Table 5

Support for proposition 1 - the context within which complex interventions occur is crucial to understanding how that intervention works and how it might be generalisable
Findings in support of proposition Evidence source
Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Trial 7
1. Complex interventions (to a greater extent than simple interventions) may require or result in changes in care delivery and therefore demand the involvement and commitment of practitioners and participants within practice settings to make them happen. ✓✓ ✓✓ ✓✓ ✓✓
2. Complex interventions are shaped or co-constructed by aspects of context. Indeed, it seems that the context of a complex intervention may in fact be considered to be a part of that intervention. ✓✓ ✓✓ ✓✓ ✓✓
3. The problem context encompasses:
· The nature and stability of the scale, distribution and causal mechanism of the problem that the intervention is designed to address. These can change over time and in relation to changes in personal and organisational contexts. ✓✓ ✓✓ ✓✓ ✓✓ ✓✓
4. The personal context encompasses:
· Factors related to the practitioners involved - perceptions of relevance and interest in the intervention, skills, motivation, beliefs, preferences, affinity for intervention, ability to fit it in. ✓✓ ✓✓ ✓✓ ✓✓ ✓✓
· Relationships between the practitioner and the participant may become sufficiently important that it becomes a mechanism of action that may facilitate or hinder the effectiveness of the intervention. ✓✓ ✓✓ ✓✓ ✓✓
5. The organisational context encompasses:
· Organisation of services, managerial support, practicality of delivering interventions, staff availability, venue and timing. ✓✓ ✓✓ ✓✓ ✓✓
6. The trial context encompasses:
· Personal and interpersonal factors related to the researcher(s) - beliefs and preferences, commitment, role in trial, relationships with practitioners and participants, background and allegiances. ✓✓ ✓✓ ✓✓ ✓✓ ✓✓ ✓✓ ✓✓
· Real and perceived knock-on effects of the intervention on the practice setting. ✓✓ ✓✓ ✓✓

✓ = some evidence (one instance); ✓✓ = strong evidence (more than one instance); no check/tick = no specific evidence.

Wells et al.

Wells et al. Trials 2012 13:95   doi:10.1186/1745-6215-13-95

Open Data