Table 2 Elements of PAR, RCT's, and integrated PAR/RCT

Collaborative design

Externally created, standardized interventions

Key elements of intervention are locally implemented based on collaborative discussion

Use of site PIs in each unique study site as collaborators with study PIs in intervention design

Internal control

External control

Joint control

Site PIs with local or shared authority

Local applicability

Generalizeable

Use local findings to inform universal understanding

Consider local insights gleaned from the implementation process as data that will form the basis for a general understanding

Acknowledge unique local environments

Uniqueness minimized through random assignment

Incorporation of local conditions into overarching approaches

Address local barriers in intervention implementation

Reveal biases

Reduce bias

Use bias to form basis of generalizeable understanding

Allowing bias into the design may lead to a better understanding of the implementation process.

Reflective process throughout intervention

Endpoints/measurement set in advance

Time function or endpoints may vary within boundaries

Reflection both within and across sites

Modify endpoints based on results

Incorporate reflection periods into study design.

No comparisons, internal focus

Comparisons between arms

Comparisons based on 'content analysis' of internal understandings and lessons

Use of qualitative methods to probe themes from implementation experiences between sites