Evaluating learning curves and competence in colorectal EMR among advanced endoscopy fellows: a pilot multicenter prospective trial using cumulative sum analysis.

BACKGROUND AND AIMS: Data on colorectal EMR (C-EMR) training are lacking. We aimed to evaluate C-EMR training among advanced endoscopy fellows (AEFs) by using a standardized assessment tool (STAT). METHODS: This multicenter prospective study used a STAT to grade AEF training in C-EMR during their 12-month fellowship. Cumulative sum analysis was used to establish learning curves and competence for cognitive and technical components of C-EMR and overall performance. Sensitivity analysis was performed by varying failure rates. AEFs completed a self-assessment questionnaire to assess their comfort level with performing C-EMR at the completion of their fellowship. RESULTS: Six AEFs (189 C-EMRs; mean per AEF, 31.5 ± 18.5) were included. Mean polyp size was 24.3 ± 12.6 mm, and mean procedure time was 22.6 ± 16.1 minutes. Learning curve analyses revealed that less than 50% of AEFs achieved competence for key cognitive and technical C-EMR endpoints. All 6 AEFs reported feeling comfortable performing C-EMR independently at the end of their training, although only 2 of them achieved competence in their overall performance. The minimum threshold to achieve competence in these 2 AEFs was 25 C-EMRs. CONCLUSIONS: A relatively low proportion of AEFs achieved competence on key cognitive and technical aspects of C-EMR during their 12-month fellowship. The relatively low number of C-EMRs performed by AEFs may be insufficient to achieve competence, in spite of their self-reported readiness for independent practice. These pilot data serve as an initial framework for competence threshold, and suggest the need for validated tools for formal C-EMR training assessment.

Factors Associated With Health Care Utilization of Recurrent Clostridium difficile Infection in New York State.

BACKGROUND: The incidence of infection due to Clostridium difficile infection (CDI) and subsequent economic burden are substantial. GOALS: The impact of changing practice patterns on demographics at risk and utilization of health care resources for recurrence of CDI remains unclear. STUDY: A total of 291,163 patients hospitalized for CDI were identified from 1995 to 2014 from the New York SPARCS database. The χ test, the Welch t test, and multivariable logistic regression analysis were performed to evaluate factors related to readmission. RESULTS: Hospital admissions and readmissions for CDI peaked in 2008 at 20,487 and 13,795, respectively, and have since decreased (linear trend, 0.9706 and 0.9464, respectively; P<0.0001). In total, 60,077 (21%) patients required ≥2 admissions. Risk factors for readmission included: age 55 to 74, government insurance, hypertension, diabetes, anemia, hypothyroidism, chronic pulmonary disease, rheumatoid arthritis, renal failure, peripheral vascular disease, and depression (all P<0.05). Trends in surgery showed a similar peak in 2008 at 165 and have since decreased (linear trend, 0.8660; P<0.0001). A total of 1830 (0.63%) patients with CDI underwent surgery, with emergent being more common than elective (71% vs. 29%). CONCLUSIONS: Hospital admissions and readmissions for CDI peaked in 2008 and have since been steadily declining. These trends may be secondary to improved diagnostic capabilities and evolving antibiotic regimens. More than 1 in 5 hospitalized patients had at least 1 readmission. Numerous risk factors for these patients have been identified. Although <1% of all patients with CDI undergo surgery, these rates have also been declining.

Endoscopic resection outcomes and predictors of failed en bloc endoscopic mucosal resection of colorectal polyps ≤ 20 mm among advanced endoscopy trainees.

Background and study aims En bloc endoscopic mucosal resection (EMR) is preferred over piecemeal resection for polyps ≤ 20 mm. Data on colorectal EMR training are limited. We aimed to evaluate the en bloc EMR rate of polyps ≤ 20 mm among advanced endoscopy trainees and to identify predictors of failed en bloc EMR. Methods This was a multicenter prospective study evaluating trainee performance in EMR during advanced endoscopy fellowship. A logistic regression model was used to identify the number of procedures and lesion cut-off size associated with an en bloc EMR rate of ≥ 80 %. Multivariate analysis was performed to identify predictors of failed en bloc EMR. Results Six trainees from six centers performed 189 colorectal EMRs, of which 104 (55 %) were for polyps ≤ 20 mm. Of these, 57.7 % (60/104) were resected en bloc. Trainees with ≥ 30 EMRs (OR 6.80; 95 % CI: 2.80-16.50; P  = 0.00001) and lesions ≤ 17 mm (OR 4.56;95 CI:1.23-16.88; P  = 0.02) were more likely to be associated with an en bloc EMR rate of ≥ 80 %. Independent predictors of failed en bloc EMR on multivariate analysis included: larger polyp size (OR:6.83;95 % CI:2.55-18.4; P  = 0.0001), right colon location (OR:7.15; 95 % CI:1.31-38.9; P  = 0.02), increased procedural difficulty (OR 2.99; 95 % CI:1.13-7.91; P  = 0.03), and having performed < 30 EMRs (OR: 4.87; 95 %CI: 1.05-22.61; P  = 0.04). Conclusions In this pilot study, we demonstrated that a relatively low proportion of trainees achieved en bloc EMR for polyps ≤ 20 mm and identified procedure volume and lesion size thresholds for successful en bloc EMR and independent predictors for failed en bloc resection. These preliminary results support the need for future efforts to define EMR procedure competence thresholds during training.