Challenges to undertaking randomised trials with looked after children in social care settings

Abstract

BackgroundRandomised controlled trials (RCTs) are widely viewed as the gold standard for assessing effectiveness in health research; however many researchers and practitioners believe that RCTs are inappropriate and un-doable in social care settings, particularly in relation to looked after children. The aim of this article is to describe the challenges faced in conducting a pilot study and phase II RCT of a peer mentoring intervention to reduce teenage pregnancy in looked after children in a social care setting.MethodsInterviews were undertaken with social care professionals and looked after children, and a survey conducted with looked after children, to establish the feasibility and acceptability of the intervention and research design.ResultsBarriers to recruitment and in managing the intervention were identified, including social workers acting as informal gatekeepers; social workers concerns and misconceptions about the recruitment criteria and the need for and purpose of randomisation; resource limitations, which made it difficult to prioritise research over other demands on their time and difficulties in engaging and retaining looked after children in the study.ConclusionsThe relative absence of a research infrastructure and culture in social care and the lack of research support funding available for social care agencies, compared to health organisations, has implications for increasing evidence-based practice in social care settings, particularly in this very vulnerable group of young people.Electronic supplementary materialThe online version of this article (doi:10.1186/s13063-015-0708-z) contains supplementary material, which is available to authorized users.

The Healthy Eating, Active Living (HEAL) Study: Outcomes, Lessons Learnt and Future Recommendations

“…there is far less research in residential or kinship care); (3) quantitative research has relied heavily on descriptive data rather than statistical examination of the relationships between variables; and (4) a lack of randomised trials (prior to HEAL, only three other studies explicitly stated the use of a randomised trial design) (Bromfield and Osborn, ; McDonald et al ., ; Shlonsky et al ., ). Other researchers have also questioned the feasibility and acceptability of using randomised trials in this setting and outlined barriers to recruitment and intervention implementation, including a lack of research infrastructure and culture in social care, concerns and/or misunderstandings about the need/purpose of randomisation, and limited resources, which make it difficult to prioritise research alongside other competing demands (Dixon et al ., ; Mezey et al ., ). Clearly, the methodological issues associated with conducting research in this setting are not exclusive to the current study.…”

Systematic review of educational interventions for looked‐after children and young people: Recommendations for intervention development and evaluation

“…The utilisation of RCTs within educational and social care settings have been plagued by a wealth of problems that have arguably contributed to their poor conduct, with this issue being further compounded by the declining quality of reporting (Torgerson et al ., ; Flynn et al ., ; Dixon et al ., ; Mezey et al ., ). As Dixon et al .…”

“…(, p. 1564) surmise, controversy over trials are derived from the interpretation that they are ‘unethical, positivist, uncritically imported from other disciplines, and unable to yield the certainty they promise’. Indeed, debates have abounded about the perceived unequitable or unethical assignment of individuals to control groups (Gueron, ; Dixon et al ., ; Mezey et al ., ), as though the expenditure of public resources on interventions without evidence of effectiveness does not pose its own ethical challenges. Further, more pragmatic limitations have included insufficient recruitment to ensure statistical power, practitioners circumventing randomisation processes by the introduction of their own rationing strategies, and the struggle to secure funding for the conduct of the trial itself (Gueron, ; Dixon et al ., ; Mezey et al ., ).…”

“…Indeed, debates have abounded about the perceived unequitable or unethical assignment of individuals to control groups (Gueron, ; Dixon et al ., ; Mezey et al ., ), as though the expenditure of public resources on interventions without evidence of effectiveness does not pose its own ethical challenges. Further, more pragmatic limitations have included insufficient recruitment to ensure statistical power, practitioners circumventing randomisation processes by the introduction of their own rationing strategies, and the struggle to secure funding for the conduct of the trial itself (Gueron, ; Dixon et al ., ; Mezey et al ., ). However, regardless of these challenges, RCTs within educational and social care settings are increasingly gaining traction, and the research community has openly called for a stronger culture around this evaluation design (Gueron, ).…”

“…Third, and perhaps most fundamentally, existing evidence of effect is compromised by a lack of scientific rigour, with limited utilisation of RCTs, which arguably constitute the most appropriate methodology to answer questions of effectiveness (Mezey et al ., ). Studies have been further undermined by small sample sizes, with many failing to provide sample size calculations, making it unclear whether evaluations are sufficiently powered to detect anticipated effect sizes.…”