- Email: [email protected]
792 Assessment of Screening Colonoscopy Performance Measures Across Diverse Practice Settings
Abstracts
792 Assessment of Screening Colonoscopy Performance Measures Across Diverse Practice Settings Jennifer Nayor*1, Jennifer Inra1,2, Margery Rosenblatt3, Muthoka L. Mutinga1,2, Sarathchandra I. Reddy4, Sapna Syngal1,2, Fay Kastrinos5 1 Division of Gastroenterology, Brigham and Women’s Hospital, Boston, MA; 2Harvard Medical School, Boston, MA; 3Division of Population Sciences, Dana-Farber Cancer Institute, Boston, MA; 4Division of Gastroenterology, South Shore Hospital, South Weymouth, MA; 5Division of Digestive and Liver Diseases, Columbia University Medical Center, New York, NY
Figure 1. Endoscopist Polyp Detection Practice Patterns: PDR vs PDR/ PDR2+
791 Timed Screening Colonoscopy: a Randomized Trial Eduardo Coghlan*, Luis Laferrere, Elisa Zenon, Juan M. Marini, German L. Rainero, Angel Nadales Gastroenterology, Hospital Universitario Austral, Buenos Aires, Argentina Background: Missed adenomas are more likely to be in the proximal colon and insufficient detection of these lesions might explain the occurrence of a proportion of interval carcinomas. Studies have demonstrated increased detection of significant neoplastic lesions in colonoscopic examinations where the withdrawal time is 6 minutes or more, but there are no recommendations on how much time to spend in each colonic segment. Objective: To assess the difference in adenoma detection rate (ADR) between two colonoscopic withdrawal timed techniques. Design: Randomized trial. Setting: Tertiary academic center. Patients: Consecutive patients referred for screening colonoscopy, without prior colonoscopies in the last 10 years. Main Outcome measurements: ADRs for patients subjected either to a timed colonoscopy with fixed withdrawal times of a minimum of 2 minutes in the cecum and right colon, 1 minute in the transverse colon and 3 minutes in the left colon, or a standard timed colonoscopy with free withdrawal time of at least 6 minutes. Results: 631 patients were enroled. Five were excluded because of incomplete colonoscopies (3 patients with Boston Bowel Preparation Scores [BBPS] of 0, and 2 patients with fixed angulations). Of the 626 patients included, 322 (51%) were randomized to the group with fixed withdrawal times (Group A) and 49% to conventional withdrawal (Group B). Median age was 57 years (SD 6), 251 patients (40%) were men and the mean withdrawal time was 7:05 minutes (SD 1). 389 adenomas were found in 230 patients, with 15 advanced lesions and 3 adenocarcinomas. Global ADR was 37.6% with no significant statistical differences between the two groups (36,8% vs 36,6%, p 0,96) respectively. On a bivariate analysis (Crude Odds Ratio [cOR]), the finding of adenomas was not related to either withdrawal technique (cOR 0,99 p 0,96) or the BBPS rating (cOR 0,91 p 0,17), but was associated to increasing age (cOR 1,03 p0, 029), male sex (cOR 1,6 p 0,008) and the time spent during withdrawal (cOR 1, p!0,001). A multivariate analysis confirmed these findings with clear relation between the finding of adenomas and age (Adjusted Odds Ratio [aOR] 1,03 p 0, 02), male sex (aOR 1,78 p, 0,002) and time of withdrawal (aOR 1, p !0,001), while no association was seen with either withdrawal technique (aOR 0,81 p 023) or BBPS rating (aOR 0,94 p 0,38). There was no statistical significant difference between the two groups (21,7 % vs. 24,3%, p 0,44) concerning right-sided adenomas. Limitations: Difficulties in assessing the extent of each colonic segment. Conclusion: The variation of withdrawal timing techniques was studied and these results, with an ADR higher than current benchmarks, support the view that longer withdrawal times, even with fixed pre-established time per colonic segment, are associated with better proximal and distal adenoma detection.
www.giejournal.org
Background: Prior studies have not assessed the impact of different endoscopy practice settings on screening colonoscopy quality metrics. Our goal was to compare screening colonoscopy performance measures between different practice settings, determine whether a scorecard intervention to assess performance measures varies by practice setting, and evaluate whether practice setting is associated with adenoma detection. Methods: We prospectively collected data on 1,987 screening colonoscopies in asymptomatic individuals R50 years old performed at a tertiary-care hospital (site 1), a community-based hospital (site 2) and a private practice group (site 3). After 12-weeks of data collection, a scorecard was distributed to each endoscopist with individualized results of adenoma detection rates (ADR), cecal intubation rate and withdrawal time. Data was collected for a second 12-week period to assess the impact of the intervention. The primary outcomes were whether scorecards improved endoscopist performance and variability by practice setting. The secondary outcome was to identify predictors associated with adenoma detection. Statistical tests included Chi Squared, Fisher’s Exact Test and Kruskal-Wallis and generalized linear mixed models to identify factors associated with adenoma detection. Results: Endoscopist and patient characteristics were similar across the sites except for more male patients at site 3 compared to sites 1 and 2 (56.1%, 42.8% and 45.0%, respectively; p!0.0001). Adequate preparation quality was also higher at sites 2 and 3 compared to site 1 (88.3%, 92.1% and 70.6%, respectively; p!0.0001). Baseline performance metrics were high across sites. After the scorecard intervention, there was no significant improvement in overall male ADR (pre 31.9%, post 30.0%; pZ0.89), female ADR (21.6%, 18.4%; pZ0.21), cecal intubation rate (98.2%, 98.5%; pZ0.56) or mean withdrawal time (9.2min, 9.8min; pZ0.79) and these quality indicators did not vary by practice site. When looking at underperforming endoscopists, scorecards did not consistently improve performance. Predictors associated with increased adenoma detection included male patient sex (OR 1.69; 95% CI 1.31-2.18), adequate preparation quality (OR 1.55; 95% CI 1.07-2.23), withdrawal time greater R6 minutes (OR 1.15; 95% CI 1.12-1.17) and increasing patient age (OR 1.03; 95% CI 1.01-1.05). Practice setting was not significantly associated with adenoma detection. Conclusions: Implementation of performance scorecards did not improve colonoscopy quality measures across different practice settings. Practice setting does not influence adenoma detection while male sex, age, adequate bowel preparation and withdrawal time is associated with increased adenoma detection. Interventions to improve the performance of screening colonoscopy can be applied similarly among diverse practice settings.
793 A Novel Metric to Assess Quality and Efficiency in Screening Colonoscopy Mark E. Benson*, Deepak V. Gopal, Michael R. Lucey, Mark Reichelderfer, Patrick R. Pfau University of Wisconsin Medical School, Madison, WI Introduction: Currently accepted minimal standards for quality in screening colonoscopy include; 1) Cecal intubation rates R 95 % 2) Withdrawal times R 6 minutes and 3) Adenoma detection rates R25% for men and R15% women. Longer withdrawal times have been correlated with increased adenoma detection rates (ADR). However, longer withdrawal times may lead to longer overall procedure times, the need for increased sedation, and less overall procedures performed (patients screened). We propose a novel metric as a means to determine an endoscopist’s colonoscopic efficiency. The measurement, adenoma detection rate divided by mean withdrawal time (min.), generates a ratio which is an assessment of procedure efficiency. The aim of this project was to directly assess the validity and applicability of this novel endoscopic quality metric in a large academic tertiary care center. Methods: Quality measures including mean withdrawal times and adenoma detection rates (% patients screened in which adenomatous polyps detected) were prospectively recorded for all consecutive colonoscopies performed by 22 endoscopists comprising gastroenterologists, colorectal surgeons and family practice physicians between April 1, 2013 and Oct 31, 2014. This information was used to calculate the novel quality metric and to compare it to recognized quality standards (cecal intubation rate, withdrawal time, and ADR). Results: 14,805 screening colonoscopies were performed in average risk patients R 50 years old (mean age 56 6, 52 % female) during the study period. The adenoma detection rate (ADR) was 34 % overall for the group (range 15 % to 55 %). The mean withdrawal time was 14 min. (range 9 min. to 23 min.). The ADR/withdrawal time ratio mean was 2.4 (range 1.0 to 3.0). There was a significant positive correlation between the ratio and ADR (R 0.68 P ! 0.01). There was a significant negative correlation between the ratio and cecal intubation time (R -0.54 P Z 0.02). All of the endoscopists with ADRs below the minimally accepted quality
Volume 81, No. 5S : 2015 GASTROINTESTINAL ENDOSCOPY AB171
792 Assessment of Screening Colonoscopy Performance Measures Across Diverse Practice Settings Jennifer Nayor*1, Jennifer Inra1,2, Margery Rosenblatt3, Muthoka L. Mutinga1,2, Sarathchandra I. Reddy4, Sapna Syngal1,2, Fay Kastrinos5 1 Division of Gastroenterology, Brigham and Women’s Hospital, Boston, MA; 2Harvard Medical School, Boston, MA; 3Division of Population Sciences, Dana-Farber Cancer Institute, Boston, MA; 4Division of Gastroenterology, South Shore Hospital, South Weymouth, MA; 5Division of Digestive and Liver Diseases, Columbia University Medical Center, New York, NY
Figure 1. Endoscopist Polyp Detection Practice Patterns: PDR vs PDR/ PDR2+
791 Timed Screening Colonoscopy: a Randomized Trial Eduardo Coghlan*, Luis Laferrere, Elisa Zenon, Juan M. Marini, German L. Rainero, Angel Nadales Gastroenterology, Hospital Universitario Austral, Buenos Aires, Argentina Background: Missed adenomas are more likely to be in the proximal colon and insufficient detection of these lesions might explain the occurrence of a proportion of interval carcinomas. Studies have demonstrated increased detection of significant neoplastic lesions in colonoscopic examinations where the withdrawal time is 6 minutes or more, but there are no recommendations on how much time to spend in each colonic segment. Objective: To assess the difference in adenoma detection rate (ADR) between two colonoscopic withdrawal timed techniques. Design: Randomized trial. Setting: Tertiary academic center. Patients: Consecutive patients referred for screening colonoscopy, without prior colonoscopies in the last 10 years. Main Outcome measurements: ADRs for patients subjected either to a timed colonoscopy with fixed withdrawal times of a minimum of 2 minutes in the cecum and right colon, 1 minute in the transverse colon and 3 minutes in the left colon, or a standard timed colonoscopy with free withdrawal time of at least 6 minutes. Results: 631 patients were enroled. Five were excluded because of incomplete colonoscopies (3 patients with Boston Bowel Preparation Scores [BBPS] of 0, and 2 patients with fixed angulations). Of the 626 patients included, 322 (51%) were randomized to the group with fixed withdrawal times (Group A) and 49% to conventional withdrawal (Group B). Median age was 57 years (SD 6), 251 patients (40%) were men and the mean withdrawal time was 7:05 minutes (SD 1). 389 adenomas were found in 230 patients, with 15 advanced lesions and 3 adenocarcinomas. Global ADR was 37.6% with no significant statistical differences between the two groups (36,8% vs 36,6%, p 0,96) respectively. On a bivariate analysis (Crude Odds Ratio [cOR]), the finding of adenomas was not related to either withdrawal technique (cOR 0,99 p 0,96) or the BBPS rating (cOR 0,91 p 0,17), but was associated to increasing age (cOR 1,03 p0, 029), male sex (cOR 1,6 p 0,008) and the time spent during withdrawal (cOR 1, p!0,001). A multivariate analysis confirmed these findings with clear relation between the finding of adenomas and age (Adjusted Odds Ratio [aOR] 1,03 p 0, 02), male sex (aOR 1,78 p, 0,002) and time of withdrawal (aOR 1, p !0,001), while no association was seen with either withdrawal technique (aOR 0,81 p 023) or BBPS rating (aOR 0,94 p 0,38). There was no statistical significant difference between the two groups (21,7 % vs. 24,3%, p 0,44) concerning right-sided adenomas. Limitations: Difficulties in assessing the extent of each colonic segment. Conclusion: The variation of withdrawal timing techniques was studied and these results, with an ADR higher than current benchmarks, support the view that longer withdrawal times, even with fixed pre-established time per colonic segment, are associated with better proximal and distal adenoma detection.
www.giejournal.org
Background: Prior studies have not assessed the impact of different endoscopy practice settings on screening colonoscopy quality metrics. Our goal was to compare screening colonoscopy performance measures between different practice settings, determine whether a scorecard intervention to assess performance measures varies by practice setting, and evaluate whether practice setting is associated with adenoma detection. Methods: We prospectively collected data on 1,987 screening colonoscopies in asymptomatic individuals R50 years old performed at a tertiary-care hospital (site 1), a community-based hospital (site 2) and a private practice group (site 3). After 12-weeks of data collection, a scorecard was distributed to each endoscopist with individualized results of adenoma detection rates (ADR), cecal intubation rate and withdrawal time. Data was collected for a second 12-week period to assess the impact of the intervention. The primary outcomes were whether scorecards improved endoscopist performance and variability by practice setting. The secondary outcome was to identify predictors associated with adenoma detection. Statistical tests included Chi Squared, Fisher’s Exact Test and Kruskal-Wallis and generalized linear mixed models to identify factors associated with adenoma detection. Results: Endoscopist and patient characteristics were similar across the sites except for more male patients at site 3 compared to sites 1 and 2 (56.1%, 42.8% and 45.0%, respectively; p!0.0001). Adequate preparation quality was also higher at sites 2 and 3 compared to site 1 (88.3%, 92.1% and 70.6%, respectively; p!0.0001). Baseline performance metrics were high across sites. After the scorecard intervention, there was no significant improvement in overall male ADR (pre 31.9%, post 30.0%; pZ0.89), female ADR (21.6%, 18.4%; pZ0.21), cecal intubation rate (98.2%, 98.5%; pZ0.56) or mean withdrawal time (9.2min, 9.8min; pZ0.79) and these quality indicators did not vary by practice site. When looking at underperforming endoscopists, scorecards did not consistently improve performance. Predictors associated with increased adenoma detection included male patient sex (OR 1.69; 95% CI 1.31-2.18), adequate preparation quality (OR 1.55; 95% CI 1.07-2.23), withdrawal time greater R6 minutes (OR 1.15; 95% CI 1.12-1.17) and increasing patient age (OR 1.03; 95% CI 1.01-1.05). Practice setting was not significantly associated with adenoma detection. Conclusions: Implementation of performance scorecards did not improve colonoscopy quality measures across different practice settings. Practice setting does not influence adenoma detection while male sex, age, adequate bowel preparation and withdrawal time is associated with increased adenoma detection. Interventions to improve the performance of screening colonoscopy can be applied similarly among diverse practice settings.
793 A Novel Metric to Assess Quality and Efficiency in Screening Colonoscopy Mark E. Benson*, Deepak V. Gopal, Michael R. Lucey, Mark Reichelderfer, Patrick R. Pfau University of Wisconsin Medical School, Madison, WI Introduction: Currently accepted minimal standards for quality in screening colonoscopy include; 1) Cecal intubation rates R 95 % 2) Withdrawal times R 6 minutes and 3) Adenoma detection rates R25% for men and R15% women. Longer withdrawal times have been correlated with increased adenoma detection rates (ADR). However, longer withdrawal times may lead to longer overall procedure times, the need for increased sedation, and less overall procedures performed (patients screened). We propose a novel metric as a means to determine an endoscopist’s colonoscopic efficiency. The measurement, adenoma detection rate divided by mean withdrawal time (min.), generates a ratio which is an assessment of procedure efficiency. The aim of this project was to directly assess the validity and applicability of this novel endoscopic quality metric in a large academic tertiary care center. Methods: Quality measures including mean withdrawal times and adenoma detection rates (% patients screened in which adenomatous polyps detected) were prospectively recorded for all consecutive colonoscopies performed by 22 endoscopists comprising gastroenterologists, colorectal surgeons and family practice physicians between April 1, 2013 and Oct 31, 2014. This information was used to calculate the novel quality metric and to compare it to recognized quality standards (cecal intubation rate, withdrawal time, and ADR). Results: 14,805 screening colonoscopies were performed in average risk patients R 50 years old (mean age 56 6, 52 % female) during the study period. The adenoma detection rate (ADR) was 34 % overall for the group (range 15 % to 55 %). The mean withdrawal time was 14 min. (range 9 min. to 23 min.). The ADR/withdrawal time ratio mean was 2.4 (range 1.0 to 3.0). There was a significant positive correlation between the ratio and ADR (R 0.68 P ! 0.01). There was a significant negative correlation between the ratio and cecal intubation time (R -0.54 P Z 0.02). All of the endoscopists with ADRs below the minimally accepted quality
Volume 81, No. 5S : 2015 GASTROINTESTINAL ENDOSCOPY AB171