DEFINING POLICY DECISION(S) AND INTEGRATING CONTEXTUAL EVIDENCE: A MULTIPLE CASE STUDY OF COLORECTAL CANCER SCREENING POLICY DEVELOPMENT

Purpose:    Colorectal cancer (CRC) screening policy is complex, dynamic and multi-factorial. There are many potential screening modalities with new options emerging, however with only a handful of randomized controlled trials supporting the effectiveness of the fecal occult blood test, and more recently, flexible sigmoidoscopy, the principal scientific evidence base is slow to evolve.  Despite this, CRC screening programs vary widely across different jurisdictions.  Building on past work, the study's purpose was to compare/contrast CRC screening policy processes in five Canadian provinces regarding (1) what constituted the formal policy decision and (2) how contextual evidence was used.

Method:    A multiple case study design was employed with cases representing the policy-making processes in the five provinces (four provinces made major CRC screening policy announcements in 2007 while the fifth made recommendations in 2008). Key informant interviews and document analysis were the primary methods used. A purposive sample of clinical leaders, screening experts, administrative leaders and government officials were invited to participate. Documents sought included policy reports/announcements, planning documents, expert group guidance, and media reports. Interview data were transcribed and coded using NVivo qualitative analysis software.  

Result:    Considerable efforts were made to unpack what constituted a ‘policy-go’ decision, among many other important decisions in the process to establish a population-based CRC screening program in each province.  Formal policy announcements varied in terms of funding approaches/commitment, program comprehensiveness, and implementation timeline.  Several issues dominated policy deliberations, including optimal screening modality, clinical/laboratory infrastructure capacity, programmatic approach (e.g., public health vs. primary care), quality assurance methods and budget impact.  Although scientific evidence is lacking for many of these key issues, particularly as focus shifts from effectiveness to implementation issues, the five cases varied widely in terms of their approach and/or use of contextual evidence.  Some cases exhibited formal and strategic approaches to developing local contextual evidence to support their policy decisions, while other cases preferred to rely almost solely on externally derived evidence.  

Conclusion:    Population-based health policy decisions are complex, multi-factorial, dynamic and context-dependent.  While the scientific evidence base often provides the impetus to initiate major policy decisions, the notable lack of structured approaches or guidance for integrating contextual evidence to support key aspects of these decisions makes them decidedly less evidence-informed than is optimal and needs to be addressed.