Is there a weekend effect in colonoscopy? A population-based study using the National Endoscopy Database


Year:

Session type:

Linda Sharp1, Liya Lu1, Jamie Catlow2, Raphael Broughton3, Peter Rogers4
1Newcastle University, 2North Tees & Hartlepool NHS Foundation Trust, 3Royal College of Physicians (RCP), 4Other

Abstract

Background

Over the past few years, hospital activity at weekends has increased, and may rise further due to the COVID-19 pandemic. However, some studies have suggested poorer outcomes for patients seen at weekends (a so called “weekend effect”). Colonoscopy is key for colorectal cancer prevention (by detecting pre-cancerous polyps) and diagnosis. Weekend working is sometimes used to deal with long colonoscopy waiting lists, with procedures often undertaken by endoscopists who are not part of a Trust’s regular workforce. Using data from the National Endoscopy Database, we investigated whether (1) colonoscopy patient characteristics on weekdays and weekends differ and (2) polyp detection varies between weekend and weekday procedures, after accounting for patient characteristics.

Method

We conducted a retrospective cross-sectional study including colonoscopies conducted 01/01-04/04/2019.  Procedures had to be non-emergencies, undertaken on patients aged ≥18 years, and involve complete examinations. Primary outcome measures were mean number of polyps (MNP) and polyp detection rate (PDR), two key performance indicators of colonoscopy quality. Multi-level mixed-effect negative binomial and logistic regression analyses, fitting endoscopist as a random effect, were used to examine associations between weekend procedure and these outcomes.

Results

92,879 colonoscopies were performed by 2,496 endoscopists. 19,977 (21.5%) were performed at weekends. Patients who had weekend procedures were more often female and less often seen for screening-related procedures (χ2 tests:  P<0.05); age differed little. Compared to weekday procedures, MNP was significantly lower for weekend procedures, after adjusting for patient age, sex and indication for procedure (Incidence Rate Ratio (IRR)=0.87, 95%CI 0.84-0.90). A similar significant reduction was found for PDR at weekend (Odds ratio (OR)=0.86, 95%CI 0.82-0.90).

Conclusion

Colonoscopy polyp detection is significantly lower at weekends, after adjusting for case-mix. This has implications for colorectal cancer prevention and diagnosis.  Research is urgently required to understand the reasons for this unwarranted variation, especially if more weekend working is introduced to address pandemic-related backlogs.

Impact statement

This large population-based study shows that detection of polyps at colonoscopy is significantly lower at weekends, after adjusting for case-mix.