Development of a method for qualitative data integration to advance implementation science within research consortia

Abstract

Abstract Background Methods of integrating qualitative data across diverse studies and within multi-site research consortia are less developed than those for integrating quantitative data. The development ofsuchmethods is essential to support the data exchange needed for cross-study qualitative inquiry and given the increasing emphasis on data sharing and open science. We describe methods for qualitative data integration within the National Cancer Institute’s Improving the Management of symPtoms During And following Cancer Treatment (IMPACT) Consortium funded by the Cancer MoonshotSM. Data collection and analysis were guided by the Consolidated Framework for Implementation Research (CFIR). Our case study highlights potential solutions for unique challenges faced when integrating qualitative data across multiple settings in a research consortium. Methods The IMPACT consortium is comprised of three research centers (RCs) each conducting pragmatic trials examining the effectiveness of routine symptom management on patient-centered outcomes. After reaching consensus on use of CFIR as the common implementation determinant framework, RCs developed a semi-structured interview guide and tailored it to features of their healthcare setting and symptom management interventions. RCs conducted interviews/focus groups with healthcare system partners to examine contextual factors impacting implementation. RCs exchanged 1–2 transcripts (n = 5 total) for purposes of pilot testing the methodology. Results Given the heterogeneity of study settings and contexts, it was challenging to simultaneously assign codes at both domain and construct levels and the process was resource intensive. Recommendations include employing a common framework for data collection and analyses from the outset, coding at domain level first and then incorporating construct codes, and centralizing processes via a coordinating center (or similar entity) and combining coded transcripts using qualitative software. We also generated an iteratively refined codebook that employed the CFIR schema and incorporated CFIR 2.0 to provide detailed guidance for coders conducting cross-study qualitative inquiry. Conclusions Limited guidance exists on how to support qualitative data integration, data exchange, and sharing across multiple studies. This paper describes a systematic method for employing an implementation determinant framework-guided approach to foster data integration. This methodology can be adopted by other research consortia to support qualitative data integration, cross-site qualitative inquiry, and generate improved understanding of evidence-based intervention implementation.

Keywords

• Cancer
• Implementation science
• Qualitative methods
• Consolidated Framework for Implementation Research
• Data integration