Top 10 POEMs 2011-2012

Top 10 GI POEMs 2014-2015Scott M. Strayer, MD, MPHProfessor Department of Family and Preventive MedicineUniversity of South CarolinaPresident, AAPCE

ObjectivesDescribe and use the principles of Information Mastery, including relevance, validity and usefulness of information.Define patient oriented evidence that matters (POEMs) Know the top 10 GI POEMs for 2013-2014 that apply to primary care endoscopists.Develop an approach for reviewing medical literature that is based on identifying POEMs as they are published.

Otherwise known as decreasing the bed-side stack 3POEMPatient-OrientedEvidencethat Mattersmatters to the clinician, because if valid, will require a change in practiceShaughnessy AF, Slawson DC, Bennett JH. Becoming an Information Master: A Guidebook to the Medical Information Jungle. The Journal of Family Practice 1994;39(5):489-99. 4Answering the relevance question. So what is pt-oriented.see next slide.

Relevance: Type of EvidencePOE: Patient-oriented evidence mortality, morbidity, quality of lifeLonger, better or both

DOE: Disease-oriented evidencepathophysiology, pharmacology, etiology6For example, when Vioxx first came out.pts found better pain relief and it became a $2b/yr drug. However, when large population-based analyses were performed, it was determined that pts were more likely to die from CVD as a result of using this drug. Most pts would care more about that outcome as would physicians, thus necessitating a change in practice. There are many other trials like that annually, albeit less dramatic, but that should constantly be prompting us to examine our practice and standard of care. Determining ValidityLevels of Evidence (LOE):1a, b, c; 2a, b, c; etc., 5- expert opinionA, B, C, DSORT CriteriaTherapy, diagnosis, prognosis, reviews, etc. 74

8Highly Controlled ResearchRandomized Controlled TrialsSystematic ReviewsPhysiologic ResearchPreliminary Clinical ResearchCase reportsObservational studiesUncontrolled Observations&Conjecture

Effect on Patient-Oriented OutcomesSymptomsFunctioningQuality of LifeLifespanEffect on Disease MarkersDiabetesArthritisPeptic UlcerEffect on Risk Factors for DiseaseImprovement in markers (blood pressure, cholesterol)SORTA

Validity of EvidenceRelevance of Outcome SORTB

SORTC

9Another way of putting relevance and validity together is illustrated in this slide, along with integration of the SORT classification for recommendations. It aint what you dont know that gets you into trouble, its what you know for sure that just aint so.Mark Twain

Finding the POEMSReviewed all Essential Evidence Plus (http://www.essentialevidenceplus.com/) daily updates from Nov. 2014-Oct 2015 identifying studies that are relevant to primary care endoscopists. EE+ reviews over 100 journals each month including JAMA, BMJ, Lancet, NEJM, ER, Surgery, Psychiatry, Dermatology, Urology and Family Medicine journalsAlso reviews Cochrane libraryPubmed clinical queries from Nov. 2013-Oct 2014, with keywords: colonoscopy technique, screening colonoscopy.

Used to use several other foraging tools including Physicians First Watch, BMJ Updates, Cochrane Pearls, Dynamed, Medscape, etc.

But, after looking at quality of various foraging tools (show slides from USC lecture) have streamlined my email inbox to include only these

HOW TO GET EE+ in AF?

11How many are using water infusion technique?What are your reasons for using this technique?#1 Water infusion versus air insufflation for colonoscopy. BOTTOM LINEAdenoma detection, assessed with two different measures (that is adenoma detection rate and number of detected adenomas per procedure), was higher using water infusion colonoscopy. In addition, procedure-related abdominal pain was also reduced, which may enhance the acceptance of screening/surveillance colonoscopy.

Cochrane Database Syst Rev. 2015 May 26;5:CD009863. doi: 10.1002/14651858.CD009863.pub2.

To reduce pain, water infusion instead of traditional air insufflation during the insertion phase of the colonoscopy has been proposed, thereby improving patients' acceptance of the procedure. Moreover, the water infusion method may improve early detection of precancerous neoplasms.We searched the Cochrane Colorectal Cancer Group Specialized Register (February 2014), the Cochrane Central Register of Controlled Trials (CENTRAL; 2014, Issue 1), Ovid MEDLINE (1950 to February 2014), Ovid EMBASE (1974 to February 2014), and ClinicalTrials.gov (1999 to February 2014) for eligible randomised controlled trials.

included randomised controlled trials comparing water infusion (water exchange or water immersion methods) against standard air insufflation during the insertion phase of the colonoscopy.Two review authors independently assessed the studies for inclusion and extracted data from eligible studies. We performed analysis using Review Manager software (RevMan 5).We included 16 randomised controlled trials consisting of 2933 colonoscopies. Primary outcome measures were cecal intubation rate and adenoma detection; secondary outcomes were time needed to reach the cecum, pain experienced by participants during the procedure, completion of cecal intubation without sedation/analgesia, and adverse events.

Completeness of colonoscopy, that is cecal intubation rate, was similar between water infusion and standard air insufflation (risk ratio 1.00, 95% confidence interval (CI) 0.97 to 1.03, P = 0.93).Adenoma detection rate, that is number of participants with at least one detected adenoma, was slightly improved with water infusion (risk ratio 1.16, 95% CI 1.04 to 1.30, P = 0.007).Assuming the fraction of patients undergoing screening colonoscopy who had one or more adenomas detected was 20 per 100 with standard colonoscopy, the use of water colonoscopy may increase the fraction to 23 per 100 individuals. From our findings, it is possible that up to 68,000 more of the 1.7 million outpatient screening colonoscopies performed annually in the United States, could detect adenomas if water infusion colonoscopy was used.In addition, with water infusion participants experienced significantly less pain (mean difference in pain score on a 0 to 10 scale: -1.57, 95% CI -2.00 to -1.14, P < 0.00001) and a significantly lower proportion of participants requested on-demand sedation or analgesia, or both (risk ratio 1.20, 95% CI 1.14 to 1.27, P < 0.00001). No increase in adverse events based on qualitative analysis.How many are using cap assisted colonoscopy?What are the reasons for using this device? #2 Cap assisted colonoscopy for detection of serrated polypsBOTTOM LINEIn this post-hoc analysis, Cap-assisted colonoscopy is a safe technique that offers a higher detection rate of significant serrated polyps when compared to standard colonoscopy. If confirmed in future trials, this simple technique has the potential to improve protection against interval colon cancers.

BMC Gastroenterol. 2015 Feb 5;15:11. doi: 10.1186/s12876-015-0234-1.Colonoscopy offers limited protection against right-sided colon cancer, a significant proportion of which arise from the serrated pathway of carcinogenesis.The aim of this study was to compare cap-assisted colonoscopy and standard high-definition white light colonoscopy regarding serrated polyps' detection.Post hoc analysis was performed of a previously conducted randomized controlled trial comparing standard and cap-assisted colonoscopy for adenoma detection.

Post hoc analysis was performed of a previously conducted randomized controlled trial comparing standard and cap-assisted colonoscopy for adenoma detection. Randomization was stratified based on the indication of colonoscopy and all procedures were performed by three experienced endoscopists.Following cecal intubation, the colonic mucosa was carefully inspected during withdrawal of colonoscope and all polyps detected were documented for their size, location, morphology and then removed and sent for histopathology.Detection rates of significant serrated polyps between both arms were compared using the Fisher's exact test and Wilcoxon Rank Sum test.Population: 427 patients were enrolled (7 exclusions, 210 completed study in each arm, mean age of 61 years, 95% male, 75% Caucasian, 67% screening colonoscopies). There were no significant differences in baseline characteristics between both groups. Cap-assisted colonoscopy detected a significantly higher proportion of subjects with significant serrated polyps as well as a higher total number of significant serrated polyps compared to standard colonoscopy (12.8% vs. 6.6%, p =0.047 and 40 vs. 20,p = 0.03 respectively).

Do you allow any food on the day before colonoscopy?

#3 Low-residue diet vs. clear liquids for colonoscopy prep BOTTOM LINEPatients allowed to have a limited low-residue diet before colonoscopy achieve a bowel prep quality that is noninferior to patients on a strict clear liquid diet limitation. Furthermore, polyp detection rates, patient tolerance, and patient acceptance were similar between the 2 groups.

J Clin Gastroenterol. 2014 Nov-Dec;48(10):851-5. doi: 10.1097/MCG.0000000000000167.

To achieve an excellent bowel preparation, it is routine to require a clear liquid diet on the day before the procedure. Unfortunately, this dietary modification may be poorly tolerated. We examine whether a change in precolonoscopy dietary restriction can lead to better patient tolerance without compromising examination quality.This is a prospective, blinded, randomized controlled trial of patients undergoing screening or surveillance colonoscopy.

The primary objective measures the effect of dietary modification on bowel prep quality. Secondary endpoints include polyp detection, patient tolerance, withdrawal time, and patient acceptance.A total of 200 patients were randomized to either (a) a low-residue diet for breakfast and lunch the day before the procedure or (b) clear liquids all day before the procedure. All patients underwent an identical low-volume sodium sulfate split prep.Bowel prep quality was scored using the Boston Bowel Preparation Scale (BBPS).Overall, 96.5% of patients had a good or excellent bowel prep (BBPS=6, 7, 8, or 9). LRD prep quality was noninferior to CLD prep quality (LRD 7.8 vs. CLD 8.1). Polyp detection rates were similar (68% vs. 65.4%, P=0.6899). Patient tolerance and acceptance did not differ. Withdrawal times were equivalent between both groups (16.2 vs. 16.5 min, P=NS).

`#4 Improving colon cancer screening ratesBOTTOM LINEIn a primary care practice, letters plus automated telephone calls are better than either alone in increasing colorectal cancer screening rates among patients who are overdue for screening. These findings suggest the promise of a relatively inexpensive intervention to improve cancer screening

J Am Board Fam Med. 2015 Jan-Feb;28(1):46-54. doi: 10.3122/jabfm.2015.01.140174.

Low-cost interventions to improve cancer screening among primary care patients are needed. The comparative effectiveness of personalized letters, automated telephone calls, and both on breast cancer (BC) and colorectal cancer (CRC) screening is not known.A pragmatic, randomized, controlled trial was conducted in 2011 to 2012.Population: Eligible primary care patients were women ages 50 to 74 years who were past due for mammography and men or women who were past due for mammography or CRC screening of any kind (>12 months since last fecal occult blood test, >5 years since last sigmoidoscopy/double-contrast barium enema, or >10 years since last colonoscopy), respectively.Participants were randomized to 1 of 3 interventions: personalized mailed letters, automated telephone calls, or both. he primary outcome was medical record documentation of a completed mammogram or CRC screening within 36 weeks of randomization.We estimated the costs of each intervention and calculated the marginal cost-effectiveness per person screened.

The crude screening rates for CRC were 17%, 14%, and 24% for the letter, automated call, and combined (letter/automated call) groups, respectively. The combined intervention group had a statistically higher screening rate (P < .05) compared with either of the single intervention groups (letter only or automated call) for CRC in both the crude and adjusted analyses. The combined intervention costs $13.14 per additional person screened for CRC.

#5 IFOBTs moderately sensitive and highly specific for colon cancer

Bottom LineImmunochemical fecal occult blood tests (IFOBTs), such as OC-Micro, OC-Sensor, or OC-Light, are moderately sensitive (73% - 89%) and highly specific (92% - 95%) for identifying colorectal cancer. In comparison, Homoccult Sensa has a lower sensitivity (64% - 80%) and specificity (87% - 90%). IFOBTs also have the advantage of requiring only one sample. (LOE = 1c)

Lee JK, Liles EG, Bent S, Levin TR, Corley DA. Accuracy of fecal immunochemical tests for colorectal cancer: systematic review and meta-analysis. Ann Intern Med 2014;160(3):171-181.

These researchers searched 5 databases and the reference lists of included studies, finding 19 eligible studies that evaluated the diagnostic accuracy of IFOBTs. Two authors used the STARD and PRISMA protocols and independently determined study eligibility, extracted the data, and evaluated study quality. They included cohort studies and randomized studies that used colonoscopy or longitudinal follow-up as the gold standard and only included studies published in English.They excluded studies or results that evaluated only the detection of adenomas. Limiting analysis to only currently available IFOBTs found a sensitivity of 82% (95% CI, 73% - 89%) and a specificity of 94% (92% - 95%). These numbers translate into a positive likelihood ratio of 13.10 and a negative likelihood ratio of 0.19.There was no difference in performance among different commercial products , and multiple sampling was no more accurate than a single sample. Heterogeneity among the studies was acceptable when removing products that are not commercially available. There was some evidence of publication bias. There are no head-to-head studies comparing one type of test with another, and no research evaluating the effectiveness of IFOBT testing on cancer-related mortality or all-cause mortality.

Whats new for IBS?Am J Gastroenterol 2014; 109:S2 S26; doi: 10.1038/ajg.2014.187 #6 Ondensatron effective for diarrhea-predominant IBSOndansetron (Zofran) provides a modest benefit for patients with diarrhea-predominant irritable bowel syndrome (IBS-D). The main benefit is in reducing urgency and is seen within 7 days. (LOE = 1b)

Garsed K, Chernova J, Hastings M, et al. A randomised trial of ondansetron for the treatment of irritable bowel syndrome with diarrhoea. Gut 2014;63(10):1617-1625.

I look forward to a larger, longer trial of ondansetron to be assured of its safety and to assess benefits in a more typical primary care population.42Ondansetron is a widely used anti-emetic, but it also slows colonic transit.This is the first adequately powered trial to investigate the use of the drug in patients with IBS-D. Population: 120 adults, aged 18 to 75 years, with IBS-D meeting standard Rome III criteria from an English specialty clinic. The patients were carefully evaluated for other causes of digestive symptoms; the evaluations included colonoscopy, blood tests for celiac disease and inflammatory bowel disease, and a test for lactose intolerance. This was a crossover trial, with each participant receiving either placebo or ondansetron during two 5-week treatment periods. The trial was well-designed, with appropriate masking and allocation concealment. After the baseline assessment, participants were randomized to receive either ondansetron 4 mg tablets or placebo tablets for a 3-week dose titration period. The patients were told to begin taking 1 capsule once daily, but could increase to a maximum of 2 capsules 3 times daily. 44 If they achieved a dosage that adequately controlled their symptoms, they were to maintain it. All patients took the final titrated dose for the last 2 weeks of the 5-week period. They then had a washout period of at least 2 weeks to get them back to their baseline stool frequency, and then they took the other medication (either placebo or ondansetron).The advantage of crossover trials is that you can get away with smaller sample sizes; the disadvantage is that it is easier for patients to determine whether they were taking active drug or placebo, and sometimes drug effects are not completely washed out. Of the 120 patients who were randomized, 98 were available for the intention-to treat-analysis (82%). The primary outcome was the Bristol Stool Form Score [folks in Bristol must be truly honored], ranging from Type 1 (separate hard lumps, nutlike, hard to pass) to Type 7 (watery, no solid pieces, entirely liquid). During the trial, the patients taking ondansetron typically improved from Type 5 (soft blobs with clear-cut edges, passed easily), to the much more desirable Type 4 (like a sausage, but with cracks on its surface). Gut transit was measured on the last day of each treatment by plain x-ray.Compared with the period when they were taking placebo, participants taking ondansetron had approximately 1 fewer day of urgency or bloating per week (P < .001), lower frequency of defecation (11% fewer stools, 95% CI 4% - 18%), no significant decrease in pain scores or days with pain.The median dosage was 4 mg per day in responders. Constipation was more common in patients ta king ondansetron (9% vs 2%). Long-term safety data for ondansetron are not available; a similar drug, alosetron, was withdrawn from widespread use because it caused severe constipation and ischemic colitis.

#7 Blood tests for excluding IBD in adults with IBSBOTTOM LINECRP and calprotectin of 0.5 or 40, respectively, essentially excludes IBD in patients with IBS symptoms. The addition of CRP and calprotectin to symptom-based criteria may improve the confident diagnosis of IBS.

Am J Gastroenterol. 2015 Mar;110(3):444-54. doi: 10.1038/ajg.2015.6. Epub 2015 Mar 3.Irritable bowel syndrome (IBS) is viewed as a diagnosis of exclusion by most providers. The aim of our study was to perform a systematic review and meta-analysis to evaluate the utility of C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), fecal calprotectin, and fecal lactoferrin to distinguish between patients with IBS and inflammatory bowel disease (IBD) and healthy controls (HCs).

A systematic online database search was performed. Included studies were prospective, adult, diagnostic cohort studies with any of the four tests. The means and s.d. values of biomarker logarithms were estimated based on studies that gave medians and either confidence intervals for the median, interquartile ranges, or ranges. We used a Naive Bayes approach to estimate the probability of being a HC, having IBS, or having IBD based on the biomarker values.

Systematic review identified 1,252 citations. After cross-referencing medical subject headings, detailed evaluation identified 140 potentially relevant journal articles/abstracts for CRP, ESR, calprotectin, and lactoferrin of which 4, 4, 8, and 2 fulfilled our inclusion criteria, respectively.None of the biomarkers reliably distinguished between IBS and healthy controls. At a CRP level of 0.5 or calprotectin level of 40g/g, there was a 1% probability of having IBD. Individual analysis of ESR and lactoferrin had little clinical utility.#8 Spores of nontoxigenic C. difficile can prevent recurrent C. difficile infectionBottom LineThe administration of nontoxigenic Clostridium difficile spores significantly reduces the rate of recurrent C. difficile infection in adults compared with placebo (30% vs 11%; number needed to treat [NNT] = 5.3). (LOE = 1b)

Gerding DN, Meyer T, Lee C, et al. Administration of spores of nontoxigenic Clostridium difficile strain M3 for prevention of recurrent C. difficile infection. A randomized clinical trial. JAMA 2015;313(17):1719-1727.Patients, especially the elderly, with Clostridium difficile infection (CDI) have a recurrence rate of 25% to 30%.Population: adults (N = 173), 18 years or older, who were given a diagnosis of CDI within the previous 28 days as either a primary episode or a first recurrence 8 weeks or sooner after primary CDI.All patients were treated with metronidazole, oral vancomycin, or both for 10 to 21 days and had clinically recovered.Intervention: One to two days after the end of metronidazole/vancomycin therapy, eligible patients randomly received (concealed allocation assignment) 1 of 2 different doses of nontoxigenic C. difficile spores or matched placebo, for a total of 4 treatment groups, for 7 to 14 days.Patients masked to treatment group assignment Self-reported outcomes, including adverse events and recurrent symptoms of CDI using daily diaries. Stool sam ples were also collected at weeks 10, 14, 18, 22, and 26. Complete follow-up occurred for 97% of patients at 26 weeks. Results: Using intention-to-treat analysis, CDI recurrence was 30% in patients receiving placebo compared with 11% in all patients receiving active treatment, regardless of the dose and treatment period (NNT= 5.3; 95% CI 2.9-17.9).Fecal colonization with nontoxigenic C. difficile occurred in 69% of patients receiving active treatment. Adverse events, including diarrhea, abdominal pain, and headache, led to study discontinuation in 7% of patients taking placebo and 3% of those taking active treatment.Note that discontinuation rate is HIGHER among placebo!56#9 High-dose barium enemas prevent recurrent diverticular bleeding

Bottom LineThis small study demonstrates that barium impaction therapy using high-dose barium enemas is safe and effective at reducing the rate of recurrent diverticular bleeding. Note that this study was conducted in Japan, where the rate of rebleeding for patients with diverticulosis is much higher than in Western populations. (LOE = 1b-)

Nagata N, Niikura R, Shimbo T, et al. High-dose barium impaction therapy for the recurrence of colonic diverticular bleeding. Ann Surg 2015;261(2):269-275.

Clinical questionIn children with acute diarrhea, is treatment with the probiotic Saccharomyces boulardii effective in decreasing duration?Bottom lineSaccharomyces boulardii, a probiotic often found in yogurt and supplements (alone or in combination with other beneficial microorganisms), decreases the duration of acute diarrhea in children. (LOE = 1a-)ReferenceFeizizadeh S, Salehi-Abargouei A, Akbari V. Efficacy and safety of Saccharomyces boulardii for acute diarrhea. Pediatrics 2014;134(1):e176-e191. doi: 10.1542/peds.2013-3950Study design: Meta-analysis (other)Funding source: Self-funded or unfundedSetting: Various (meta-analysis)SynopsisThe authors assembled the studies for this meta-analysis by scouring 5 databases, including the Cochrane Library, to identify 22 studies of various quality that evaluated the effect of S. boulardii treatment on children with acute diarrhea. The inclusion criteria were very broad regarding the definition of diarrhea, the definition of resolution, and study design (many studies were not randomized or not conducted in a double-blind manner). The majority of patients were treated as inpatients. Two investigators independently selected studies for inclusion and abstracted the data. They included studies published in any language but did not include research on probiotic combinations (these products are commonly marketed in the United States) or studies that did not specify a dosage.The included studies, for the most part, were of low quality and the results likely overestimate benefit. There was no evidence of publication bias. For most of the studies the daily dose of S. boulardii was 250 mg to 750 mg (109 to 1010 colony-forming units). Results varied widely, but again, quantifying diarrhea is not an exact science. The overall duration of diarrhea was decreased an average of 19 hours as compared with the comparator (95% CI, 226.05 - 213.34; P < .001). The decrease in duration was less in higher quality studies (16 hours), but this result was consistent across these studies. The likelihood of diarrhea on days 2, 3, and 4 was less for patients receiving the probiotic.Allen F. Shaughnessy, PharmD, MMedEdProfessor of Family MedicineTufts UniversityBoston, MA

57A high-dose barium enema is thought to prevent recurrent diverticular bleeding through a physical tamponade of bleeding vessels, as well as by a direct hemostatic effect of the barium itself. Retained barium in colonic diverticula over time has previously been shown to be safe. Population: patients hospitalized with diverticular bleeding who had spontaneous cessation of bleeding were randomized, using concealed allocation, to receive either barium impaction therapy (n = 27) or conservative treatment (n = 27).In the barium impaction therapy group, barium sulfate was administered by gastroenterologists via an enema bag at a concentration of 200 g barium per 100 mL tap water for a total volume of 400 mL.X-ray imaging confirmed filling of multiple colonic diverticula with barium and the patient was asked to rotate positions to ensure filling of all diverticula.Baseline characteristics were similar in the 2 groups: the majority of patients were male, the average age was 70 years, and half had a prior history of diverticular bleeding.The severity of initial bleeding was also similar, as measured by number of units of blood transfused prior to randomization and the number of days until spontaneous cessation of bleeding. For the primary outcome of recurrence of bleeding at the 1-year follow-up, the barium group fared better than the conservative treatment group (15% vs 43%; P = .04; NNT=4).After adjusting for factors associated with recurrent bleeding, including hypertension, nonsteroidal anti-inflammatory drug use, and chronic renal failure, the risk of bleeding was decreased in the barium group (hazard ratio = 0.34, 95% CI 0.12-0.97). #10 Risk of GI bleeding with novel anticoagulantsBOTTOM LINEThe risk of gastrointestinal bleeding related to novel oral anticoagulants was similar to that for warfarin. Caution should be used when prescribing novel oral anticoagulants to older people, particularly those over 75 years of age.

BMJ. 2015 Apr 24;350:h1857. doi: 0.1136/bmj.h1857.Clinical questionIs radiofrequency ablation superior to endoscopic surveillance in reducing morbidity or mortality in adults with Barrett esophagus and low-grade dysplasia?Bottom lineUnfortunately, this study was terminated before the important clinical questions about mortality and morbidity could be answered. Radiofrequency ablation reduced the risk of neoplastic progression in adults with Barrett esophagus and low-grade dysplasia (compared with endoscopic surveillance). But it also caused significantly more adverse events, including the need for dilation for esophageal stricture in nearly 12% of patients. I understand the concerns of monitoring committees, but we truly need overall patient-oriented evidence to most accurately guide our clinical decisions. If a monitoring committee had stopped the clinical trial of intravenous lidocaine in the setting of acute myocardial infarction because of a concern for life-threatening cardiac arrhythmias in the placebo group, we would never know that lidocaine actually increased mortality by 10%. The optimal management of Barrett esophagus with low-grade dysplasia will remain uncertain. (LOE = 1b)ReferencePhoa KN, van Vilsteren FG, Weusten BL, et al. Radiofrequency ablation vs endoscopic surveillance for patients with Barrett esophagus and low-grade dysplasia. A randomized clinical trial. JAMA 2014;311(12):1209-1217.Study design: Randomized controlled trial (single-blinded)Funding source: IndustryAllocation: ConcealedSetting: Outpatient (specialty)SynopsisThe optimal management of Barrett esophagus with low-grade dysplasia is uncertain. These investigators identified 136 eligible patients (mean age = 63 years) who had previously undergone upper endoscopy and biopsy within the previous 18 months that confirmed low-grade dysplasia. Eligible patients randomly received (by concealed allocation) either endoscopic radiofrequency ablation or surveillance only. The ablation group received sessions every 3 months until complete endoscopic and histological eradication of Barrett esophagus or a maximum of 2 circumferential and 3 focal sessions. Subsequent endoscopy and 4-quadrant surveillance biopsies occurred annually for a total of 3 years. Patients in the surveillance-only group received endoscopy and 4-quadrant biopsies at 6 months and 12 months initially, and then annually for a total of 3 years. Individuals masked to group assignment assessed histologic and clinical outcomes. Complete follow-up occurred for all patients at 3 years. Using intention-to-treat analysis, patients in the ablation group were significantly less likely to progress to high-grade dysplasia or adenocarcinoma, but also significantly more likely to develop esophageal stricture requiring dilation than patients in the control group (11.8% vs 0.0%). The trial was terminated by the monitoring committee before any additional patient-oriented outcomes were evaluated. The progression to high-grade dysplasia/adenocarcinoma was noted to be directly related to the presence of circumferential Barrett esophagus, age, time since diagnosis of Barrett esophagus, and dysplasia noted on multiple endoscopies. No predictors of spontaneous resolution of the low-grade dysplasia in the control group could be discerned to help guide the choice of ablation therapy over surveillance in patients.David Slawson, MDDirector of Information SciencesUniversity of Virginia Health SystemCharlottesville, VA

61To determine the real world risk of gastrointestinal bleeding associated with the use of the novel oral anticoagulants dabigatran and rivaroxaban compared with warfarin.Retrospective, propensity matched cohort study. New users of dabigatran, rivaroxaban, and warfarin from 1 November 2010 to 30 September 2013.Incidence rates (events/100 patient years) and propensity score matched Cox proportional hazards models were used to estimate rates of total gastrointestinal bleeds, upper gastrointestinal bleeds, and lower gastrointestinal bleeds for the novel oral anticoagulants compared with warfarin in patients with and without atrial fibrillation.

Heterogeneity of treatment effect related to age was examined using a marginal effects model.The incidence of gastrointestinal bleeding associated with dabigatran was 2.29 (95% confidence interval 1.88 to 2.79) per 100 patient years and that associated with warfarin was 2.87 (2.41 to 3.41) per 100 patient years in patients with atrial fibrillation. In non-atrial fibrillation patients, the incidence of gastrointestinal bleeding was 4.10 (2.47 to 6.80) per 100 patient years with dabigatran and 3.71 (2.16 to 6.40) per 100 patient years with warfarin.

In propensity score matched models, the risk of gastrointestinal bleeding with novel oral anticoagulants was similar to that with warfarin in atrial fibrillation patients (dabigatran v warfarin, hazard ratio 0.79 (0.61 to 1.03); rivaroxaban v warfarin, 0.93 (0.69 to 1.25)) And, in non-AF patients (dabigatran v warfarin, hazard ratio 1.14 (0.54 to 2.39); rivaroxaban v warfarin, 0.89 (0.60 to 1.32)). The risk of gastrointestinal bleeding increased after age 65, such that by age 76 the risk exceeded that with warfarin among atrial fibrillation patients taking dabigatran (hazard ratio 2.49 (1.61 to 3.83)) and patients with and without atrial fibrillation taking rivaroxaban (2.91 (1.65 to 4.81) and 4.58 (2.40 to 8.72), respectively).