Improving Screening Colonoscopy Quality
Although screening colonoscopy is effective in reducing the risk of CRC, its effectiveness depends on the quality of procedure performance. There is increasing support in the literature for the notion that variability in colonoscopy quality, specifically in the detection of adenomatous polyps (represented by ADR), is a major contributor to interval CRC. Patients who undergo colonoscopy by endoscopists with a greater ADR are less likely to develop an interval or fatal interval CRC compared with those who receive a colonoscopy by endoscopists with a lower ADR. Consequently, measuring, reporting, and improving ADR is a prime focus of colonoscopy quality improvement; however, studies demonstrating the efficacy of interventions in improving endoscopist ADRs are scarce.
In this study, we report the cumulative positive effects of report cards and mandating SOP on ADR. First, we noted a significant increase in ADR with initial distribution of a report card. Thus, self-awareness of performance relative to peers may significantly drive performance. Data supporting the use of report cards are strong at the institutional level for surgical specialties, but individual report cards may not be informative in some procedure areas owing to a lower case volume per individual. Furthermore, mandating practice standards was associated with significantly improved individual and institutional ADRs, resulting in all endoscopists in this study achieving minimum ADR and WT benchmarks. The observed additive effect of SOP suggests an additional impact of incentivizing endoscopists who may have varying knowledge of, or commitment to, the importance of ADR. The combination of SOP in addition to report cards was associated with a greater positive improvement in ADR, translating to a predicted greater reduction in interval and fatal interval CRC, compared with baseline and that of reporting alone.
We used an institutional electronic data warehouse to generate ADRs. Multiple methods are used in practice to determine ADR. The methods include manual chart review, central data registries, and data warehouses with or without the use of natural language processing. For larger practices, manual chart review is laborious and thus impractical for continuous quality monitoring. Although national data registries can be helpful to compare provider performance nationally and to automate a portion of the process by direct transmission of the endoscopy report to a central data registry, they still require manually identifying whether an adenoma was present during screening colonoscopy, which may be cumbersome with a large number of providers. The use of a data warehouse, in which discrete electronic health record systems are integrated into a central data repository, is particularly effective in calculating ADR, as it links the pathology report to the endoscopy report to facilitate ADR calculation.
Once ADRs are calculated, it is possible to target quality improvement efforts. Multiple methods have been trialed with variable results. In a single center, an endoscopy quality improvement education program that focused on techniques to improve adenoma detection resulted in a durable improvement in ADRs. In a small study of six practitioners at a Veterans Affairs hospital, a quarterly report card was associated with improved provider ADR. The single provider with an ADR below 20% in that study also improved significantly following receipt of the report cards, similar to our findings. Some experts have advocated for the use of ADR public reporting as a mechanism for quality improvement. However, it is unclear whether public reporting has any significant effect on physician or patient behavior in other disciplines. Thus, interventions that significantly improve outcome measures, such as ADR, are of great interest.
Previous studies have not shown a significant improvement when focusing solely on minimal WT. In a study of 42 endoscopists at an academic medical center, an institution mandated minimum WT of 7 min (process measure) was not associated with an improvement in the outcome of polyp-detection rates. In our study, we found no significant increase in WT after both interventions, although all providers achieved at least a 5-min WT following implementation of SOP. This further stresses the importance of focusing on ADR as an outcome measure rather than WT as a process measure.
There are important limitations of this study. Although we have shown short-term improvements in ADRs, we have not yet demonstrated the durability of these interventions, and this requires further study. It is unclear whether additional external factors (such as the accumulating literature highlighting the importance of ADR) would have resulted in a temporal trend regardless of interventions. In addition, there may have been a "carry over" effect of the report card to Period 3, which we have attributed to implementing SOP. In addition, these interventions took place at a single academic medical center with 20 endoscopists, and it is unclear whether these findings are generalizable to other practice settings. Finally, recent guidelines have proposed a minimum ADR of 25%, and further study should explore whether these interventions can achieve that minimum ADR.
In summary, our data suggest that distributing colonoscopy quality report cards resulted in a significant improvement in adenoma-detection rate. Further, we report evidence that implementing a universal standard of practice significantly improved ADR beyond report card distribution and resulted in all endoscopists meeting minimum benchmarks. This suggests that report cards and SOP may have an additive effect in improving colonoscopy quality, and their implementation in endoscopy labs should be encouraged.
Discussion
Although screening colonoscopy is effective in reducing the risk of CRC, its effectiveness depends on the quality of procedure performance. There is increasing support in the literature for the notion that variability in colonoscopy quality, specifically in the detection of adenomatous polyps (represented by ADR), is a major contributor to interval CRC. Patients who undergo colonoscopy by endoscopists with a greater ADR are less likely to develop an interval or fatal interval CRC compared with those who receive a colonoscopy by endoscopists with a lower ADR. Consequently, measuring, reporting, and improving ADR is a prime focus of colonoscopy quality improvement; however, studies demonstrating the efficacy of interventions in improving endoscopist ADRs are scarce.
In this study, we report the cumulative positive effects of report cards and mandating SOP on ADR. First, we noted a significant increase in ADR with initial distribution of a report card. Thus, self-awareness of performance relative to peers may significantly drive performance. Data supporting the use of report cards are strong at the institutional level for surgical specialties, but individual report cards may not be informative in some procedure areas owing to a lower case volume per individual. Furthermore, mandating practice standards was associated with significantly improved individual and institutional ADRs, resulting in all endoscopists in this study achieving minimum ADR and WT benchmarks. The observed additive effect of SOP suggests an additional impact of incentivizing endoscopists who may have varying knowledge of, or commitment to, the importance of ADR. The combination of SOP in addition to report cards was associated with a greater positive improvement in ADR, translating to a predicted greater reduction in interval and fatal interval CRC, compared with baseline and that of reporting alone.
We used an institutional electronic data warehouse to generate ADRs. Multiple methods are used in practice to determine ADR. The methods include manual chart review, central data registries, and data warehouses with or without the use of natural language processing. For larger practices, manual chart review is laborious and thus impractical for continuous quality monitoring. Although national data registries can be helpful to compare provider performance nationally and to automate a portion of the process by direct transmission of the endoscopy report to a central data registry, they still require manually identifying whether an adenoma was present during screening colonoscopy, which may be cumbersome with a large number of providers. The use of a data warehouse, in which discrete electronic health record systems are integrated into a central data repository, is particularly effective in calculating ADR, as it links the pathology report to the endoscopy report to facilitate ADR calculation.
Once ADRs are calculated, it is possible to target quality improvement efforts. Multiple methods have been trialed with variable results. In a single center, an endoscopy quality improvement education program that focused on techniques to improve adenoma detection resulted in a durable improvement in ADRs. In a small study of six practitioners at a Veterans Affairs hospital, a quarterly report card was associated with improved provider ADR. The single provider with an ADR below 20% in that study also improved significantly following receipt of the report cards, similar to our findings. Some experts have advocated for the use of ADR public reporting as a mechanism for quality improvement. However, it is unclear whether public reporting has any significant effect on physician or patient behavior in other disciplines. Thus, interventions that significantly improve outcome measures, such as ADR, are of great interest.
Previous studies have not shown a significant improvement when focusing solely on minimal WT. In a study of 42 endoscopists at an academic medical center, an institution mandated minimum WT of 7 min (process measure) was not associated with an improvement in the outcome of polyp-detection rates. In our study, we found no significant increase in WT after both interventions, although all providers achieved at least a 5-min WT following implementation of SOP. This further stresses the importance of focusing on ADR as an outcome measure rather than WT as a process measure.
There are important limitations of this study. Although we have shown short-term improvements in ADRs, we have not yet demonstrated the durability of these interventions, and this requires further study. It is unclear whether additional external factors (such as the accumulating literature highlighting the importance of ADR) would have resulted in a temporal trend regardless of interventions. In addition, there may have been a "carry over" effect of the report card to Period 3, which we have attributed to implementing SOP. In addition, these interventions took place at a single academic medical center with 20 endoscopists, and it is unclear whether these findings are generalizable to other practice settings. Finally, recent guidelines have proposed a minimum ADR of 25%, and further study should explore whether these interventions can achieve that minimum ADR.
In summary, our data suggest that distributing colonoscopy quality report cards resulted in a significant improvement in adenoma-detection rate. Further, we report evidence that implementing a universal standard of practice significantly improved ADR beyond report card distribution and resulted in all endoscopists meeting minimum benchmarks. This suggests that report cards and SOP may have an additive effect in improving colonoscopy quality, and their implementation in endoscopy labs should be encouraged.
SHARE
