Colonoscopy Quality Measures Research Articles

Performance report cards increase adenoma detection rate.

Background and study aims Adenoma detection rate (ADR) is an important measure of colonoscopy quality, as are polyp, advanced ADR, and adenocarcinoma detection rates. We investigated whether performance report cards improved these outcome measures. Patients and methods Endoscopists were given report cards comparing their detection rates to the institutional mean on an annual basis. Detection rates were evaluated at baseline, 1 year after report cards (Year 1), and 2 years after report cards (Year 2). Endoscopists were unaware of the study and received no other interventions. The primary outcome was ADR and secondary outcomes were polyp detection rate (PDR), advanced ADR, and adenocarcinoma detection rate. Multivariate regression was performed to adjust for temporal trends in patient, endoscopists, and procedural factors. Results Seventeen physicians performed 3,118 screening colonoscopies in patients with positive FOBT or family history of colon cancer. The ADR increased from 34.5 % (baseline) to 39.4 % (Year 1) and 41.2 % (Year 2) ( P = 0.0037). The PDR increased from 45 % (baseline) to 48.8 % (Year 1) and 51.8 % (Year 2) ( P = 0.011). There was no significant improvement in advanced ADR or adenocarcinoma detection rates. On multivariate analysis, the ADR increased by 22 % in Year 1 ( P = 0.03) and 30 % in Year 2 ( P = 0.008). Among physicians with a baseline ADR < 25 %, improvement in ADR was even greater, increasing 2.2 times by the end of the study ( P = 0.004). Improvements in ADR were not correlated with specialty although gastroenterologists were 52 % more likely to find an adenoma than general surgeons. Conclusions Annual performance report cards increased adenoma detection rates, especially among physicians with low ADR < 25 %.

Provider-specific quality measurement for ERCP using natural language processing

Natural language processing (NLP) is an information retrieval technique that has been shown to accurately identify quality measures for colonoscopy. There are no systematic methods by which to track adherence to quality measures for ERCP, the highest risk endoscopic procedure widely used in practice. Our aim was to demonstrate the feasibility of using NLP to measure adherence to ERCP quality indicators across individual providers. ERCPs performed by 6 providers at a single institution from 2006 to 2014 were identified. Quality measures were defined using society guidelines and from expert opinion, and then extracted using a combination of NLP and data mining (eg, ICD9-CM codes). Validation for each quality measure was performed by manual record review. Quality measures were grouped into preprocedure (5), intraprocedure (6), and postprocedure (2). NLP was evaluated using measures of precision and accuracy. A total of 23,674 ERCPs were analyzed (average patient age, 52.9 ± 17.8 years, 14,113 were women [59.6%]). Among 13 quality measures, precision of NLP ranged from 84% to 100% with intraprocedure measures having lower precision (84% for precut sphincterotomy). Accuracy of NLP ranged from 90% to 100% with intraprocedure measures having lower accuracy (90% for pancreatic stent placement). NLP in conjunction with data mining facilitates individualized tracking of ERCP providers for quality metrics without the need for manual medical record review. Incorporation of these tools across multiple centers may permit tracking of ERCP quality measures through national registries.

Sa1062 Colonoscopy Quality Measures of Gi Fellow Colonoscopy Performance by Quarter of Training

Sa1062 Colonoscopy Quality Measures of Gi Fellow Colonoscopy Performance by Quarter of Training

Tu1724 - The Effect of Mid-Year Report Cards on Colonoscopy Quality Measures

Tu1724 - The Effect of Mid-Year Report Cards on Colonoscopy Quality Measures

Quality of Colonoscopy Performed in Rural Practice: Experience From the Clinical Outcomes Research Initiative and the Oregon Rural Practice-Based Research Network.

Colon cancer screening is effective. To complete screening in 80% of individuals over age 50 years by 2018 will require adequate colonoscopy capacity throughout the country, including rural areas, where colonoscopy providers may have less specialized training. Our aim was to study the quality of colonoscopy in rural settings. The Clinical Outcomes Research Initiative (CORI) and the Oregon Rural Practice-based Research Network (ORPRN) collaborated to recruit Oregon rural practices to submit colonoscopy reports to CORI's National Endoscopic Database (NED). Ten ORPRN sites were compared to non-ORPRN rural (n = 11) and nonrural (n = 43) sites between January 2009 and October 2011. Established colonoscopy quality measures were calculated for all sites. No ORPRN physicians were gastroenterologists compared with 82% of nonrural physicians. ORPRN practices reached the cecum in 87.4% of exams compared with 89.3% of rural sites (P = .0002) and 90.9% of nonrural sites (P < .0001). Resected polyps were less likely to be retrieved (84.7% vs 91.6%; P < .0001) and sent to pathology (77.1% vs 91.3%; P < .0001) at ORPRN practices compared to nonrural sites. The overall polyp detection (39.0% vs 40.3%) was similar (P = .217) between ORPRN and nonrural practices. Of exams with polyps, the rate for largest polyp on exam 6-9 mm was 20.8% at ORPRN sites, compared to 26.8% at nonrural sites (P < .0001), and for polyps >9mm 16.6% vs 18.7% (P = .106). ORPRN sites performed well on most colonoscopy quality measures, suggesting that high-quality colonoscopy can be performed in rural settings.

Adenoma and Polyp Detection Rates in Colonoscopy according to Indication.

Background Adenoma detection rate (ADR) is a validated quality measure for screening colonoscopy, but there are little data for other indications. The distribution of adenomas is not well described for these indications. Aim To describe ADR and the adenoma distribution in the proximal and distal colon based on colonoscopy indication. Methods Outpatient colonoscopies are subdivided by indication. PDR and ADR for the entire colon and for proximal and distal colon. Data were compared using generalized estimating equations to adjust for clustering amongst endoscopists while controlling for patient age and gender. Results 3436 colonoscopies were reviewed (51.2%: men (n = 1759)). Indications are screening 49.2%, surveillance 29.3%, change in bowel habit 8.4%, bleeding 5.8%, colitides 3.0%, pain 2.8%, and miscellaneous 1.5%. Overall ADR was 37% proximal ADR 28%, and distal ADR 17%. PDR and ADR were significantly higher in surveillance than in screening (PDR: 69% versus 51%; ADR: 50% versus 33%; p = 0.0001). Adenomas were more often detected in the proximal than in the distal colon, for all indications. Conclusions Prevalence of polyps and adenomas differs based on colonoscopy indication. Adenoma detection is highest in surveillance and more commonly detected in the proximal colon. For quality assurance, distinct ADR and PDR targets may need to be established for different colonoscopy indications.

The Impact of Patient Navigation on the Quality of Colorectal Cancer Screening: Results from the Provider and Community Engagement (PACE) Program in Franklin County, Ohio

Introduction: Colorectal cancer (CRC) screening rates are lowest among racial/ethnic minorities and those of low socioeconomic status (SES). These groups are also at greatest risk for suboptimal bowel prep, one of several colonoscopy quality indicators. Patient navigation (PN) is an evidence-based intervention that was developed as a means to address barriers to cancer screening. In March 2015, we piloted a PN program within our Provider and Community Engagement (PACE) Program to reduce disparities in the quality of colonoscopy-specific colorectal cancer screening. Aim: Determine the impact of PN on colonoscopy quality indicators among racial/ethnic minorities with low SES. Methods: We performed a retrospective chart review of all patients who participated in the pilot PN program. Only those who underwent an average-risk screening colonoscopy (SC) were included in the primary analysis. They were compared to all individuals who underwent a SC by one of the eight endoscopists participating in the PN program within the prior six months. Propensity score matching (based on age, race, ethnicity, gender, preferred language, smoking and diabetes status, and aspirin usage) was used to identify controls from this group. Patients with multiple colonoscopies were excluded. Demographic characteristics and colonoscopy quality indicators (bowel prep score, cecal intubation rate, adenoma detection rate, serrated polyp detection rate, advanced adenoma detection rate and adenoma per colonoscopy) were compared between matched groups by mixed effect models for continuous variables or conditional logistic regression for categorical variables. Insurance status was used as a surrogate for SES and we equated Medicaid and self-pay with low SES. Results: During the study period, 45 out of 51 (88.2%) colonoscopies in the PN group were average risk, compared to 608 out of 710 (85.6%) in the control group (p = 0.61). Following propensity score matching, 90 controls were compared to 45 individuals in the PN group in the primary analysis. Overall, a significantly greater proportion of participants in the PN cohort were self-pay or on Medicaid as compared to the control group (93% vs 48%; p Conclusion: Patients of low SES who received PN had colonoscopy quality measures that did not differ from the control group with higher SES. These results inform clinical practice and support the use of PN for improving the quality of SC among low SES patients, a population at greatest risk for suboptimal bowel prep and low SC uptake.Table 1: Demographic characteristics among average-risk participants who underwent a screening colonoscopyTable 2: Colonoscopy quality indicators among average-risk participants who underwent a screening colonoscopy

Endoscopist Serrated Polyp Detection Rate (SDR) Highly Correlates with Adenoma Detection Rate (ADR)

Introduction: Sessile serrated polyps are important precursors of colon cancer, but are more difficult to detect at colonoscopy. While serrated polyp detection rate (SDR) may be an important measure of colonoscopy quality, an accurate diagnosis of a serrated polyp is dependent upon pathologist expertise. It is unclear whether established quality metrics of withdrawal time (WT) and adenoma detection rate (ADR) sufficiently correlate with SDR. The primary aim is to determine the correlation of SDR with WT and ADR. The secondary aim is to determine the SDR in high ADR endoscopists. Methods: This is a retrospective cohort study of all providers at a single academic medical center performing ≥200 eligible screening colonoscopies over a 26-month period (4/2014-5/2016). An eligible screening colonoscopy was a colonoscopy in a patient age 50-75 without history of neoplastic polyps. SDR and ADR are the proportion of screening colonoscopies in which a serrated polyp (SDR) or adenoma (ADR) was biopsied or removed. WT is the time spent on colonoscope withdrawal in screening colonoscopies in which no biopsy/polypectomy occurred. Endoscopists were stratified into tertiles by ADR (High/Intermediate/Low ADR Endoscopists). All data was abstracted from our institutional data warehouse which integrates pathology, demographics, and endoscopy reports. Results: The 24 endoscopists who met inclusion criteria performed a total of 15,168 screening colonoscopies (median volume 486; range 200-1126). All endoscopists had a mean WT of ≥6 minutes (range 6-20 min). The mean institutional ADR was 39% (range 21-59%) and SDR was 8% (range 0.5-19%). There was a strong positive correlation between WT and both ADR (r=0.72, p < 0.0001) and SDR (r=0.71, p < 0.0001; Fig 1). There was a very strong positive correlation between ADR and SDR (r=0.86, p < 0.0001; Fig 2). The pooled SDR among High ADR endoscopists was 13%, which was significantly higher than Intermediate (SDR 7%, p < 0.01) and Low (SDR 4%, p < 0.001) ADR endoscopists.Figure 1Figure 2Conclusion: Screening colonoscopy SDR varies widely between endoscopists, identifying important variations in screening colonoscopy quality. However, there is a very strong correlation between endoscopist ADR and SDR. As SDR is more dependent on pathologist expertise than ADR, there appears to be no additional benefit in calculating endoscopist SDR to measure screening colonoscopy quality. Improvements in ADR may lead to simultaneous improvements in SDR, though this requires further study.

PTU-030 Number of Significant Polyps Detected Per Six Minutes of Withdrawal Time at Colonoscopy (SP6): A New Measure of Colonoscopy Efficiency and Quality: Abstract PTU-030 Table 1

Introduction There is an increasing focus on quality, safety and efficiency of colonoscopy procedures. Colonoscopic efficiency could be defined as the ability to detect clinically relevant pathologies with a minimum expenditure of time and effort without compromising the safety of the procedure. Adenoma Detection Rate (ADR) is considered as a surrogate marker for quality of colonoscopic examination. However, ADR does not take into account sessile serrated polyps/adenomas (SSP/A) or the efficiency of detection. We therefore propose a new measure of colonoscopy efficiency the SP6 that can be used to evaluate both individual endoscopist’s performance and to compare different detection interventions. Our objective was to assess the SP6 for an individual colonoscopist during standard and Endocuff –assisted colonoscopy (EAC) Methods A prospective service evaluation of screening colonoscopies was performed by an experienced endoscopist between October 2014 and September 2015. For consecutive colonoscopies, patient demographics and procedural data were collected. Results 96 patients had screening colonoscopy during this period. The median age was 65 years (55–74 years). A distal disposable attachment such as Endocuff VisionTM was used at the endoscopists’ discretion. 49 patients had Endocuff Vision-assisted colonoscopy and the remainder had standard colonoscopy. Figure 1 summarises the main findings. There was no significant difference in caecal intubation time and withdrawal time between the two groups. Both ADR and SP6 were significantly improved and SP6 demonstrated that EAC appears to significantly improve colonoscopy efficiency with approximately twice as many pre-cancerous lesions detected and removed per 6 minutes of withdrawal time (1.11 vs 0.6, p = 0.004). Conclusion From this preliminary data an SP6 (colonoscopy efficiency metric) appeared significantly higher when Endocuff VisionTM is used. An SP6 >1 can be achieved in the context of bowel cancer screening FOBT positive colonoscopy and may act as a new benchmark to demonstrate high quality examinations. Disclosure of Interest None Declared

Su1653 Higher Adenoma Detection Rate With Endocuff-Assisted Colonoscopy in a Screening Population

The Adenoma Detection Rate (ADR) is one of the quality measures in screening colonoscopy. The lower the ADR, the higher is the risk of cancers after colonoscopy. Endocuff is an endoscopic cap with plastic projections which permits to flatten the colonic fold during withdrawal. Endocuff-Assisted colonoscopy (EAC) is potentially able to ameliorate the ADR which is crucial especially in a screening population.

Tu1030 Primary Care Physician's Perception of Colonoscopy Quality Measures and Their Influence on Colonoscopy Referral Patterns

Tu1030 Primary Care Physician's Perception of Colonoscopy Quality Measures and Their Influence on Colonoscopy Referral Patterns

The burden of bowel preparations in patients undergoing elective colonoscopy.

An adequate bowel preparation is an important quality measure for optimal colonoscopy. The aim of this article is to study the burden of bowel preparations by examining seven specific variables (hunger, taste, volume, sleep, social, work, and adverse events (AEs)). Ambulatory patients undergoing elective colonoscopy completed a questionnaire regarding their experience with the prescribed preparation. The seven study variables were graded using a numerical scale of 0-10 (best to worst). A score >6 was considered to indicate a significant impact and used as primary outcome. Patients were also asked to grade in descending order what they perceived as the worst aspect of the preparation. A total of 216 patients completed the survey. Preparations consisted of split-dose sodium picosulfate (SPS) (n = 49), split-dose 4 l PEG ± menthol (n = 49), full-dose PEG (n = 68), and 2 l split-dose PEG + ascorbic acid (n = 50). Except for work and AEs, all variables were considered to have a negative impact by >20% of patients (range 20.4-34.2). SPS was superior to PEG regimens in taste (4.1% vs. 35.9%) and volume (0% vs. 44.9%) (p < 0.05 for both) but inferior for hunger (30.6% vs. 19.2%; p = 0.09). The addition of menthol to PEG significantly improved taste (22.4% vs. 41.5%; p = 0.02). Sleep disturbances were most common with SPS and least with split-dose PEG (30.6% vs. 17.4%; p < 0.05). Overall, patients ranked volume, taste, and hunger as most burdensome. The burden of bowel preparation is substantial. An informed personalized choice of preparation may improve adherence, tolerability and colon cleansing.

A Novel Electronic Medical Record-Based Workflow to Measure and Report Colonoscopy Quality Measures.

When I was in private practice, pulling adenoma detection rates (ADRs) by physician was easy. In 2007, we developed a method to calculate ADRs, which was a bit cumbersome but effective, and we began tracking them by physician. We provided every partner with his or her ADR as part of a package of quarterly performance metrics. By 2009, we had incorporated endoscopy and pathology into our electronic medical record (EMR) so ADR analysis became automatic. This allowed us to begin publishing our results on our website and ultimately in peer-reviewed journals (Clin Gastroenterol Hepatol 2009;7:1335–1340 and Gastroenterology 2015;149:952–957). I then moved to an academic medical center where the enterprise-wide EMR (Epic) did not interface with either our endoscopy reporting software or pathology results, and all hope of routine ADR measurement vanished. Everyone working in similar situations can relate to my frustration. In this month’s Road Ahead column, Leiman and colleagues provide a straightforward method to extract ADRs from an Epic EMR. While this still requires physician input, it helps those of us in large health systems measure what is important in our ongoing efforts to reduce the incidence of colon cancer.

Temporal trends and variability of colonoscopy performance in a gastroenterology practice.

Quality measures for colonoscopy are operator dependent and vary. It is unclear whether quality measures change over time. In this study, time-dependent variation in colonoscopy performance was examined in a gastroenterology practice. Colonoscopy and histopathology records that were collected at three hospitals (one university and two non-university hospitals) over three time periods (2007, 2010, and 2013) were reviewed. Data from colonoscopists performing at least 100 procedures per year were analyzed. Inter-colonoscopist variation in performance (i. e. adjusted cecal intubation rate [aCIR], adenoma detection rate [ADR], advanced ADR, mean adenomas per procedure [MAP], proximal ADR, nonpolypoid ADR, and serrated polyp detection rate) were examined using coefficients of variation. Logistic regression analyses were also performed, adjusting for covariates. A total of 23 colonoscopists performing 6400 procedures were included. Overall, the mean aCIR, ADR, MAP, and proximal ADR improved significantly over time, from 91.9 %, 22.5 %, 0.37, and 10.2 % in 2007 to 95.3 %, 25.8 %, 0.45, and 13.4 %, respectively, in 2013 (P < 0.05). The inter-colonoscopist variation in ADR decreased from 37 % in 2007 to 15 % in 2013 (P < 0.05). In the non-university hospitals, mean values for quality measures increased significantly over time, whereas they remained stable in the university hospital. Variability in performance among colonoscopists decreased significantly within the gastroenterology clinical practice. Core quality measures improved over time, mainly through improvement of the lower performers. Measurement of inter-colonoscopist variation in performance helps to identify factors that stimulate or hinder performance, and forms the basis for interventions. http://www.trialregister.nl.

Differences in Colonoscopy Quality Among Facilities: Development of a Post-Colonoscopy Risk-Standardized Rate of Unplanned Hospital Visits

Colonoscopy is a common procedure, yet little is known about variations in colonoscopy quality among outpatient facilities. We developed an outcome measure to profile outpatient facilities by estimating risk-standardized rates of unplanned hospital visits within 7 days of colonoscopy. We used a 20% sample of 2010 Medicare outpatient colonoscopy claims (331,880 colonoscopies performed at 8140 facilities) from patients ≥65 years or older, and developed a patient-level logistic regression model to estimate the risk of unplanned hospital visits (ie, emergency department visits, observation stays, and inpatient admissions) within 7 days of colonoscopy. We then used the patient-level risk model variables and hierarchical logistic regression to estimate facility rates of risk-standardized unplanned hospital visits using data from the Healthcare Cost and Utilization Project (325,811 colonoscopies at 992 facilities), from 4 states containing 100% of colonoscopies per facility. Outpatient colonoscopies were followed by 5412 unplanned hospital visits within 7 days (16.3/1000 colonoscopies). Hemorrhage, abdominal pain, and perforation were the most common causes of unplanned hospital visits. Fifteen variables were independently associated with unplanned hospital visits (c= 0.67). A history of fluid and electrolyte imbalance (odds ratio [OR]= 1.43; 95% confidence interval [CI]: 1.29-1.58), psychiatric disorders (OR= 1.34; 95% CI: 1.22-1.46), and, in the absence of prior arrhythmia, increasing age past 65 years (aged >85 years vs 65-69 years: OR= 1.87; 95% CI: 1.54-2.28) were most strongly associated. The facility risk-standardized unplanned hospital visits calculated using Healthcare Cost and Utilization Project data showed significant variation (median 12.3/1000;5th-95th percentile, 10.5-14.6/1000). Median risk-standardized unplanned hospital visits were comparable between ambulatory surgery centers and hospital outpatient departments (each was 10.2/1000), and ranged from 16.1/1000 in the Northeast to 17.2/1000 in the Midwest. We calculated a risk-adjusted measure of outpatient colonoscopy quality, which shows important variation in quality among outpatient facilities. This measure can make transparent the extent to which patients require follow-up hospital care, help inform patient choices, and assist in quality-improvement efforts.

The impact of exclusion criteria on a physician’s adenoma detection rate

The adenoma detection rate (ADR) is a validated and widely used measure of colonoscopy quality. There is uncertainty in the published literature as to which colonoscopy examinations should be excluded when measuring a physician's ADR. To examine the impact of varying the colonoscopy exclusion criteria on physician ADR. We applied different exclusion criteria used in 30 previous studies to a dataset of endoscopy and pathology reports. Under each exclusion criterion, we calculated physician ADR. A private practice colonoscopy center affiliated with the University of Illinois College of Medicine. Data on 20,040 colonoscopy examinations performed by 11 gastroenterologists from July 2009 to May 2013 and associated pathology notes. ADRs across all colonoscopy examinations, each physician's ADR, and ADR ranking. There were 28 different exclusion criteria used when measuring the ADR. Each study used a different combination of these exclusion criteria. The proportion of all colonoscopy examinations in the dataset excluded under these combinations of exclusion criteria ranged from 0% to 92.2%. The mean ADR across all colonoscopy examinations was 39.1%. The change in mean ADR after applying the 28 exclusion criteria ranged from -5.5 to +3.0 percentage points. However, the exclusion criteria affected each physician's ADR relatively equally, and therefore physicians' rankings via the ADR were stable. ADR assessment was limited to a single private endoscopy center. There is wide variation in the exclusion criteria used when measuring the ADR. Although these exclusion criteria can affect overall ADRs, the relative rankings of physicians by ADR were stable. A consensus definition of which exclusion criteria are applied when measuring ADR is needed.

OC-029 Non-invasive screening for colorectal adenomas using plasma micrornas

Introduction In the UK BSCP, patients aged 60–75 are screened using biennial faecal occult blood testing (FOBt), FOBt+ patients undergo colonoscopy. At screening colonoscopy, yield for adenomas is 46.5%, adenocarcinomas in 6%. 1 Whilst effective, FOBt lacks high sensitivity, specificity and accuracy – half of screening colonoscopies are normal or reveal haemorrhoids/diverticular disease (DD). 1 One off sigmoidoscopy screening (“bowel scope”) is being introduced for patients aged 55 after a large trial showed 23% reduction in CRC incidence. 2 However, endoscopy based screening is invasive, resource intensive and can cause harm. Uptake of FOBt screening is 1 Early reports suggest this is lower for bowel scope. A blood based biomarker screening test is appealing. Method We investigated microRNAs (miRs) – short non-coding RNA molecules as potential biomarkers. 181 FOBt+ patients undergoing BCSP colonoscopy and 29 patients undergoing planned polypectomy were recruited (blood samples taken pre-procedure). 210 patients – 128 male, 82 female. 117 with adenomas, 12 with CRC, 81 controls (normal or non-adenoma/adenocarcinoma). RNA was extracted from plasma and processed. Pooled groups of patients with adenomas and controls were analysed using array cards. MiRs 19a, 98, 146b, 186, 331–5p, 452 and 625 were identified as candidate biomarkers. Cases were analysed for candidate miRs using quantative polymerase chain reaction. Results Candidate miRs showed significant levels of expression in patients with adenomas compared to controls. MiRs 19a, 331–5p, 452 p = Modelling of miR panels were performed, ROC curve analysis showed: Polyps (excl hyperplastic) in males, panel; miRs 98, 186, 452, sensitivity 0.600/specificity 0.872. Adenomas in males, panel; miRs 98, 186, 452, sens 0.606/spec 0.875 or sens 0.591/spec 0.900. Polyps (excl hyperplastic) in pts with DD/haemorrhoids, panel; miRs 186, 452, 331–5p, sens 0.600/spec 0.889 or sens 0.633/spec 0.867. Adenomas in pts with DD/haemorrhoids, panel; miRs 625, 452, 331–5p, sens 0.714/spec 0.864. Conclusion This study suggest plasma microRNAs are potential screening biomarkers for male patients with colorectal polyps/adenomas and patients with adenomas and background DD/ haemorrhoids. Further study is needed to validate these exciting findings. Disclosure of interest None Declared. References Lee TJW, Rutter MD, Blanks RG, et al . Colonoscopy quality measures: experience from the NHS Bowel Cancer Screening Programme. Gut 2012;61:1050–1057 Atkin WS, Edwards R, Kralk-Hans I, et al . Once-only flexible sigmoidoscopy screening in prevention of colorectal cancer: a multicentre randomised controlled trial. Lancet 2010;375:1624–1633

Sa1400 Natural Language Processing for Detection of Quality Measures in Endoscopic Retrograde Cholangiopancreatography

ERCP is one of the riskiest procedures in endoscopy and has the potential for devastating complications. In 2006 the ASGE/ACG Taskforce on Quality in Endoscopy published quality indicators for ERCP; however these indicators required manual review for validation and are not widely in use. Natural language processing (NLP), a technique for discovering information from free text documents, has been shown to identify quality measures in colonoscopy; however, the use of NLP to identify quality indicators related to ERCP has not yet been studied.

Sa1451 Development and Validation of the Saint Paul's Endoscopy Comfort Scale (Specs) for Colonoscopy

Patient comfort during colonoscopy is an important measure of colonoscopy quality and is associated with improved patient satisfaction and compliance with future procedures. The Gloucester Scale (GS) is the most commonly used; however, there are few studies evaluating its validity.

Multi-center colonoscopy quality measurement utilizing natural language processing.

An accurate system for tracking of colonoscopy quality and surveillance intervals could improve the effectiveness and cost-effectiveness of colorectal cancer (CRC) screening and surveillance. The purpose of this study was to create and test such a system across multiple institutions utilizing natural language processing (NLP). From 42,569 colonoscopies with pathology records from 13 centers, we randomly sampled 750 paired reports. We trained (n=250) and tested (n=500) an NLP-based program with 19 measurements that encompass colonoscopy quality measures and surveillance interval determination, using blinded, paired, annotated expert manual review as the reference standard. The remaining 41,819 nonannotated documents were processed through the NLP system without manual review to assess performance consistency. The primary outcome was system accuracy across the 19 measures. A total of 176 (23.5%) documents with 252 (1.8%) discrepant content points resulted from paired annotation. Error rate within the 500 test documents was 31.2% for NLP and 25.4% for the paired annotators (P=0.001). At the content point level within the test set, the error rate was 3.5% for NLP and 1.9% for the paired annotators (P=0.04). When eight vaguely worded documents were removed, 125 of 492 (25.4%) were incorrect by NLP and 104 of 492 (21.1%) by the initial annotator (P=0.07). Rates of pathologic findings calculated from NLP were similar to those calculated by annotation for the majority of measurements. Test set accuracy was 99.6% for CRC, 95% for advanced adenoma, 94.6% for nonadvanced adenoma, 99.8% for advanced sessile serrated polyps, 99.2% for nonadvanced sessile serrated polyps, 96.8% for large hyperplastic polyps, and 96.0% for small hyperplastic polyps. Lesion location showed high accuracy (87.0-99.8%). Accuracy for number of adenomas was 92%. NLP can accurately report adenoma detection rate and the components for determining guideline-adherent colonoscopy surveillance intervals across multiple sites that utilize different methods for reporting colonoscopy findings.