Availability and Quality of Colonoscopy Resources and the Colorectal Cancer Diagnostic Interval
Background: There is concern that patients are waiting too long to be diagnosed with colorectal cancer (CRC) after presenting to the healthcare system. Previous research has found issues with the availability and quality of colonoscopy resources, a key procedure used to diagnose CRC, including long colonoscopy wait times, regional variations in colonoscopy utilization, and variations in colonoscopy quality. However, no research has described colonoscopy resource availability and quality and its relationship with the CRC diagnostic interval. Methods: This thesis examined colonoscopy resource availability and quality and its relationship with the CRC diagnostic interval in two phases using administrative health data from the Institute for Clinical Evaluative Sciences. Phase One used a population-based cross-sectional design to describe colonoscopy resource availability and quality in Ontario between 2007 and 2013. We described regional variations in colonoscopy resources and evaluated associations between colonoscopy resources and colonoscopy utilization. Phase Two used a population-based retrospective cohort design to describe the CRC diagnostic interval for Ontario CRC patients diagnosed between 2009 and 2012 and to evaluate the associations between colonoscopy resource availability and quality and the diagnostic interval. We used quantile regression to model these relationships while controlling for confounding and evaluating effect modification. Results: In Phase One, we found regional variation in the availability and quality of colonoscopy resources in Ontario. Significant correlations between colonoscopy resource characteristics and colonoscopy utilization indicated that reduced resource availability and quality were associated with reduced colonoscopy utilization. In Phase Two, the median CRC diagnostic interval was 84 days (90th percentile 323 days). We observed significant associations between the diagnostic interval and the availability and quality of colonoscopy resources. Patients residing in networks with lower colonoscopist density, colonoscopy completion rates and private clinic access had longer diagnostic intervals. These measures of colonoscopy resource availability and quality were also associated with the care that patients received within the diagnostic interval. Conclusion: This research demonstrated substantial variation in colonoscopy resource availability and quality in Ontario and identified the availability and quality of colonoscopy resources as an important determinant of the CRC diagnostic interval.