How evidence-based are clinical practice guidelines in anaesthesiology: the cases of preoperative care and difficult airway management

Article type
Authors
Ciapponi A1, Tapia López E1, Virgilio SA1, Perelli L1, Cohen Arazi H1, Bardach A1
1Institute for Clinical Effectiveness and Health Policy (IECS-CONICET)
Abstract
Background: clinical practice guidelines (CPGs) in anaesthesiology have diverse methodological quality and have used various evidence-rating systems. It is difficult to properly compare their recommendations without a systematic grading approach.

Objectives: to analyze and describe evidence-based CPGs' (EB-CPGs) approach on topics frequently addressed in anaesthesiology.

Methods: overview of EB-CPGs on two topics: general preoperative care (PC) and difficult airway (DA) management, following Cochrane methods.

Inclusion criteria:
1) description of the development expert panel;
2) use of standard methods for identification, data collection and study risk of bias assessment;
3) reporting of the level of evidence that supports each recommendation.

We searched CPGs published in the last 10 years without language limitation in the main electronic databases, meta-search engines, specific sources of CPG and consulting experts. Pairs of independent review authors selected EB-CPGs, rated their methodological quality using the AGREE-II instrument and classified the strength and the level of evidence for each recommendation according to the GRADE methodology. We resolved discrepancies by consensus.

Results: we identified 2262 references and included 16 EB-CPGs for PC and from 2591 references found for VA, we included 11 EB-CPGs. Only nine (56%) and eight (73%) respectively had searches during the last five years. The most used grading system was GRADE (44% for PC and 64% for DA). The AGREE-II average, which scores for rigour in the elaboration, was 56 ± 23%. The worst score was for the ‘applicability’ domain (25 ± 24% for PC and 37 ± 10% for DA) of this tool (Table 1). We found a low proportion of diagnostic recommendations supported by high/moderate certainty of evidence, and a relatively higher one for therapeutic/preventive recommendations. The proportion of 'strong' diagnostic recommendations were 41% to 51% and 76% respectively (Table 2). We found higher consistency among guidelines in the direction (for or against) and strength of the recommendations than in their supporting level of evidence.

Conclusions: the included EB-CPGs showed significant heterogeneity in terms of quality and rating systems. Applicability and monitoring were deficient; however, almost two-thirds of EB-CPGs were updated in the last five years. These findings reveal several opportunities for quality improvement.

Patient or healthcare consumer involvement: none, since it was a methodological paper about published guidelines.