Audit plus peer comparison improved documentation of tumour features and resection completeness in non-muscle-invasive bladder cancer surgery.

This cluster randomised controlled trial enrolled 14,915 patients across 201 sites undergoing transurethral resection of bladder tumour surgery for non-muscle-invasive bladder cancer. One hundred sites were randomised to receive audit and feedback plus peer comparison and education, while 101 sites served as the control group receiving audit alone. The study evaluated four coprimary outcomes: single-instillation chemotherapy use, detrusor muscle sampling, documentation of tumour features, and resection completeness, alongside early recurrence as a secondary outcome.

The intervention arm demonstrated a statistically significant improvement in the documentation of tumour features, with an adjusted mean difference of 6.0 (95% CI: 1.8, 10; p = 0.005). Similarly, documentation of resection completeness improved significantly in the intervention arm, showing an adjusted mean difference of 5.5 (95% CI: 1.5, 9.5; p = 0.007). In contrast, there was no statistically significant difference between arms regarding the use of single-instillation chemotherapy (adjusted mean difference 0.3; 95% CI: -4.7, 5.3; p = 0.9) or detrusor muscle sampling (adjusted mean difference 2.6; 95% CI: -1.3, 6.4; p = 0.2). Early recurrence rates also showed no significant difference between groups (adjusted odds ratio 1.02; 95% CI: 0.8, 1.4; p = 0.9).

Safety data, including adverse events, serious adverse events, discontinuations, and tolerability, were not reported in the study. A key limitation noted is that improvements observed in the control arm may explain a lack of effect of the intervention in some outcomes. The findings suggest that while audit and feedback with education enhance the documentation of important surgical findings that influence clinical management, the intervention does not alter the performance of detrusor muscle sampling, adjuvant chemotherapy use, or early recurrence rates. Causality is supported by the randomised design, though results should be interpreted within the context of the large sample size and specific documentation metrics.