br Disclaimer br Introduction Adolescent substance misuse is

Disclaimer
Introduction Adolescent substance misuse is a significant problem in developed countries (Currie et al., 2012; U.S. Department of Health and Human Services, 2007) and early initiation of substance use is associated with higher levels of substance-related harm during adulthood (Dawson, Goldstein, Chou, Ruan & Grant, 2008; Grant & Dawson, 1998). Because the consequences of early initiation are difficult to modify, an important response has been the development of family-based prevention interventions (Cuijpers, 2003; Kumpfer, Alvarado & Whiteside, 2003). One such intervention, the Strengthening Families Programme (SFP), aims to delay substance use initiation and prevent later misuse through strengthening family-based protective factors. In the United States of America trials of SFP 10-14 - a universal version of SFP for families with children aged 10–14, have found evidence of long-term effectiveness (Spoth, Redmond & Shin, 2001; Spoth et al., 2013; Spoth, Redmond, Trudeau & Shin, 2002), though the methodological rigour of these studies has been criticised (Gorman, 2015). The evidence aa-utp for family-based prevention interventions such as SFP 10-14 is dominated by studies from the USA and there is a need for more research on whether effective interventions can be successfully ‘transported’ to other national contexts (Petrie, Bunn & Byrne, 2007), where they are more likely to be implemented under ‘real-world’ conditions, and without extensive input from programme developers (Axford & Morpeth, 2013). Family-based programmes are complex interventions, with multiple components designed to work synergistically. Process evaluations, which analyse implementation, aid interpretation of complex outcome effects and understanding of intervention theory (Durlak, 1998; Durlak & DuPre, 2008). An important purpose of process evaluations is to assess the extent to which interventions are implemented with fidelity (Carroll et al., 2007; Moore, et al., 2014). This includes adherence (whether planned activities are delivered), dose (how much of an intervention is delivered/received), delivery quality, and reach and recruitment (Baranowski & Stables, 2000; Dusenbury, Brannigan, Falco & Hansen, 2003). Alongside these quantitative measures, qualitative research can provide important data on the processes which influence implementation, and their variation across contexts (Moore et al., 2014). New interventions must operate within existing delivery systems and biochemistry depend upon cooperation from individuals and organisations, especially when delivered on a multi-agency basis (May, 2013) - a common social service delivery mechanism in the UK and elsewhere. Delivery settings are typically complex systems - characterised by the interaction of multiple individuals, social networks and organisations. Within these systems practitioners make meaning of interventions in ways which shape how they are delivered (Bisset, Daniel & Potvin, 2009; May, 2013) – though the study of these phenomena is limited (Bisset, Potvin & Daniel, 2013; Hill, Maucione & Hood, 2007). Practitioner engagement with an intervention may be emergent (and therefore hard to predict), and self-adaptive rather than centrally controlled (Sterman, 2006; Tan, Wen & Awad, 2005). Although studies in many countries have encountered variation in implementation across delivery contexts (Cantu, Hill & Becker, 2010; Durlak & DuPre, 2008; Lendrum & Humphrey, 2012), the role of intervention-context interaction in shaping this has often been overlooked (Bisset et al., 2009; Bonell, Fletcher, Morton, Lorenc & Moore, 2012; Glasgow, Lichtenstein & Marcus, 2006; Hawe, Shiell, Riley & Gold, 2004), and the narrow focus of process evaluations on quantitative assessment of pre-specified structural aspects of interventions (e.g. coverage of intervention activities) has been criticised for paying insufficient attention to the processes through which they occur (). A previous trial of SFP 10-14 (conducted in the United States) – in which the programme was delivered by community-university partnerships, found no significant association between implementation team functioning and levels of adherence, but suggested that potential relationships may have been masked by the consistently high rates of adherence across programmes (Spoth, Guyll, Lillehoj, Redmond & Greenberg, 2007). However, evaluation of the programme in the USA as part of ‘real world’ dissemination found greater variation in adherence and other aspects of implementation (staffing levels, group size, children\'s age range), though no clear association between facilitator characteristics and fidelity (Cantu et al., 2010; Hill et al., 2007). Questions therefore remain about the key influences on the quality of implementation of SFP 10-14, the role of individual facilitators and their teams, and the influence of wider contextual factors.