Design and Setting
This study is based at the Implementation Science Center for Cancer Control Equity (ISCCCE), a National Cancer Institute-funded center with collaboration between the Massachusetts League of Community Health Centers (Mass League), Harvard T.H. Chan School of Public Health, Massachusetts General Hospital, and Dana Farber Cancer Institute. The Mass League is the Primary Care Association that provides workforce development, policy analysis, information technology development, clinical quality improvement (QI), training, and education to 52 FQHCs across Massachusetts (14). The core ISCCCE structure includes an Implementation Lab that is responsible for building the research capacity of 31 FQHCs that share a common population management platform and for supporting engagement in pilot studies.
We used an explanatory sequential mixed methods design to assess the implementation of cancer prevention EBIs in FQHCs (15). The study began with a quantitative survey of staff to determine the frequency with which established EBIs are being implemented. These items were embedded in a broader organizational survey that assessed aspects of the inner setting, outer setting, and characteristics of the individual staff. Following the survey, we conducted qualitative one-on-one interviews with a sub-sample of staff to explore how the EBIs selected on the survey were implemented. The study was approved by the Harvard Longwood Campus Institutional Review Board. The Good Reporting of a Mixed Methods Study Checklist was used to ensure study rigor.
The study was guided by two implementation science frameworks. First, Proctor’s Model for Implementation Outcomes (16), describes the core implementation outcomes that lay on the pathway between interventions and intended clinical patient outcomes. Our survey focused on adoption (e.g., the degree to which Massachusetts FQHCs offer EBIs) and penetration (e.g., the proportion of eligible patients offered or referred to these EBIs). The Consolidated Framework for Implementation Research (CFIR) guided our qualitative exploration (17, 18). This determinants framework describes the potential multilevel contextual influences on EBI implementation, including aspects of the implementation process, characteristics of the intervention itself (e.g., complexity), characteristics of individuals (e.g., staff responsible for delivery), the inner setting (e.g., structures and culture within the FQHC), and the outer setting (e.g., influences outside the FQHC).
Quantitative Surveys
Participants
The ISCCCE team invited staff members from 31 Massachusetts FQHCs to participate. We sampled one to three people within five diverse job type categories – leadership, clinical, QI, community direct-service, or community outreach/engagement. Respondents also self-reported their role on the survey, which includes these categories as well as management, administrative, referrals, consultation, and data support. The survey used branching logic to have respondents who selected community direct-service, community outreach/engagement, or management roles complete items pertaining to EBI implementation given the alignment of their positions to the community partnership aims of this study.
Measures
The research team fielded a 10–20-minute online survey via REDCap between November 2020 and August 2021. To ensure culturally appropriate engagement and build rapport with each FQHC, a Mass League leader emailed notification of the survey several days before the invitation was emailed from the ISCCCE principal investigator via REDCap. Up to three reminders to complete the survey were made by email following the initial invitation.
The survey included close-ended items to identify specific primary prevention EBIs (see Fig. 1) that can be delivered by FQHCs or community-based organizations focused on nutrition, physical activity, and tobacco cessation. We included EBIs that focused on prevention of other chronic diseases (e.g. diabetes) if the primary behavior change targets of intervention were nutrition and physical activity. We included individual or group-delivered EBIs, drawing from The Community Guide (19) as well as input from the research team. Following Proctor’s Model for Implementation Outcomes (16), staff were asked to estimate the proportion of eligible people that are impacted by each intervention (i.e., penetration) on a five-point scale, ranging from “none” to “most or all.” We utilized this self-reported survey measure because it was easily understood by a wide range of FQHC staff with and without prior engagement in research and allowed participants to share their perception of patient impact (vs. yes/no adoption). Surveys also included open-ended questions for participants to describe interventions that were not pre-specified or were “home-grown” (e.g., developed by the FQHC). We also measured participants’ FQHC tenure and demographics (e.g., gender, race/ethnicity, age). Cognitive interviews were conducted with three staff from centers not included in the study sample to review the EBIs included and improve survey language and flow prior to launch of the survey. Respondents were compensated $25 for participation. FQHCs that were new to ISCCCE research projects also received $100 site-level incentives to provide meals or snacks for staff.
Analysis
We calculated descriptive statistics for participants’ time in current role and demographics (e.g., race, ethnicity, age, gender), as well as number of eligible people that are offered or referred to the intervention. EBI adoption was summarized at the site level.
Qualitative Interviews
Participants
Researchers invited 23 staff members (1–4/site) from 12 sites to participate in qualitative interviews, beginning with those who completed the survey and then adding interviewee referrals of individuals, such as those in quality improvement and community engagement roles who had in-depth knowledge of FQHC implementation and outreach processes for the types of prevention interventions under investigation. Interviewees included QI and population health leaders, staff managing prevention programming and community engagement activities, and practitioners delivering interventions.
Measures
We explored the experience of adoption and implementation of cancer prevention EBIs in greater depth via one-hour, one-on-one semi-structured interviews. Interviews were audio-recorded and then transcribed for accuracy. Researchers asked interviewees to describe how and why each intervention was initially adopted, implemented, and sustained using adoption data from quantitative surveys to structure interview prompts. Participants were also asked whether they implemented other tobacco, nutrition, and physical activity interventions in the past or hoped to in the future. To address Aim 2, participants were asked how partners were involved in implementation of each intervention and where intervention activities were situated (e.g., FQHC or another community setting). Contextual influences were explored using interview probes structured following CFIR to explore multilevel determinants on implementation (17, 18), APPENDIX A). Participants were compensated $50 for participation in interviews.
Analysis
We utilized reflexive, thematic approaches, as described by Braun and Clarke (20), beginning deductively with codes from CFIR, then inductively coding additional categories using NVivo Software. To increase credibility and rigor, we utilized researcher triangulation; interviews were coded by the principal investigator and the research project manager, who both have backgrounds conducting public health and implementation science research, to ensure deep engagement with the data and integration of multiple perspectives (21). A third member of the research team, who has a background in medical anthropology, created summaries of codes prioritized for this analysis. Drawing on concepts of information power, we determined that our sample size would be sufficient given fairly broad research goals, a somewhat homogeneous sample (in terms of work focus), rich data collection, and strong reliance on an existing framework (22). Interpretation of results was support by three co-authors who are FQHC staff and researchers with extensive experience in community-based cancer implementation science.