Methods: We designed a structure and process that led to an evidence-based and consensus-driven intervention and research design. We weighted and applied selection criteria, grounded in prior research, to invite 3 FQHCs to join this study as intervention sites. We engaged FQHCs in identifying a minimum set of evidence-based activities and in agreeing to a uniform set of measures for evaluation. Our structure included face-to-face meetings, site visits, regular conference calls complemented with current data and ongoing feedback. Using a multiple case study approach with an embedded pre/post difference-in-difference analysis, we gathered prospective qualitative and quantitative data from multiple sources over multiple points in time. We obtained data from 6 FQHCs in 3 states/jurisdictions, 592 low-income and underserved pediatric patients with poorly-controlled, moderate-to-severe asthma, 449 caretaker attitude and satisfaction surveys, interviews (100) and follow-up web-based surveys with FQHC staff and leadership, web-based logs by asthma educators, and telephone interviews with state primary care associations. We analyzed qualitative process data with NVivo v. 10.
Results: Findings indicate smooth integration of CHAMPS into intervention FQHCs. The engagement of FQHCs in decision making and in the collaborative design of a common intervention protocol that aligned both with research priorities and clinic workflows contributed to success. Our adaptive research design allowed us to effectively respond to unanticipated challenges as they arose. FQHC capacity to undertake CHAMPS and succeed varied by site and evolved over time in unexpected directions but our approach promoted sustainable practice changes.
Conclusion: Our case study design, with strengthened external validity and a common research and intervention protocol across study sites, offers a rigorous and pragmatic option for translational and implementation research by balancing external validity with fidelity to the evidence base. It also generated actionable data for future dissemination.