Background: Evidence-based guideline development presumes a foundation of scientific evidence for recommendations. However, there are fields of research that are not particularly suitable to study patients in a (randomised) controlled setting. For example, fragile patients. In those situations, a guideline can still be developed to help clinicians make informed treatment choices.

Objectives: In this study, we aim to evaluate how guidelines on topics with little evidence are developed, if they include recommendations, and how decisions are made.

Methods: We studied a cohort of clinical guidelines published by NICE (n=182) and the Knowledge Institute of Medical Specialists (n=248). Criteria for inclusion were: the guideline described a subject on fragile patients (children, frail elderly, mentally incompetent patients), life-threatening situations, and low- prevalence diseases (5 per 100.000 patients). Guidelines without GRADE evaluations were excluded. We assessed the grading of literature and, if recommendations were formulated, the strength of the recommendation, and the evidence-to-decision framework.

Results: Out of 330 guidelines 86 fulfilled the inclusion criteria. Evidence was GRADEd low to very low in over 95%. Recommendations were strong 56%, conditional, 26%, or weak 18%. 0% of guidelines made no recommendations because of the lack of evidence. Transparent methods from evidence to decisions were lacking. Factors supporting decisions were experience of care providers, patient perspective, costs, and duration of an action. A framework from evidence to decisions was missing in X% of guidelines.

Conclusions: Evidence in our cohort of guidelines was graded low in most cases, nevertheless the recommendations were frequently strong. Although clinical expertise is part of evidence-based care, it would be useful to have an insight on which considerations the decisions were based. A framework from evidence to decisions, like GRADE proposes, would help to make this process more transparent.