1093 — Developing and Testing an Electronic Measure of Screening Colonoscopy Overuse in a Large Integrated Healthcare System

Saini SD, VA Ann Arbor HSR&D Center for Clinical Management Research and University of Michigan; Dominitz JA, VA Puget Sound Health Care System and the University of Washington School of Medicine; Fisher DA, Durham VA Center for Health Services Research in Primary Care and Duke University Medical Center; Vijan S, VA Ann Arbor HSR&D Center for Clinical Management Research and University of Michigan; Pittmann KS, VA National Center for Health Promotion and Disease Prevention; Powell AA,  Schoenfeld P, VA Ann Arbor HSR&D Center for Clinical Management Research and University of Michigan; Francis J, VA Clinical Analytics and Reporting; Kinsinger L, VA National Center for Health Promotion and Disease Prevention; Kerr EA, VA Ann Arbor HSR&D Center for Clinical Management Research and University of Michigan

Objectives:
Most existing performance measures focus on underuse of care, but there is growing interest in identifying and reducing overuse. We sought to develop an electronic measure of screening colonoscopy overuse and quantify overuse in VA.

Methods:
We convened an expert workgroup to specify measures of screening colonoscopy overuse. We then developed a high-specificity electronic algorithm to identify screening colonoscopies that met measure definitions, using data from FY11 to FY13. This algorithm was validated using detailed chart review from a random national sample of colonoscopies. We examined variation in overuse and temporal trends in overuse at the level of the facility and VISN. We also performed multivariable analysis to identify facility-level predictors of overuse.

Results:
The electronic algorithm achieved a specificity of 97% for overuse compared to gold-standard chart review (N = 2,915). 88,754 average-risk, screening colonoscopies were electronically identified in FY13. 20,530 (23%) of these met the definition for probable (17%) or for possible (6%) overuse. The most common reasons for overuse of screening colonoscopy were performance of colonoscopy < 6 months after negative fecal occult blood test (35%) and performance of colonoscopy < 9 years after prior colonoscopy (31%). Substantial variation in overuse was noted between VISNs and facilities, with a nearly 2-fold difference between the maximum and minimum rates at the VISN level and a nearly 8-fold difference at the facility level. Furthermore, overuse was relatively stable over time. Between FY11 and FY13, all VISNs in the bottom quartile of performance (high overuse) remained in the bottom quartile, and all VISNs in the top quartile (low overuse) remained in the top quartile. Likewise, 70 of 122 facilities remained in the same quartile of performance for all 3 years, and 103 of 122 facilities improved or worsened by no more than one quartile. No facility-level characteristics were found to predict overuse of screening colonoscopy.

Implications:
Overuse of screening colonoscopy is common in a large integrated healthcare system that utilizes underuse measures to encourage screening uptake. Overuse varies widely between facilities and VISNs but is relatively stable over time.

Impacts:
Overuse measures, such as those we have specified through a consensus workgroup process, could be combined with underuse measures to enhance the appropriateness of colorectal cancer screening and other preventive services.