Chapters 4 and 5 showed that while there are some similarities between guidelines, there are also some important areas of difference between medical practice guidelines in the Netherlands and England. This chapter draws together the findings from the case studies presented in those chapters and considers the implications drawing upon all four different cases. This chapter outlines six key areas for analysis that have emerged from the case studies: evidence, consensus, institutions, external factors, conflicts of interest and cost-effectiveness.

EVIDENCE: SOCIAL AND INSTITUTIONAL CONSTRUCTION OF OBJECTIVITY

A central aim of this study has been to understand how guideline developers select, use, appraise and understand medical evidence in constructing guidelines that purport to be evidence-based. It has endeavoured to add to a more nuanced understanding of evidence, conceptualising it as a social and institutional construction used in a collaborative and communicative process aimed at creating ‘objective facts’. This approach rejects other theories that offer either normative description of how evidence should be used or see evidence as the mechanical application of knowledge to specific medical cases. This section examines the results of the study as they relate to social and institutional choices in evidence selection, appraisal and use in the two case studies and across all four guidelines.

The case studies provide a varied image in how the different groups deal with evidence. The first factor that explains some of the difference between the Netherlands and England is the level of formality employed in searching for medical studies in medical databases. The institutional environment of NICE provides a more formal way of searching evidence, as they have a dedicated team of technical support staff and specialists on searching the medical literature. This results in a more formal selection of evidence in the guidelines in English case studies. A more formal way of searching for evidence in theory should produce more varied evidence to support decision making. However, in England the opposite was observed as guidelines were based on fewer studies and fewer types of studies than the Dutch guidelines. This was caused by the fact that the ‘search strings’, which are the words that are used to search these medical databases, were themselves based on specific clinical questions that had been formulated. The search strings copied very closely the precise words employed in the specific clinical questions posed by the guideline group.