Guidance documents for colorectal and anal cancer treatment: A systematic quality and reporting assessment

Evidence-based medicine is essential for clinical practice. Clinical practice guidelines (CPGs) and consensus statements (CSs) ought to follow a consistent methodology to underpin high-quality healthcare. We systematically analysed the quality and reporting of colorectal (CRC) and anal cancer CPGs and CSs.

Embase, MEDLINE, Scopus, Web of Science, the Cochrane Database of Systematic Reviews and online sources (59 professional society websites and eight guideline databases) were systematically searched following prospective registration (PROSPERO no. CRD42021286146) by two reviewers independently, without language restrictions. CPGs and CSs about CRC and anal cancer treatment were included from January 2018 to November 2021 and were assessed using the AGREE II tool (per cent of maximum score) and the RIGHT tool (per cent of total 35 items) for quality and reporting respectively.

The median overall quality and reporting of the 59 guidelines analysed were 55.0% (interquartile range 47.0-62.0) and 58% (interquartile range 50.0-67.9), respectively, with a proportion scoring less than half (<50%) for quality (20/59, 33.9%) and reporting (15/59, 25.4%). Guidance reported that following AGREE II methodology scored better on average than that without (AGREE II 77.7% vs. 47.6%, P = 0.001; RIGHT 50.0% vs. 33.9%, P = 0.001). Guidelines based on systematic reviews had better quality and reporting on average than those based on consensus (AGREE II 56.5% vs. 46.6%, P = 0.001; RIGHT 36.9% vs. 35.4%, P = 0.019).

The quality and reporting of colorectal and anal cancer treatment CPGs and CSs were poor. Despite AGREE II and RIGHT inherent methodological limitations, few high-quality guidelines were found. Despite wide variability in scoring different domains, they merit urgent improvement in all areas. It has also been demonstrated that CPGs and CSs should be underpinned by systematic reviews collecting the best available clinical research findings.