Is it feasible to nest a randomised Trial within a Cohort Study (TwiCS) to evaluate an early years parenting programme? Results from a Born in Bradford’s Better Start (BiBBS) study

Background It is crucial to find pragmatic design alternatives to randomised trials to evaluate early childhood parenting interventions in practice, which can suffer from potentially low uptake and disappointment bias. In the first study of its kind, we assess the feasibility of nesting a randomised Trial within a Cohort Study (TwiCS) to evaluate an early years parenting intervention delivered to parents of 1-2-year-old children living in inner-city Bradford, England. The study evaluated: (1) whether TwiCS methodology can be implemented to create control and intervention arms, whilst documenting any incidences of contamination; (2) whether satisfactory rates of intervention uptake are achieved; and (3) whether satisfactory rates of retention of participants are achieved. Methods A stratified random sample of 240 mother-child dyads were selected from an eligible population in the Born in Bradford’s Better Start cohort. Mother-child dyads were individually randomised 1:1 to the intervention or control, with stratification by child age (1 or 2 years old at the time of randomisation) and ethnicity (White British, South Asian, or other). A Red Amber Green (RAG) rating system supported the assessment of each objective: red (not achieved), amber (partly achieved), and green (achieved). Results TwiCS methodology created balanced control and intervention arms, and 73% of parents could be contacted about the intervention (green), though rates of contamination were higher than desired (2%, amber). Satisfactory rates of intervention uptake could not be achieved: 43% of parents consented to a referral (amber), and 14% of parents enrolled in IY-T (red). Satisfactory rates of retention in the intervention may be achievable, as 80% of enrolled parents participated (amber), and all participating parents completed the intervention (green). Conclusions It was feasible to use TwiCS methodology to randomly select, allocate, and contact parents, and document contamination in this setting. However, intervention uptake was low, and this would need improving for a future TwiCS evaluating this intervention to be feasible. This study suggests a TwiCS is not a viable approach to evaluating early years interventions delivered >12 months after cohort recruitment. Other pragmatic and implementable methods of evaluation may be more feasible, such as quasi-experimental designs, or target trials. Trial registration The study was prospectively registered on ISRCTN (ISRCTN16150114).