The quality assessment of clinical practice guidelines for intracranial aneurysms: a systematic appraisal.

Abstract

Intracranial aneurysms are common in adults. The relevant guidelines for patients with intracranial aneurysms aim to standardize the clinical practice and decision making for these patients. However, their management is controversial, and the quality of the guidelines has not been assessed. We aim to evaluate the quality of the guidelines for intracranial aneurysms as well as to compare and analyze the recommendations between different guidelines. Systematic searches were conducted to identify the guidelines for intracranial aneurysms from general electronic and guideline databases. Two independent reviewers identified the guidelines and extracted the data, and four reviewers independently evaluated the eligible guidelines through the AGREE II tool. Agreement among reviewers was measured using the intraclass correlation coefficient. A total of 12 guidelines, which were published from 1997 to 2016, were included. The agreement among reviewers was high (intraclass correlation coefficient, 0.85 (95% CI: 0.8-0.89)). The mean scores of six domains ranged from 16.5 to 57.5% (scope and purpose 57.5% (39-68%); stakeholder 30.8% (19-46%); rigor 31.9% (19-52%); clarity 57.2% (42-79%); applicability 24.9% (16-42%); and editorial independence: 16.5% (0-58%)). Furthermore, 202 recommendations related to intracranial aneurysms were collected from the included guidelines. Of these, 143 reported the quality of evidence and/or strength, and 119 reported both the quality of evidence and the strength. Of the 119 recommendations, there were six class A and 20 class B recommendations based on level III evidence. There were 12 recommendations in which the contents were similar between different guidelines and two recommendations with the opposite contents. The AGREE II scores of the guidelines for intracranial aneurysms were relatively low. The majority of recommendations were rated as classes A and B and based on levels II and III evidence. Approximately a fifth of strong recommendations was based on a low quality of evidence without interpretation or explanation.