Complex systems approaches to social intervention research are increasingly advocated. However, there have been few attempts to consider how models of intervention science, such as the UK’s Medical Research Council complex interventions framework, might be reframed through a complex systems lens. This article identifies some key areas in which this framework might be reconceptualized, and a number of priority areas where further development is needed if alignment with a systems perspective is to be achieved. We argue that a complex systems perspective broadens the parameters of ‘relevant’ evidence and theory for intervention development, before discussing challenges in defining feasibility in dynamic terms. We argue that whole systems evaluations may be neither attainable, nor necessary; acknowledgment of complexity does not mean that evaluations must be complex, or investigate all facets of complexity. However, a systems lens may add value to evaluation design through guiding identification of key uncertainties, and informing decisions such as timings of follow-up assessments.

BackgroundAdapting interventions that have worked elsewhere can save resources associated with developing new interventions for each specific context. While a developing body of evidence shows benefits of adapted interventions compared with interventions transported without adaptation, there are also examples of interventions which have been extensively adapted, yet have not worked in the new context. Decisions on when, to what extent, and how to adapt interventions therefore are not straightforward, particularly when conceptualising intervention effects as contingent upon contextual interactions in complex systems. No guidance currently addresses these questions comprehensively. To inform development of an overarching guidance on adaptation of complex population health interventions, this systematic review synthesises the content of the existing guidance papers.MethodsWe searched for papers published between January 2000 and October 2018 in 7 bibliographic databases. We used citation tracking and contacted authors and experts to locate further papers. We double screened all the identified records. We extracted data into the following categories: descriptive information, key concepts and definitions, rationale for adaptation, aspects of adaptation, process of adaptation, evaluating and reporting adapted interventions. Data extraction was conducted independently by two reviewers, and retrieved data were synthesised thematically within pre-specified and emergent categories.ResultsWe retrieved 6694 unique records. Thirty-eight papers were included in the review representing 35 sources of guidance. Most papers were developed in the USA in the context of implementing evidence-informed interventions among different population groups within the country, such as minority populations. We found much agreement on how the papers defined key concepts, aims, and procedures of adaptation, including involvement of key stakeholders, but also identified gaps in scope, conceptualisation, and operationalisation in several categories.ConclusionsOur review found limitations that should be addressed in future guidance on adaptation. Specifically, future guidance needs to be reflective of adaptations in the context of transferring interventions across countries, including macro- (e.g. national-) level interventions, better theorise the role of intervention mechanisms and contextual interactions in the replicability of effects and accordingly conceptualise key concepts, such as fidelity to intervention functions, and finally, suggest evidence-informed strategies for adaptation re-evaluation and reporting.Trial registrationPROSPERO 2018, CRD42018112714.

The integration of realist evaluation principles within randomised controlled trials (‘realist RCTs’) enables evaluations of complex interventions to answer questions about what works, for whom and under what circumstances. This allows evaluators to better develop and refine mid-level programme theories. However, this is only one phase in the process of developing and evaluating complex interventions. We describe and exemplify how social scientists can integrate realist principles across all phases of the Medical Research Council framework. Intervention development, modelling, and feasibility and pilot studies need to theorise the contextual conditions necessary for intervention mechanisms to be activated. Where interventions are scaled up and translated into routine practice, realist principles also have much to offer in facilitating knowledge about longer-term sustainability, benefits and harms. Integrating a realist approach across all phases of complex intervention science is vital for considering the feasibility and likely effects of interventions for different localities and population subgroups.

Recent years have seen a growing emphasis on the value of building and testing middle range theory throughout the development and evaluation of complex population health interventions. We agree that a coherent theoretical basis for intervention development, and use of evaluation to test key causal assumptions and build theory, are crucial. However, in this editorial, we argue that such recommendations have often been operationalised in somewhat simplistic terms with potentially perverse consequences, and that an uncritical assumption that an intervention explicitly based on theory is inherently superior carries significant risks. We first argue that the drive for theory-based approaches may have exacerbated a propensity to select ‘off-the-shelf’ theories, leading to the selection of inappropriate theories which distract attention from the mechanisms through which a problem is actually sustained. Second, we discuss a tendency toward over-reliance on individual-level theorising. Finally, we discuss the relatively slow progress of population health intervention research in attending to issues of context, and the ecological fit of interventions with the systems whose functioning they attempt to change. We argue that while researchers should consider a broad range of potential theoretical perspectives on a given population health problem, citing a popular off-the-shelf theory as having informed an intervention and its evaluation does not inherently make for better science. Before identifying or developing a theory of change, researchers should develop a clear understanding of how the problem under consideration is created and sustained in context. A broader conceptualisation of theory that reaches across disciplines is vital if theory is to enhance, rather than constrain, the contribution of intervention research. Finally, intervention researchers need to move away from viewing interventions as discrete packages of components which can be described in isolation from their contexts, and better understand the systems into which change is being introduced.

To date, research on the role of the Internet in self-harm has focused on young people's interaction via the medium of text, with limited consideration of the effect of images. This qualitative study explores how young people understand and use online images of self-harm. Semi-structured interviews were conducted with a community sample of 21 individuals aged 16–24 living in Wales, UK, with a previous history of self-harm. Interviewees reported the role of the Internet in normalising young people's self-harm. Images rather than textual interactions are the primary reason cited for using the Internet for self-harm purposes. Images invoke a physical reaction and inspire behavioural enactment, with Tumblr, which permits the sharing of images by anonymous individuals, being the preferred platform. Viewing online images serves a vital role in many young people's self-harm, as part of ritualistic practice. Online prevention and intervention need to attend to the importance of images.

Rates of hospital presentation for self-harm have increased in recent years, and although clinical practice guidelines on clinical provision prioritize positive patient experiences, the quality of provision remains variable. This systematic review provides an updated and extended synthesis of qualitative research on the following: (a) patients’ experiences of treatment following presentation to hospital; and (b) patients’ perceptions of the impact of treatment on recurrent self-harm and/or suicidal ideation, and future help-seeking. Twenty-six studies were identified for inclusion in the final synthesis. Three meta-themes emerged: (a) individuals undertake extensive identity work when presenting with self-harm, navigating the process of becoming a patient, and negotiating the type of patient they want to be; (b) care ranges from gentle to hostile, with care at admission and discharge being particularly disorientating; and (c) negative experiences of clinical treatment may increase future self-harm. Emerging research gaps include the need for further theoretically informed qualitative research in this area.

Background: Evidence reports that schools influence children and young people's health behaviours across a range of outcomes. However there remains limited understanding of the mechanisms through which institutional features may structure self-harm and suicide. This paper reports on a systematic review and meta-ethnography of qualitative research exploring how schools influence self-harm and suicide in students. Methods: Systematic searches were conducted of nineteen databases from inception to June 2015. English language, primary research studies, utilising any qualitative research design to report on the influence of primary or secondary educational settings (or international equivalents) on children and young people's self-harm and suicide were included. Two reviewers independently appraised studies against the inclusion criteria, assessed quality, and abstracted data. Data synthesis was conducted in adherence with Noblit and Hare's meta-ethnographic approach. Of 6744 unique articles identified, six articles reporting on five studies were included in the meta-ethnography. Results: Five meta-themes emerged from the studies. First, self-harm is often rendered invisible within educational settings, meaning it is not prioritised within the curriculum despite students' expressed need. Second, where self-harm transgresses institutional rules it may be treated as 'bad behaviour', meaning adequate support is denied. Third, schools' informal management strategy of escalating incidents of self-harm to external 'experts' serves to contribute to non-help seeking behaviour amongst students who desire confidential support from teachers. Fourth, anxiety and stress associated with school performance may escalate self-harm and suicide. Fifth, bullying within the school context can contribute to self-harm, whilst some young people may engage in these practices as initiation into a social group.

No abstract