Colonoscopy is used for primary colorectal cancer (CRC) screening, surveillance of patients with prior colon neoplasia or cancer and for diagnosis in patients with lower gastrointestinal (GI) tract symptoms or positive fecal occult blood test (FOBT). There is evidence that utilization of colonoscopy has increased dramatically in the past few years, largely due to increased rates of CRC screening. The effectiveness and safety of colonoscopy depends on the quality of examination, and a growing body of evidence suggests that the quality of colonoscopy in clinical practice varies. Quality of colonoscopy in VA medical centers is unknown.
In 2002, the Multi-Society Task Force on colorectal cancer (MSTF-CRC) published specific recommendations to improve the quality and effectiveness of colonoscopy. In 2007, the Quality Assurance Task Force of the National Colorectal Cancer Roundtable developed a colonoscopy reporting and data system (CO-RADS) to assist endoscopists in monitoring quality indicators in their practice. Structured, computerized endoscopic reporting systems provide an ideal tool for monitoring quality indicators. The Clinical Outcomes Research Initiative (CORI) developed a computerized endoscopic reporting system to study outcomes and quality of endoscopy in the United States. Data are transmitted electronically to a central data repository for analysis. The CORI consortium includes 70 diverse practice sites (including 9 VA centers) in 24 states, and receives about 260,000 endoscopic reports annually.
To determine if specific quality indicators are being reported by endoscopists at VA centers, and to measure actual performance relative to specific targets.
Aim 1: Measure adherence to completion of specific quality indicators for colonoscopy in 9 VA medical centers (2004-2007)
Hypothesis: There will be practice variation among endoscopists and areas for quality improvement will be identified
Aim 2: Measure actual performance relative to specific targets in the VA centers. Results will be compared to outcomes in non-VA centers which participate in CORI
Hypothesis: There will be practice variation. VA endoscopists will be similar to non-VA physicians in performance. Areas of quality improvement will be identified.
This study will analyze prospectively collected data from nine participating VA centers in the Clinical Outcomes Research Initiative (CORI) during the time period of January 1, 2004 through December 31, 2007. All VAMC endoscopists who perform more than 100 colonoscopy exams per year (2 per week) will be included in the analysis. We will query the database to identify specific quality indicators and analyze on a per-provider basis. We will determine if the specific field was completed (Aim #1) and if so, we will measure the specific outcome. We will also compare quality indicators among VA sites to non-VA sites overall. More than 600,000 colonoscopies were contributed to CORI between 2004 and 2007. The VA sample (66,000) represents 11% of the reports.
The CORI database received 588,139 colonoscopy reports from 2004-2007. Data was collected from 8 VA sites, representing 11.3% of colonoscopy reports from 2004-2007.
We compared demographics of VA and non-VA sites. VA patients were predominantly male, older and more likely to be minority. Screening, as an indication for colonoscopy was similar (45.6% for VA; 47.4% for non-VA). However, VA sites were more likely to have patients referred for positive FOBT than non-VA sites (10.8% vs 2.6%).
We compared quality indicators from VA and non-VA sites. VA reports were less likely to include bowel prep quality (81.9% vs 87.6%; p < 0.001), and when reported, preps were less adequate at VA (91.7% vs 94.5%; p <0.001). Cecal intubation rates were slightly lower at VA sites (94.6% vs 96.6%).
The key endpoint of colonoscopy screening is detection of polyps >9mm. Rates were higher in VA practices (7.6% vs 6.2%), which is most likely due to higher rates of men and older patients. Rates of interventions for unplanned events were similar (1.2% vs 1.0%).
We then performed a "by-endoscopist" analysis for the VA sites, analyzing data of 58 endoscopists who performed more than 100 colonoscopies during at least one year of the study period. The first analysis were reporting elements which should appear in every report. We set an arbitrary 95% threshold for each element in the computerized report. The proportion of endoscopists who reached this benchmark were: 1) report of prep quality 44.8%; 2) report of ASA class (79.3%); 3) report of prior endoscopy in patients receiving polyp surveillance (15.5%); 4) report of polyp size (82.8%); and 5) report of polyp morphology (29.3%).
Finally we analyzed four specific indicators for each provider.
1)Bowel prep was adequate in more than 90% of exams for 50% of endoscopists
2)Cecal intubation rate was >905 in 52 or 58 endoscopists (89.7%)
3)Polyp retrieval rates were >94% in 44 of 58 endoscopists (75.9%)
4)Unplanned interventions were reported in more than 2% of exams by 14 of 58 endoscopists (24.1%).
This study provides valuable quality assurance data to the VA, and will enable quality improvement programs in colonoscopy. These data highlight important areas for quality improvement. These data can also be used as feedback to individual providers and sites to help identify areas for process improvement. For example, if a site has low rates of adequate bowel prep, they can analyze their bowel prep procedures, and take steps to improve quality.
External Links for this Project
- Lieberman DA, Faigel DO, Logan JR, Mattek N, Holub J, Eisen G, Morris C, Smith R, Nadel M. Assessment of the quality of colonoscopy reports: results from a multicenter consortium. Gastrointestinal endoscopy. 2009 Mar 1; 69(3 Pt 2):645-53. [view]