Article type
Year
Abstract
Background: Ideally, clinical practice guidelines (CPGs) should be based on high quality evidence such as Cochrane systematic reviews. However, there is concern that only few reviews are being used to support guideline recommendations. Objectives: To estimate what proportion of the Australian-New Zealand Guidelines for Chronic Kidney Disease (CKD) are supported by Cochrane and non-Cochrane reviews and identify strategies for improving the contribution Cochrane reviews can make to guidelines groups. Methods: All CPGs (published 2004–2007) produced by Caring for Australasians with Renal Impairment (CARI) were reviewed. We assessed the guideline subtopics for the following questions: (1) Was a Cochrane or non-Cochrane review used? (2) Did the guideline recommendations concur with the Cochrane recommendations? (3) Were there relevant Cochrane reviews available which were not used in the guideline, and, if so, what were the reasons? Results: We identified 131 CARI guidelines for CKD which included dialysis and transplantation topics. Of the 30 guidelines which cited systematic review evidence, 17 guidelines cited Cochrane reviews. Ten guideline recommendations were consistent with Cochrane recommendations. One guideline recommendation offered a different recommendation to the review, which was explainable because the guideline author stated that one large, well-designed RCT was not included in the review. The remaining six guidelines cited outdated reviews of which five reviews have since been updated after the guidelines were published. Cochrane reviews were not cited in 114 (87.0%) guidelines for four reasons: no reviews were available on the guideline topic (64/131); the guideline did not address an intervention question (28/131); the guideline was written before the relevant review was published (21/131); and the guideline did not cite the review (1/131). Conclusions: Currently, less than 25% of CARI guideline recommendations were supported by systematic reviews. We identified reasons why Cochrane reviews were not cited. No relevant Cochrane reviews existed for 70% of guidelines, and less than 1% was due to non-citing of a published review. If we want reviews to inform guidelines, we need a more comprehensive range of reviews on interventions and develop reviews of non-intervention topics including diagnostic tests.