Clinical guidelines for acute stroke management: Which recommendations should remain consensus-based?

Although guideline recommendations should be based on the highest levels of evidence available, it may be appropriate for some recommendations to remain consensus-based. This study investigated stroke clinicians' and academics' opinions about which consensus-based recommendations from the Australian National Stroke Foundation Clinical Guidelines for Acute Stroke Management (2007) should remain as such, using a self-administered questionnaire. In the study, 43 people participated (62% response rate). Of the 50 recommendations presented, at least half the participants believed that 35 (70%) of these should remain consensus-based. In an adjusted multivariable analysis, recommendations perceived to be "highly likely'' or "likely'' to have an impact on patient outcomes were less likely to be perceived by respondents as appropriate to remain consensus-based (OR = 0.16 [95% CI 0.13-0.19], P < .001). Further, although females were less likely to state that recommendations should remain consensus-based (OR = 0.26 for female gender, [95% CI 0.080.88], P = .031), those with higher level of educational qualifications (OR = 1.45 per an incremental unit increase in education level [95% CI 1.04-2.03], P = .028) and those from nursing and medicine disciplines in comparison with allied health (OR = 25.2 for medical, nursing, and "other'' disciplines in comparison with allied health [95% CI 1.54-413.1], P = .024) were more likely to state that a recommendations should remain consensus-based. The results indicated there was agreement that it may be appropriate that a proportion of recommendations in clinical guidelines remain consensusbased. Assessing the views of what areas require more robust research evidence according to academic and clinical experts may allow for prioritization and optimal allocation of scarce research resources.