Su1276Water Exchange May Be Superior to Water Immersion -Minimizes Colonoscopy Discomfort and Increases AdenomaDetection Rate (ADR) - Review of Recent RandomizedControlled Trials (RCT)Felix W. Leung*1,2, Arnaldo Amato3, Christian Ell4, Shai Friedland5,6,Judith O. Harker1, Yu-Hsi Hsieh7, Joseph W. Leung8,9,Surinder K. Mann9, Silvia Paggi3, Juergen Pohl4, Franco Radaelli3,Francisco C. Ramirez10, Rodelei M. Siao-Salera8, Vittorio Terruzzi31Gastroenterology, VA Sepulveda Ambulatory Care Center, North Hills,CA; 2Medicine, UCLA, Los Angeles, CA; 3Gastroenterology, ValduceHospital, Como, Italy; 4Gastroenterology, Medical School of theUniversity of Mainz, Wiesbaden, Germany; 5Gastroenterology, PaloAlto VAMC, Palo Alto, CA; 6Gastroenterology, Stanford University, PaloAlto, CA; 7Gastroenterology, Buddhist Dalin Tzu Chi General Hospital,Chia-Yi, Taiwan; 8Gastroenterology, Sacramento VAMC, VANCHCS,Mather, CA; 9Gastroenterology, UC Davis, Sacramento, CA;10Gastroenterology, Mayo Clinic, Scottsdale, AZIntroduction: Water-related methods for colonoscopy are distinguished by thetiming of removal of infused water, predominantly during withdrawal (waterimmersion), or during insertion (water exchange). Water immersion was firstdescribed in the US literature by Falchuk et al. (NEJM 1984;310:598). Subsequentvariations, commonly known as “air-water hybrids”, have been widely used asadjuncts to usual air insufflation. The water method with water exchange as thesole modality used in conjunction with air exclusion (air pump turned off andresidual colonic air removed by suction) to aid insertion was reviewed recentlyby Leung et al. (Endoscopy 2011;43:816). Objective: To determine the impact ofthese approaches on colonoscopy pain and adenoma detection rate (ADR).Design: Systematic review. Setting: RCT that compared water-related methodsand air insufflation during colonoscope insertion. Patients: Colonoscopy patients.Interventions: Medline, PubMed and Google searches (2008 to 2011) andpersonal communications of manuscripts published or in press were consideredto identify appropriate RCT. Main Outcome: Pain during colonoscopy and ADR.Measurements: RCT were grouped according to whether water immersion orwater exchange was used. Pain scores and ADR were tabulated based on groupassignment. Results: Compared with traditional air insufflation pain duringcolonoscopy is significantly reduced by both water immersion (n�9 RCT) andwater exchange (n�3 RCT). Reduction of pain scores is of a higher order ofmagnitude with water exchange (55%) compared to water immersion (27%)(Table 1). A mixed pattern of increases and decreases in ADR with a significant(p�0.023) net reduction of 7% is observed with water immersion (n�6 RCT). Aconsistently higher ADR is obtained when water exchange (n�3 RCT) isimplemented. A significant (p�0.0261) net increase of 9.7% in overall ADR, 11%in proximal overall ADR (p�0.0072), and 12% in proximal �10 mm ADR(p�0.0043) is observed (Table 2). Limitations: Heterogeneity of reports restrictsapplication of meta-analysis. Conclusions: Both water immersion and waterexchange significantly reduce colonoscopy pain compared with air insufflation.Water exchange may be superior to water immersion in minimizing colonoscopydiscomfort and in increasing both overall and proximal ADR. [Supported by VAand ACG research funds (FWL)].

Table 1. Colonoscopy pain scores during insertion

Air Water ImmersionReduction

(Average�27%) pN Pain score N Pain score

9 RCT 114 5.3 (2.7) 112 4.1 (2.7) �1.2 (23%) 0.001�

170 4.6 [1.8-9.2]a 170 2.9 [1.0-5.8]a �1.7a (37%) 0.001‡39 2.6 (2.2)b 41 2.4 (2.2)b �0.2b (8%) 0.89489 3.4 (2.8) 90 2.5 (2.5) �0.9 (26%) 0.021*51 4.4 (2.6) 51 3.0 (2.2) �1.4 (32%) 0.004*

51 3.3 (2.4) �1.1 (25%) 0.028*114 3.9 [1.4-5.4]a 116 2.8 [1.2-4.4]a �1.1a (28%) 0.001‡31 5.5 31 3.6 �1.9 (35%) �0.05*56 4.2 (2.3) 48 2.8 (1.9) �1.3 (31%) 0.02

Air Water ExchangeReduction

(Average�55%) PN Pain score N Pain score

3 RCT 28 4.1 (3.4) 28 1.3 (1.8) �2.8 (68%) 0.0002*40 6 [—] 42 3 [—] �3 (50%) 0.002*50 4.9 (2.0) 50 2.3 (1.7) �2.6 (53%) 0.0012*

[—], not reported; [IQR], interquantile range; median score [IQR]; mean score (SD);aConverted from 0 to 100 scale to 0 to 10 scale;bConverted from 0 to 5 scale to 0 to 10 scale;*t test;‡rank sum.

Table 2. Combined ADR

Air Water Immersion Difference Pa

(6 RCT) Overall 159/457 (35%) 127/458 (28%) � 7% .023

AirWater

Exchange

(3 RCT) Overall 115/281 (40.9%) 136/269 (50.6%) 9.7% .0261Proximal overall 82/281 (29%) 108/269 (40%) 11% .0072Proximal �10 mm 72/281 (25%) 100/269 (37%) 12% .0043aFisher’s exact test

Su1277Balanced Propofol Sedation vs. Propofol Monosedation inTherapeutic ERCP and EUS: A Prospective Randomized, Double-Blind StudyTae Hoon Lee*1, Chang Kyun Lee2, Sang-Heum Park1, Suck-Ho Lee1,IL Kwun Chung1, Hyun Jong Choi3, Jong Ho Moon3,Young Deok Cho3, Sun-Joo Kim1

1Division of Gastroenterology, Department of Internal Medicine,Soonchunhyang University, College of Medicine, Cheonan, Republic ofKorea; 2Internal Medicine, Kyung Hee University School of Medicine,Seoul, Republic of Korea; 3Division of Gastroenterology, Department ofInternal Medicine, Soonchunhyang University, College of Medicine,Seoul, Republic of KoreaBackgrounds: Propofol sedation demonstrated higher sedation efficacy andshorter patient recovery during ERCP than the traditionally usedbenzodiazepines. However, despite the adequate safety data,nonanesthesiologist-administered propofol sedation for endoscopy remains ahighly controversial issue. Balanced propofol sedation (BPS) can reduce the doseof propofol required, representing an effective alternative to treatment withpropofol alone and leading to fewer propofol-related cardiopulmonarycomplications. However, no studies have compared BPS and propofol alonetitrated to moderate levels of sedation under supervision in therapeutic ERCP orEUS. Objective: To compare the sedation efficacy, safety, and procedureoutcomes of BPS and propofol monosedation by trained registered nurses undersupervision in therapeutic ERCP and interventional EUS. Materials and Methods:A total of 218 consecutive patients who were scheduled for therapeutic ERCP orEUS were recruited at single tertiary-care referral center as a prospective,randomized, double-blind study from April to September, 2011. BPS (singleinduction doses of midazolam 0.05 mg/kg and fentanyl 25-50 �g, followed by 10to 20 mg propofol) or propofol monosedation (bolus injection of propofol 0.5mg/kg, followed by 10 to 20 mg propofol) titrated to a moderate level ofsedation based on the ASA levels was performed under careful monitoring. Allsedatives and analgesics used were administered by trained, registered sedationnurses who had advanced cardiac life support certification under endoscopists’supervision. The primary objective was to compare sedation efficacy. Secondaryoutcomes included sedation safety, procedure outcomes, and complications.Results: A total of 218 patients were randomized into the BPS (n � 109) andpropofol alone (P) (n � 109) groups. Seven patients in the BPS group and fivein the P group were excluded from final analysis. There were no significantdifferences in sedation efficacy, with the exception of recovery timemeanrecovery time (standard deviation) was 18.37 (7.86) min for BPS and 13.4 (6.24)min for P group, respectively (P � .001) (Table). In a safety analysis,cardiopulmonary complication rates were 7.8% (8/102) and 9.6% (10/104),respectively (P � .652). No patient required assisted ventilation or permanenttermination of a procedure in either group. Procedure outcomes and procedure-related complications did not differ depending on sedation protocols. Technicalsuccess rates were 96.3% and 97.2%, respectively (P � .701).Limitations: Single-center study. Conclusions: Propofol monosedation by trained registered sedationnurses provided a rapid recovery time without differences in sedation efficacy,quality, and safety. Procedure-related adverse events and outcomes between BPSand propofol monosedation did not differ.

Table. Sedation efficacy parameters with visual analog scale (VAS)

BPS group Propofol group P value

Induction time, min, mean (SD)(range)*

3.4 (1.78) (1-13) 3.83 (1.96) (1-11) .098

Recovery time† 18.37 (7.86) (5-52) 13.4 (6.24) (5-45) � .001Patient cooperation byendoscopists‡, mean (SD), VAS

7.91 (2.05) 7.65 (2.18) .384

Overall satisfaction withsedation by endoscopist‡

7.96 (1.84) 7.80 (1.81) .525

Overall satisfaction withprocedures by endoscopist‡

8.14 (1.48) 8.03 (1.64) .620

Abstracts

www.giejournal.org Volume 75, No. 4S : 2012 GASTROINTESTINAL ENDOSCOPY AB276

BPS group Propofol group P value

Difficulty of maintenance byendoscopist§

2.57 (2.15) 2.90 (2.32) .284

Patient cooperation by nurse‡,mean (SD), VAS

8.15 (2.20) 7.84 (2.23) .316

Overall satisfaction withsedation by nurse‡

8.25 (1.82) 7.87 (2.06) .164

Difficulty of maintenance bynurse§

3.25 (2.61) 3.57 (2.69) .385

Patient satisfaction withsedation‡, mean (SD), VAS

9.13 (1.16) 8.90 (1.69) .270

Remembered endoscopicinsertion, n (%)

12 (11.8) 18 (17.3) .259

Remembered endoscopicwithdrawal, n (%)

12 (11.8) 13 (12.5) .878

Felt pain during procedure, n(%)

8 (7.8) 17 (16.3) .062

*Time from sedation start to procedure start.†Time from completion of ERCP or EUS to reaching Aldrete score of 10.‡0 � poor; 10 � excellent.§0 � easy; 10 � very difficult.

Su1278Efficacy of Endoluminal Gastroplication in Japanese PatientsWith Proton Pump Inhibitor-Resistant, Non-Erosive EsophagitisKunio Kasugai*, Kentaro Tokudome, Yasushi Funaki, Yasuhiro Tamura,Shinya Izawa, Akihito Iida, Mari Mizuno, Naotaka Ogasawara,Makoto SasakiGastroenterology, Aichi Medical University, Nagakute, JapanBackground and Aims: Non-erosive reflux disease (NERD) accounts for morethan half of cases of gastro-esophageal reflux disease. Its pathology is complex,and the disease is often not amenable to treatment with proton pump inhibitor(PPI), which are used as front-line drugs. Though surgical therapy has beenreported to be effective as an alternative to drug therapy in NERD, the efficacy ofendoscopic treatment in PPI-refractory NERD has not yet been studied. The aimof this study was to evaluate efficacy, safety, and long-term outcomes ofendoluminal gastroplication (ELGP) in patients with PPI-resistant NERD.Methods: We evaluated GERD symptoms, QOL, esophageal pH, symptom index(SI), symptom sensitivity index (SSI), number of plications and PPI medicationbefore and after ELGP in patients with PPI-resistant NERD. Results: The meanfrequency scale for symptoms of the GERD (FSSG) score decreased significantlyfrom 19.1�10.5 before ELGP to 10.3�7.4 after 3 months and to 9.3�9.9 after 12months. There were no changes in the number of reflux episodes or SI for refluxevents, but the number of symptom events and SSI decreased significantly frombefore ELGP to 3 months after. After 12 months, 3 patients (16.7%) were able toreduce the amount of PPI medication by 50% or more, and 12 patients (66.7%)were able to discontinue PPI medication altogether. One or more plicationsremained in 80% and 43% after 3 and 12 months, respectively. There were noserious complications. Conclusions: ELGP was safe, resulted in significantimprovement in subjective symptoms, and allowed less medication to be usedover the long term, thus demonstrating that it could become an option for thetreatment of PPI-refractory NERD.

Su1279Pre and Post Endoscopic Retrograde CholangiopancreatographyAnesthesia Time in Patients Receiving Monitored AnesthesiaCare Versus General AnesthesiaVernon J. Carriere*1, Jason Conway1, Girish Mishra1, Raymond Roy2,John A. Evans1

1Gastroenterology, Wake Forest University Medical Center, WinstonSalem, NC; 2Anesthesia, Wake Forest University Medical Center,Winston Salem, NCPurpose: Many factors determine whether patients will receive monitoredanesthesia care (MAC) versus general anesthesia (GA) for an EndoscopicRetrograde Cholangiopancreatography (ERCP). Minimizing delays both pre andpost ERCP will maximize throughput, thus allowing for increased procedurevolumes.Aim: To determine the contribution of anesthesia to total length ofprocedure time. Methods: All patients undergoing an ERCP from 1/1/11 to 3/31/11 were included. Patients were stratified by indications, type of anesthesia (GAversus MAC) and American Society of Anesthesiologist (ASA) classification. Thefollowing times were collected: room in, endoscope insertion, endoscoperemoval and room out. The time difference between room in and the endoscopeinsertion was considered the “pre-procedure time”. The time difference between

the endoscope removal and room out was considered the “post-procedure time”.Results: A total of 90 patients were included in the study with 39 MAC cases and51 GA cases. The mean age was 54.3 years and 58/90 (64.4%) were female.There was no significant difference between the GA and MAC groups in regardsto ASA classification (p�0.3) with a majority of patients being class II 34/90(37.8%) or class III 51/90 (56.7%). No statistical significance was found betweenthe groups with regard to indication (p�0.8), sex (p�0.62) or age (p�0.2). Themean pre procedure time for the GA group was 20.8 minutes and 11.1 minutesfor the MAC group (p �0.001). The mean post procedure time was 12.9 minutesfor the GA group and 6.3 minutes for the MAC group (p �0.001.) The time ofthe procedure was 33.8 minutes for the GA group and 23.2 minutes for the MACgroup (p � 0.03). The total time was 67.6 minutes for GA and 40.6 minutes forMAC (p �0.001). There were no major complications reported in any of theprocedures. Conclusion: The pre and post procedure time, as well as the totalprocedure time, was significantly shorter for the MAC anesthesia group. In theGA group, the actual procedure time accounted for only 50% of the total time ofthe procedure. Procedure length and complexity should precipitate discussionbetween the gastroenterologist and anesthesiologist regarding MAC versus GA.The merits of MAC anesthesia appear implicit for maximizing efficiency whilemaintaining patient safety when performing ERCP.

Su1280A Novel Procedure for Gastrocutaneous Fistula ClosureOmer J. Deen*1, Keely R. Parisian1, Campbell Harris2,Donald F. Kirby1,2

1Digestive Disease Institute, Cleveland Clinic Foundation, Cleveland,OH; 2Gastroenterology, Medical College of Virginia, Richmond, VABackground: PEG tubes have allowed for a safe and efficient way to feedpatients who cannot tolerate oral feeding, yet have a functioning gastrointestinaltract. Gastrocutaneous fistulas after PEG removal are an unusual and rarecomplication in adults and may be in part due to poor tissue healing, delayedgastric emptying or increased gastric acid production. Various approaches havebeen reported to treat PEG-related gastric fistulashowever, their success rate isvariable and patients frequently require repeat procedures or more than onetechnique in combination, including acid suppression therapy, silver nitrateablation of the PEG tract lining, argon plasma coagulation, fibrin glue, and/orendoclipping. Upon our review, there have been no published case seriesreporting the use of endoscopic banding to close persistent gastrocutaneousfistulas after PEG removal. Study Design: Four patients with persistentgastrocutaneous fistulas after PEG removal were taken for EGD with banding ofthe fistula site. This procedure was chosen due to its relative ease of application,low likelihood of needing repeat procedures, and the ability to seal off thegastric opening immediately upon application. Patient follow-up was bytelephone within 3 days of having the procedure and then again 1-2 weeksafterwards, to ensure that there was no persistent leakage through the fistulatract. Results: Of the four patients who had persistent gastrocutaneous fistulasafter PEG removal, endoscopic banding resulted in complete closure of thefistula in 3 of our 4 patients. In one case, banding was unsuccessful secondary toscarring from prior radiation treatment as well as having a previous PEG tubeplaced 1 inch from the current fistula site. In this case, a second PEG tube wasplaced through the original PEG stoma, leading to cessation of the gastric leak.The first case resulted in no recurrence after 3 years. The second and third caseshave shown no recurrence after 3 months. The fourth case resulted in a secondPEG tube to manage persistent drainage through the tract after unsuccessfulbanding of the site due to complex endoscopic and anatomical issues.Conclusion: Endoscopic closure of a gastrocutaneous fistula, regardless oftechnique used, can help avoid surgical intervention. Anatomical changes fromany previous treatment modalities may decrease the success rate of fistulabanding. However, in most patients, endoscopic banding is a safe and relativelysimple alternative in closing persistent gastrocutaneous fistulas due to prior PEGtubes.

Su1281ERCP Quality Assessment and Outcomes in a Tertiary ReferralCenterVivian E. Ekkelenkamp*1, Arjun D. Koch1, Jelle Haringsma2,Robert A. De Man1, Ernst J. Kuipers1

1Department of Gastroenterology and Hepatology, Erasmus MC,Rotterdam, Netherlands; 2Department of Gastroenterology andHepatology, Maasstad Ziekenhuis, Rotterdam, NetherlandsIntroduction: The ASGE Committee on Outcomes Research recommendedmonitoring 9 ERCP-specific quality indicators for optimal quality assurance inERCP. There are however little data available on these indicators and measuresfor quality are sparse. With the development of a self-assessment tool for ERCP(Rotterdam Assessment Form for ERCP - RAF-E), important key parameters (i.e.appropriateness of indication, assessment of procedural difficulty, cannulationrate, and success rates of sphincterotomy, common bile duct stone extraction

Abstracts

AB277 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 4S : 2012 www.giejournal.org