Although multimodal interventions like extra training with periodic feedback showed some signs of improving ADR, withdrawal time monitoring was not significantly associated with a better detection rate, reported Anshul Arora, MD, of Western University, London, Ont., and colleagues.
“Given the increased risk of postcolonoscopy colorectal cancer associated with low ADR, improving [this performance metric] has become a major focus for quality improvement,” the investigators wrote in Clinical Gastroenterology and Hepatology.
They noted that “numerous strategies” have been evaluated for this purpose, which may be sorted into three groups: endoscopy unit–level interventions (i.e., system changes), procedure-targeted interventions (i.e., technique changes), and technology-based interventions.
“Of these categories, endoscopy unit–level interventions are perhaps the easiest to implement widely because they generally require fewer changes in the technical aspect of how a colonoscopy is performed,” the investigators wrote. “Thus, the objective of this study was to conduct a systematic review and meta-analysis to identify endoscopy unit–level interventions aimed at improving ADRs and their effectiveness.”
To this end, Dr. Arora and colleagues analyzed data from 34 randomized controlled trials and observational studies involving 1,501 endoscopists and 371,041 procedures. They evaluated the relationship between ADR and implementation of four interventions: a performance report card, a multimodal intervention (e.g., training sessions with periodic feedback), presence of an additional observer, and withdrawal time monitoring.
Provision of report cards was associated with the greatest improvement in ADR, at 28% (odds ratio, 1.28; 95% confidence interval, 1.13-1.45; P less than .001), followed by presence of an additional observer, which bumped ADR by 25% (OR, 1.25; 95% CI, 1.09-1.43; P = .002). The impact of multimodal interventions was “borderline significant,” the investigators wrote, with an 18% improvement in ADR (OR, 1.18; 95% CI, 1.00-1.40; P = .05). In contrast, withdrawal time monitoring showed no significant benefit (OR, 1.35; 95% CI, 0.93-1.96; P = .11).
In their discussion, Dr. Arora and colleagues offered guidance on the use of report cards, which were associated with the greatest improvement in ADR.
“We found that benchmarking individual endoscopists against their peers was important for improving ADR performance because this was the common thread among all report card–based interventions,” they wrote. “In terms of the method of delivery for feedback, only one study used public reporting of colonoscopy quality indicators, whereas the rest delivered report cards privately to physicians. This suggests that confidential feedback did not impede self-improvement, which is desirable to avoid stigmatization of low ADR performers.”
The findings also suggest that additional observers can boost ADR without specialized training.
“[The benefit of an additional observer] may be explained by the presence of a second set of eyes to identify polyps or, more pragmatically, by the Hawthorne effect, whereby endoscopists may be more careful because they know someone else is watching the screen,” the investigators wrote. “Regardless, extra training for the observer does not seem to be necessary because the three RCTs [evaluating this intervention] all used endoscopy nurses who did not receive any additional polyp detection training. Thus, endoscopy unit nurses should be encouraged to speak up should they see a polyp the endoscopist missed.”
The investigators disclosed no conflicts of interest.
