Table 2 Summary of evolution of simplified framework version 1 to the AIMD framework (simplified framework version 2)

Intended Targets to AIMS

Conceptually overlaps with causal mechanisms via reference to intended effects

Renamed as ‘Aims’ in version 2 and redefined to reflect that the intended effects are the aims of the intervention and the beneficiaries are who the aims are directed towards

Evaluation could be considered part of intended target in that it quantifies the expected change; therefore intended target could be redefined to consider aim of study

See above

Active Ingredients to INGREDIENTS

Conceptually overlaps with causal mechanism (e.g., an active ingredient such as ‘persuasion’ implies a causal mechanism)

The word active was removed from version 1 to avoid confusion between the intervention ingredients and the mechanism by which the intervention works. This resulted in ‘Ingredients’ in version 2

Best defined as ‘what it is’ (i.e., remove bulleted points, 3 and 4) or ‘as empirically established’

The single term ‘Ingredients’ acts as a prompt to provide details (i.e., the component parts) rather than a broad nominal description

Causal Mechanisms to MECHANISM

Conceptually overlaps with active ingredients as exemplified above

As above

The term could refer to ‘how it is known to work’ (empirically established) or ‘how it is thought to work’ (theoretical rationale) – the definition could be refined or the concept of ‘rationale’ could be incorporated differently?

‘Causal’ was felt to indicate an empirically established mechanism, rather than a hypothesized mechanism. Therefore, ‘causal’ was removed from version 1 to allow for consideration of theoretical or empirical rationale. This resulted in ‘Mechanism’ in version 2

Mode of Delivery or Application to DELIVERY

Conceptually overlaps with active ingredient (e.g., local opinion leader implies active ingredient, i.e., a local opinion leader can deliver an intervention, but a local opinion leader is also an active ingredient, and could be delivered in multiple modes of delivery such as phone, face-to-face)

There can be more to delivery than just mode. Therefore, this component was renamed as ‘Delivery’ in version 2, and redefined to encompass information such as mode, content and dosage

Mode of delivery alone is insufficient for replicability. Furthermore, combining mode of delivery with application is problematic as one may or may not be covered; this category may need to be redefined to include other vital information (e.g., eligibility criteria, mode, delivery personnel, content, dosage (i.e., duration, intensity), audience and its size of audience, number of care providers and centres, intervention fidelity and its measurement, the identification of breaches and how the intervention was modified, context, standardisation and tailoring strategies, clustering, blinding, enrolment, and allocation

Per above, ‘Delivery’ now defined more broadly to improve applicability to non-clinical settings (e.g., public health, policy)

Across all framework components

Consider how post-intervention information with potential application to future interventions could be used (e.g., fidelity, what was learned about causal mechanisms, contexts in which intervention may or may not be effective, financial considerations such as cost-benefit, etc.). This information could be considered as part of the description of a published study intervention or it could be assumed that this information is fed into future intervention studies (i.e., methods description)

The framework exists within the limits of the intervention itself; it does not extend to other contextual concepts (e.g. fidelity, rationale). Therefore, no refinements to version 1 were made pertaining to this issue

Further guidance is needed on how to handle situations where there is more than one intervention (i.e., is the tool used per intervention or can a group of interventions be scored together?)

Multiple intervention components could be considered within AIMD. No refinements to version 1 were made pertaining to this issue

‘Rationale’ for the intervention could be more explicitly covered, perhaps as a stand-alone component by the components and/or as a separate component

Not discussed at length in meeting. Changes were made to version 1 to more explicitly define ‘rationale’ within ‘Mechanism’ (see below)

The control condition should be described in the same terms as the intervention and/or the control condition justified

Not discussed at length in meeting. No refinements to version 1 were made pertaining to this issue