Cultural adaptation and intervention integrity: a response to Skärstrand, Sundell and Andréasson

A key issue in the cultural adaptation of interventions concerns the tension between modifying programmes to meet the needs of new populations and implementation settings, and the importance of retaining fidelity to the original intervention design. 1 Skärstrand et al.'s recent randomized controlled trial (RCT) of the Swedish adaptation of Strengthening Families Programme (SFP) 10-14—a family-based substance misuse prevention intervention—found that it was ineffective in preventing drunkenness or drug or tobacco use. 2 We are concerned that the null result could be interpreted as evidence that the SFP 10-14 is ineffective, whereas adaptations made to the intervention may have undermined intervention fidelity and theory by removing key family-based activities that play an important role in its hypothesized pathways to behaviour change. The original American SFP 10-14 3 (and its use in countries such as the UK and Poland 4,5) comprises delivery to groups of 10–12 families, which parents and their child(ren) attend together. Parents and young people undertake separate activities in the first hour of each weekly session, and families are engaged in joint activities in the second hour to practise together the skills they have learnt. In many cases a meal is also offered to families at the halfway point of each session, providing an opportunity for social interaction within and between families. The focus on family level change, and the involvement of parents and children working together to develop new skills, thus differentiates SFP 10-14 from both traditional school-based health education (delivered only to pupils) and parenting (delivered only to parents) programmes. Changes made to the SFP 10-14 during its adaptation for Sweden 6 included: delivery of young people and parent sessions at different times (with the former during the school day and the latter in the evening); removal of the family hour in 6 of the 7 weeks of the programme sessions; providing the intervention to children whose parents/carers did not participate; delivery of the young people's hour to whole classes of 25–30 young people, rather than groups of 10–12 participants; and inclusion of the booster programme sessions as a continuation of the 7-week programme (with no family session in 3 of the 4 weeks). The removal of the family sessions, and the decision to deliver the parent and young people's sessions at different times, appear to have been driven primarily by local implementation needs and available resources (the cost of facilitators and the need to use …