Quality evaluation of guidelines on genetic screening, surveillance and management of hereditary colorectal cancer

BACKGROUND: We examined the methodological quality of guidelines on colorectal cancer genetic susceptibility syndromes. METHODS: PubMed, EMBASE, and Google were searched up to July 2010. Adjourned guidelines in English were included. The Appraisal of Guidelines, Research and Evaluation (AGREE) instrument was used to assess their quality, and the reported evidence base of the guidelines was evaluated. RESULTS: The search yielded 17 eligible guidelines: 11 were developed by independent associations, while 6 had national\state endorsement. Only three guidelines performed satisfactorily, achieving a score >50% in all 6 AGREE domains. Mean standardized scores for the 6 AGREE domains were: 'scope and purpose', 83.9 ± 22.5%; 'stakeholder involvement', 35.6 ± 24.9%; 'rigour of development', 48.6 ± 25.3%; 'clarity and presentation', 71.6 ± 19.3%; 'applicability', 33.8 ± 30.1%; 'editorial independence', 42.2 ± 39.7%. Guidelines with national endorsement performed better in all the domains, with a statistically significant difference in three domains. Fifteen guidelines out of 17 were found to be based on research evidence. CONCLUSIONS: There is scope, in many areas, for improving the guidelines analysed, among which are the involvement of various professional figures and patients' representatives, and policies for their application. The AGREE instrument is a useful tool and could also be used profitably by guideline developers to improve the quality of recommendations.