Updating an Electronic Measure of Screening Colonoscopy Overuse in a Large Integrated Healthcare System to Examine Trends and Variation in Overuse

Abstract

Research ObjectiveOveruse of screening colonoscopy can lead to patient harm and wasteful use of resources. We previously developed an ICD‐9 based measure to detect screening colonoscopy overuse in a large integrated healthcare system. This measure was highly specific, suggesting that cases identified as overuse were true positives, but had low sensitivity (likely to miss cases of overuse). We sought to update and test this previously validated measure for use in ICD‐10 and assess trends and variation in colonoscopy overuse in a large integrated healthcare delivery system.Study DesignRetrospective cohort study of Veterans Health Administration (VHA) administrative data, with measure validation via manual record review.Population StudiedIndex screening colonoscopy encounters at 117 VHA facilities in 2017.Principal Findings269,572 colonoscopies were performed in VHA in 2017. After applying exclusion criteria (non‐index procedures, procedures in patients at increased risk of colorectal cancer, inpatient procedures, colonoscopy for non‐screening indication within 12 months), 88,143 colonoscopy encounters remained. Validating the updated ICD‐10 based electronic overuse measure (“Updated Measure”) against the gold standard of manual record review in a random sample of 511 cases, the Updated Measure had similar specificity to the ICD‐9 based measure (96% vs. 97%) but was significantly more sensitive (92% vs. 20%). The sensitivity and specificity of the Updated Measure were robust both among sites with the lowest levels of overuse (sensitivity 100%, specificity 97%) and sites with the highest levels of overuse (sensitivity 93%, specificity 97%).Applying the Updated Measure, 24.5% of screening colonoscopy encounters (21,600/88,143) met the definition of overuse (as defined in J Gen Intern Med 2016;31[Suppl 1]:53–60), similar to levels in 2011–13 (23%). Of these 21,600 colonoscopies meeting overuse criteria, the top 2 reasons for overuse in both periods were screening colonoscopy performed <9 years after previous colonoscopy (45% in 2017 vs. 35% in 2011–13) and screening colonoscopy performed <6 months after negative FOBT (23% in 2017 vs. 31% in 2011–13). Median facility‐level overuse was 22.5% (IQR 19.1%–26.3%), with four to five‐fold variability among facilities. The performance of 55/117 (47%) of facilities remained stable over time (same quartile), while the performance of 93/117 (79%) facilities improved or worsened by no more than one quartile. The 13 lowest performing facilities remained in the bottom quartile of performance in both time periods.ConclusionsOur updated ICD‐10 based measure reliably measures screening colonoscopy overuse with similar specificity but markedly better sensitivity than the ICD‐9 based measure, allowing VHA to track facility‐level performance over time. Despite increased focus on reducing low value care and enhancing access, levels of colonoscopy overuse in VHA remained stable between 2011–13 and 2017, with continued facility‐level variability.Implications for Policy or PracticeThese findings are among the first to suggest that ICD‐10 codes can substantially improve the validity of quality measures relying on administrative data. Moreover, this updated measure can be successfully deployed by large healthcare systems to track facility‐level procedural overuse over time. This data can facilitate efforts to improve care quality and access and, in integrated healthcare systems, expand capacity by limiting low‐value colonoscopy.Primary Funding SourceDepartment of Veterans Affairs.