Natural Language Processing Accurately Identifies Dysphagia Indications for Esophagogastroduodenoscopy Procedures in a Large US Integrated Healthcare System: Implications for Classifying Overuse and Quality Measurement.

Recent evidence suggests almost half of repeat esophagogastroduodenoscopy procedures (EGDs) are overused; this prior research relied on administrative data that are often inaccurate. Our primary objective was to determine and compare the accuracy of natural language processing and administrative data to manual chart review to identify dysphagia indications for EGD procedures within the national VA healthcare system. From 396,856 EGD notes identified from 2008-2014, we classified 119,920 as "index" procedures in 2010-2012. We compared the performance of our NLP to ICD codes to correctly identify dysphagia indications in the index EGD procedures and in repeat EGD procedures. We used linked pathology data to describe esophageal biopsies performed during these EGDs. NLP performed significantly better and identified significantly more index and repeat EGD procedures with dysphagia indications than ICD codes, which has critical implications for determining appropriateness of EGD procedures.

Clinical outcomes in patients with bile leaks treated via ERCP with regard to the timing of ERCP: a large multicenter study.

BACKGROUND AND AIMS: Postsurgical or traumatic bile duct leaks (BDLs) can be safely and effectively managed by endoscopic therapy via ERCP. The early diagnosis of BDL is important because unrecognized leaks can lead to serious adverse events (AEs). Our aim was to evaluate the relationship between timing of endotherapy after BDL and the clinical outcomes, AEs, and long-term results of endoscopic therapy. METHODS: We conducted a multicenter, retrospective study on patients with BDLs who underwent ERCP between 2006 and 2014. Data were assembled on patient demographics, etiology of BDL, and procedural details. Endotherapy for BDLs were classified a priori into 3 groups based on timing of ERCP from time of biliary injury: within 1 day of BDL, on day 2 or 3 after BDL, and greater than 3 days after BDL. The relationship among timing of ERCP after BDL injury and outcomes, procedure-related AEs, and patient AEs and mortality were evaluated. RESULTS: From February 2006 to June 2014, 518 patients (50% male; mean age, 51.7 years) underwent ERCP for therapy of BDLs. The etiology of the BDL was laparoscopic cholecystectomy (70.7%), post-liver transplantation (11.2%), liver resection (14.1%), trauma (2.5%), and other causes (1.5%). Endotherapy was performed by placing a transpapillary stent alone (73.5%) or with a sphincterotomy (26.5%). The timing of ERCPs was as follows: ≤1 day = 57 patients, day 2 or 3 = 140 patients, and >3 days = 321 patients. There was no statistical difference in patient demographics, etiology/site of BDL, or type of endotherapy performed among the 3 groups. On multivariate analysis there was no statistically significant difference in BDL success rate for ERCPs performed within 1 day compared with those performed on day 2 or 3 or after 3 days of bile duct injury (91.2%, 90%, and 88.5%, respectively; P = .77). Similarly, there was no significant difference in the overall patient AE rate among the 3 groups (21.1%, 22.9%, and 24.6%, respectively; P = .81). AEs in men occurred significantly more frequently when compared with women, even after adjusting for age, BDL etiology, and location of leak (27.6% vs 19.9%; OR, 1.53; P = .04). Patients whose BDL was due to a cholecystectomy had a lower AE and mortality rate compared with those who had biliary injury from other etiologies (OR, .42; P < .001). CONCLUSIONS: The overall success rates and AEs after ERCP were not dependent on the timing of the procedure relative to the discovery of the bile leak. This suggests that ERCP in these patients can usually be performed in an elective, rather than an urgent, manner.

Safety Analysis of Bariatric Patients Undergoing Outpatient Upper Endoscopy with Non-Anesthesia Administered Propofol Sedation.

BACKGROUND: Non-anesthesia administered propofol (NAAP) has been shown to be a safe and effective method of sedation for patients undergoing gastrointestinal endoscopy. Bariatric surgery patients are potentially at a higher risk for sedation-related complications due to co-morbidities including obstructive sleep apnea. The outcomes of NAAP in bariatric patients have not been previously reported. METHODS: In this retrospective cohort study, severely obese patients undergoing pre-surgical outpatient esophagogastroduodenoscopy (EGD) were compared to non-obese control patients (BMI ≤ 25 kg/m2) undergoing diagnostic EGD at our institution from March 2011-September 2015 using our endoscopy database. Patients' demographics and procedural and recovery data, including any airway interventions, were statistically analyzed. RESULTS: We included 130 consecutive pre-operative bariatric surgical patients with average BMI 45.8 kg/m2 (range 34-80) and 265 control patients with average BMI 21.9 kg/m2 (range 14-25). The severely obese group had a higher prevalence of sleep apnea (62 vs 8%; p < 0.001), experienced more oxygen desaturations (22 vs 7%; p < 0.001), and received more chin lift maneuvers (20 vs 6%; p < 0.001). Advanced airway interventions were rarely required in either group and were not more frequent in the bariatric group. CONCLUSIONS: With appropriate training of endoscopy personnel, NAAP is a safe method of sedation in severely obese patients undergoing outpatient upper endoscopy.