OBJECTIVE: The objective of this review was to assess the association between quality indicators used to evaluate individual colonoscopist performance and subsequent interval colorectal cancers (CRCs) in patients participating in bowel cancer screening programs.
INTRODUCTION: Colorectal cancer is a leading cause of cancer death. Bowel cancer screening has been shown to reduce CRC mortality and morbidity, and has therefore been introduced in many countries. Endoscopy societies have developed quality assurance guidelines and guidelines on quality indicators for screening colonoscopies. These quality indicators need to be validated against a relevant outcome to assess their value.
INCLUSION CRITERIA: We included studies on screening colonoscopies conducted on participants in a bowel cancer screening program, regardless of comorbidity. Studies on procedures performed on patients with known CRC, hereditary nonpolyposis colorectal cancer or familial adenomatous polyposis were excluded. We also included studies evaluating the quality indicators of withdrawal time (WT), cecal intubation rate (CIR) and adenoma detection rate (ADR). The search did not reveal any studies evaluating the quality indicators polyp retrieval rate and incomplete adenoma resection/incomplete polyp resection. Only studies with interval CRC as an outcome were included (i.e. CRC diagnosed after a negative screening colonoscopy, but before the next recommended examination date).
METHODS: Published studies were searched in: MEDLINE, Embase, Web of Science and CINAHL. Unpublished studies were searched in: OpenGrey and Grey Literature Report. The sources were searched from 1980 to2018. Data were extracted using the JBI critical appraisal checklist for analytical cross sectional studies. A meta-analysis was conducted based on three of the colonoscopist dependent quality indicators: WT, CIR and ADR.
RESULTS: Seven prospective and retrospective cohort studies were included out of 2373 papers identified after duplicates were removed. The included studies were on bowel cancer screening programs with colonoscopy as the primary screening tool, resulting in the inclusion of a total of 616,390 screening colonoscopies performed by 1431 colonoscopists and 2319 subsequent interval CRCs. Six studies were assessed as high-quality studies, and one study was of low quality. The meta-analysis on WT revealed a 61% lower risk of interval CRC among the patients if the mean WT per colonoscopist was >6 minutes as compared to a mean WT of <6 minutes (RR: 0.39 [95% CI: 0.23 - 0.66]). The meta-analysis on CIR revealed a 31% lower risk of interval CRC among the patients if the CIR per colonoscopist was ≥90% as compared to a CIR of <85% (RR: 0.69 [95% CI: 0.56 - 0.83]). One of two meta-analyses on the individual colonoscopist ADR suggested that this should be 15-19%, as compared to an ADR <10% (RR: 0.77 [95% CI: 0.62 - 0.96]), in order to significantly reduce the risk of interval CRC. The other meta-analysis on ADR revealed a significant association between an individual colonoscopist ADR of ≥25% and a lower risk of interval CRC as compared to an ADR of <25% (RR: 0.51 [95% CI: 0.33 - 0.80]). The meta-analyses on WT and CIR showed no heterogeneity concerning the significant results (I = 0.0%). A high variability across studies due to heterogeneity concerning an ADR of ≥20% resulted in an I = 59.9%, and an I = 63.2% for an ADR of ≥25%.
CONCLUSIONS: To minimize the risk of interval CRC, it may be recommended that WT and CIRs are monitored in bowel cancer screening programs, with an optimal individual colonoscopist mean withdrawal time of >6 minutes and a cecal intubation rate of ≥90%. In bowel cancer screening programs using colonoscopy as their primary screening tool, it may be recommended that the individual colonoscopist ADR be 15-19% or better ≥25% to minimize the risk of interval CRC.
