Article Text

Download PDFPDF

When and how do ‘effective’ interventions need to be adapted and/or re-evaluated in new contexts? The need for guidance
  1. Rhiannon E Evans1,
  2. Peter Craig2,
  3. Pat Hoddinott3,
  4. Hannah Littlecott1,
  5. Laurence Moore2,
  6. Simon Murphy1,
  7. Alicia O'Cathain4,
  8. Lisa Pfadenhauer5,
  9. Eva Rehfuess5,
  10. Jeremy Segrott6,
  11. Graham Moore1
  1. 1 DECIPHer, School of Social Sciences, Cardiff University, Cardiff, UK
  2. 2 MRC/CSO Social and Public Health Sciences Unit, University of Glasgow, Glasgow, UK
  3. 3 Nursing, Midwifery and Allied Health Professions Research Unit (NMAHP RU), University of Stirling, Stirling, UK
  4. 4 Medical Care Research Unit, School of Health and Related Research (SHARR), University of Sheffield, Sheffield, UK
  5. 5 Institute for Medical Informatics, Biometry and Epidemiology, Pettenkofer School of Public Health, LMU Munich, Munich, Germany
  6. 6 DECIPHer, Centre for Trials Research (CTU), Cardiff University, Cardiff, UK
  1. Correspondence to Dr Rhiannon E Evans, DECIPHer, School of Social Sciences, Cardiff University, Cardiff cf10 3bd, UK; EvansRE8{at}

Statistics from

Rational models of evidence-informed policy have historically centred on an assumption that it is possible to identify ‘effective’ interventions, before recommending wider implementation. However, for population health interventions (which we define as inclusive of public health and health services), transferability to new contexts is often uncertain.1 Some interventions have demonstrated limited effectiveness, or even harm, when used elsewhere. For example, antenatal corticosteroids have reduced neonatal mortality among pregnant women at risk of preterm birth in high-income countries, but increased mortality and maternal infection in low-income and middle-income countries.2

While definitions are contested, context can be defined as a set of active and unique characteristics and circumstances that interact with, modify, facilitate or constrain intervention delivery and effects. It includes geographical, epidemiological, sociocultural, socioeconomic, ethical, legal and political determinants.3 The argument that every context is unique and interventions cannot translate across them should not be overstated however. Some parenting interventions, for instance, have not successfully transferred, while others have been highly resilient to contextual variation, with minimal difference in effects between ‘home-grown’ and transported approaches.4

There is increased recognition then of the need to consider context when making decisions about the transferability of evidence. Population health interventions are increasingly conceived as ‘events in systems’, which aim to modify aspect(s) of a pre-existing context, altering conditions that sustain suboptimal population health outcomes. Effects therefore may be shaped as much by changes to or displacement of prior features of the context, as by properties of the new intervention. Implementation may differ in a new context due to the complexity of the intervention components or ambiguity over its mechanisms, inhibiting high fidelity. Implementation failure may ensue if an intervention …

View Full Text

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.