Table 1 Framework and recommendations for planning, conducting and disseminating evaluations of AMR policy interventions
| Â | Researchers | Policy-makers | Research funders | Publishers |
|---|---|---|---|---|
Mechanisms for prioritising research | Prioritise research to study what works, when it works, why it works, and what elements are necessary to its success | Engage with the research community and make your evidence needs clear | Funding support for formal prioritisation processes Targeted grant competitions to drive research in priority areas | Require that research reports summarise the evidence that was already known on a topic and show systematic review evidence that supports the conduct of an intervention |
Systematic reviews | Inform interventions using rigorous systematic reviews | Partner with researchers to do evidence syntheses to inform policy-making | Strategic funding support for systematic reviews, evidence syntheses, living systematic reviews Require evaluations to be justified by systematic reviews | Support the publication of systematic review protocols, systematic reviews, evidence syntheses and living systematic reviews |
More planned evaluations | Work with other stakeholders to make rigorous evaluations of all AMR programmes the norm | Working with researchers, plan the evaluation strategy for a programme or policy before launching the programme or policy | Strategic funding support for evaluations of policy interventions and for policy-research partnerships to facilitate this research | Require authors to register study protocols |
Research that is more responsive to stakeholder needs | Ensure that research addresses all informational needs of policy-makers (e.g. including equity and cost-effectiveness) | Decide in advance what information and evidence you need to inform policy-making | Encourage integrated knowledge translation and collaboration with stakeholders when awarding grants to support AMR policy research | Ensure timely peer review and publication of research to ensure that evidence is available to support stakeholders |
Better design of AMR Interventions | Use theory, frameworks and logic models in the intervention design phase to frame how and why an intervention is expected to work | Use theory, frameworks and logic models when planning policy interventions to clarify how and why an intervention is expected to work | Do not fund interventions that do not employ theory, frameworks or logic models to describe how and why the intervention is expected to work | Require authors to report on their use of theory, frameworks, and logic in the design and conduct of AMR interventions |
Better design of evaluations | Use the most rigorous possible evaluation designs to minimise bias and maximise generalisability | Embrace research evaluation to understand what, when, why and how and intervention works | Studies using weak study designs (e.g. uncontrolled before and after designs) should not be funded | Refrain from publishing studies that use poor quality methods such as uncontrolled before and after studies for evaluation of AMR interventions |
Iterative improvement on existing trials | Conduct head-to-head comparisons of intervention variations | Promote radical incrementalism (based on rigorous evaluation) to enhance the effectiveness of extant policies | Provide funding support for head-to-head trials | Publish research with neutral and negative results |
A set of standard measures and metrics | Develop a set of core outcome measures for AMR research | Partner with researchers to ensure that core outcome measures address your key evidence needs | Funding support for the development of an AMR core outcome set Require use of core outcome measures in funded applications | Require researchers to use the core outcome measures in published evaluations |
Better reporting of interventions | Commit to full and transparent reporting of studies Use reporting guidelines and checklists to fully report a study Register intervention protocols to reduce the risk of publication bias Avoid ‘spin’ especially with weak evaluative designs | Publish or make available reports on the effectiveness of policy interventions and efforts to improve them | Make public the details of funded interventions Require full and transparent reporting of studies Require researchers to register the protocols of their interventions | Require authors to use the relevant research reporting guidelines and checklists Publish research with neutral and negative results |
More opportunities for shared learning | Disseminate research widely and embrace open data and open access opportunities Make datasets available to other researchers through data repositories Develop cross programme collaborations to encourage learning and efficient knowledge generation | Take advantage of opportunities to borrow and adapt policy interventions from other contexts Make available data on policy interventions in your setting to promote uptake in other contexts and ensure that ineffective policy is not duplicated in other settings | Provide funding for open access publishing, open data-sharing platforms, cross programme collaborations and living systematic reviews | Increased commitment and support for open access publication |