Variability in Computer-Aided Detection effect on Adenoma Detection Rate in randomized controlled trials: a meta-regression analysis

Aims The assistance of computer-aided detection (CADe) systems during colonoscopy may increase adenoma detection rate (ADR), theoretically reducing the risk of post-colonoscopy colorectal cancer (PCCRC). Although the promising results in different randomized trials, both the variability of the magnitude of relative effect of CADe systems across the previous studies, and the contradicting results in the first real-life experiences, highlighted a clear gap of knowledge when looking for those factors possibly explaining these fluctuating results. The aim of our analysis was to investigate the different variables possibly affecting the impact of CADe-assisted colonoscopy and its effect on ADR.

Methods We searched MEDLINE, EMBASE, and Scopus databases until July 2023 for RCTs reporting diagnostic accuracy of CADe systems in the detection of colorectal neoplasia (PROSPERO: CRD42023462438). The main outcome was pooled adenoma detection rate (ADR). We calculated risk ratios (RRs), and performed meta-regression analysis to explore thesources of heterogeneity. The variables examined included factors with an impact on expected prevalence of adenomas across the study populations, such as gender, age and colonoscopy indication. We also included both key (ADR), and minor (Withdrawal time) performance measures considered as quality indicators for colonoscopy.

Results Twenty-three randomized controlled trials (RCTs) on 19,077 patients were include. ADR was higher in the CADe group than in the standard colonoscopy group (45.83% versus 38.28%; RR 1.22 [95% CI 1.14-1.29]) with substantial level of heterogeneity (I²=67.69%). In univariable meta-regression analysis, patient age, ADR in control arms, and withdrawal time were the strongest predictors of CADe effect on ADR (P<.001), whereas FIT as an indication for colonoscopy was only suggestively associated with the outcome (P=0.098), and was included in the multivariable analysis. The proportion of male patients was not apparently associated with the CADe effect on ADR. In multivariable meta-regression, ADR in control arms, and withdrawal time were simultaneous significant predictors of the proportion of the CADe effect on ADR.

Conclusions In conclusion, the substantial level of heterogeneity found appeared to be associated with variability in colonoscopy quality performances across the studies. As a matter of fact, across all the studies in which the CADe system showed no relative effect, the baseline ADR was higher than 60% suggesting a possible “ceiling effect” with little room left for improvement in the intervention group. On the other hand, endoscopists with lower quality performances are going to benefit the most from the use of CADe systems during colonoscopy, irrespectively from the expected adenoma prevalence across different populations. Thus, the implementation of CADe-assisted colonoscopy is supposed to help in reducing the gap in term of detection performances between high- and low- detectors.

Conflicts of interest

Authors do not have any conflict of interest to disclose.

Publication History

Article published online:
15 April 2024