potential outcome advantage for the Helping Families Programme-Modified (effect size range 0.0–1.3). For parental quality-adjusted life-years, the Helping Families Programme-Modified dominated usual care, and child quality-adjusted life-years resulted in higher costs and more quality-adjusted life-years. At second follow-up, the Helping Families Programme-Modified was associated with higher costs and more quality-adjusted life-years than usual care. For child quality-adjusted life-years, when controlled for baseline EuroQol-5 Dimensions, three-level version, usual care dominated the Helping Families Programme-Modified. No serious adverse events were reported.ConclusionThe Helping Families Programme-Modified is an acceptable specialised parenting intervention. Trial methods using non-diagnostic criteria were largely supported. For future work, a definitive efficacy trial should consider site selection, recruitment methods, intervention efficiency and revised comparator condition.Trial registrationCurrent Controlled Trials ISRCTN14573230.FundingThis project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full inHealth Technology Assessment; Vol. 24, No. 14. See the NIHR Journals Library website for further project information.