S1934 To BRAVO or Not? Wireless pH Monitoring Versus Conventional Nasopharyngeal pH Catheter Monitoring in the Evaluation of Gastroesophageal Reflux Disease in a Multiracial Asian Cohort

a reflux episode - from the Table below this would be 3/4 or 75%; (2) the symptom sensitivity index (SSI) is the percentage of intervals with a reflux episode that are immediately followed by an interval with a symptom - from the Table below this would be 3/26 or 12%; (3) the symptom-association probability (SAP) is 1 minus the p value calculated using Fisher's Exact Test. From the Table below this value is 99.99%. This value is the probability of obtaining a difference in the proportions smaller than that obtained if the null hypothesis is true. The SI and SSI fail to consider the role of chance association of symptoms and reflux episodes, whereas the SAP calculates the probability assuming the data result from chance alone. Aim. A method is needed to determine the probability of a variety of different relationships, including chance alone, between symptoms and reflux episodes given the data. Bayes' rule provides such a method. Methods. Bayes' rule (Posterior Probability = Prior Probability x Likelihood / Total Probability) was used to analyze the data in the Table below using a “flat” prior probability that considers all possible relationships to be equally likely. Results. Using Bayes' rule, the posterior probability that the proportion of heartburn intervals with reflux is greater than the proportion without reflux is 0.998; that the proportion of heartburn intervals without reflux is greater than the proportion with reflux is 0; and of the null hypothesis (chance alone) is 0.002. Furthermore, the posterior probability is 0.90 that the proportion of heartburn intervals that are immediately preceded by a reflux episode is not greater than 0.20. That is, the probability is 0.90 that at least 80% of intervals in which heartburn occurs are not immediately preceded by a reflux episode. Conclusion. Thus, Bayes' rule can be used to calculate the probability of a relationship between gastroesophageal reflux episodes and symptoms and can provide quantitative insight that is not provided by currently existing methods. Relationship between Symptoms and Esophageal Reflux Episodes.

S1930 IBS-Symptoms in Patients with GERD Helmut Neumann, Klaus Mönkemüller, Ulrike von Arnim, Peter Malfertheiner Introduction: Gastroesophageal reflux disease (GERD) and irritable bowel syndrome (IBS) are highly prevalent in the general population. Hence, it is expected that these conditions frequently overlap. Due to its functional origin, it is believed that IBS occurs more commonly in NERD. Objective: We aimed to evaluate the presence of IBS-symptoms in patients with erosive (ERD), non-erosive reflux disease (NERD) and Barrett's esophagus (BE). Patients and methods: 164 patients were prospectively included in this study. According to the Montreal classification 85 patients had NERD (female n=53; male n=32; mean age 62 years), 44 ERD (female n=18; male n=26; mean age 62 years) and 35 BE (female n=12; male n=23; mean age 66.5 years). All patients with ERD and NERD had typical reflux-symptoms, as assessed by a validated GERD-questionnaire (RDQ). The diagnosis of BE was based on endoscopic and histological criteria according to the guidelines of the American College of Gastroenterology. The diagnosis of IBS was assessed according to the current Rome III criteria. Results: IBS was a frequent finding and not significantly different in prevalence among NERD, ERD and BE. IBS symptoms (bloating, abdominal pain, constipation and diarrhea) in different groups were shown in the enclosed table. NERD patients descripted significantly more often abdominal pain compared to patients with BE (p =0.02). Although patients with NERD and ERD had more bloating, constipation and diarrhea than patients with BE, none of these differences reached statistical significance. Conclusion: Our data show that IBS-symptoms are very common in the entire spectrum of GERD. A careful evaluation of IBS related symptoms should be part of the clinical assessment in patients with GERD.

S1933 Detection of Predictive Factors and Assessment of Therapeutic Effects for Gastro-Esophageal Reflux Diseases (GERD) Using a Novel System FSSG (Frequency Scale for Symptoms of GERD) Kyoko Ito, Takashi Wada, Hisao Tajiri, Mikio Zeniya Backgrounds & Aims: The prevalence of gastro-esophageal reflux diseases (GERD) is increasing worldwide. We investigated the prevalence of GERD in the population and factors predicting the appearance of GERD using FSSG (frequency scale for symptoms of GERD) which is a novel questionnaire for the diagnosis of GERD. The therapeutic effects of rabeprazole were assessed by FSSG. Materials & Methods: FSSG was performed on 1335 subjects (175 female and 1160 male) who attended the medical checkup for one year from December 2006. Subjects under treatments of GERD were excluded. FSSG is a simple questionnaire consisting of 12 specific questions (7 for reflux symptom and 5 for dyspepsia; full score is 48), a score higher than 7 was defined as GERD. Sensitivity, specificity and concordance rate of FSSG were assessed by endoscopical or upper GI series findings as a standard in 295 or 896 subjects, respectively. Age, gender, obesity, metabolic syndrome, smoking, alcohol intake, impaired glucose metabolism, cardiac diseases, liver dysfunctions and hyperlipidemia were simultaneously assessed for predictive factors of GERD. 56 subjects diagnosed as GERD were re-analyzed by FSSG after 4-weeks treatment with rabeprazole 20mg. Results: 230 (26%) subjects were categorized as GERD by FSSG. Sensitivity, specificity and concordance rate of FSSG was 34.7%, 72.7% and 63.4%, or 24.0%, 72.5% and 51.6% in the assessment with endoscopical or upper GI series findings, respectively. The prevalence of GERD was significantly higher in young age under 45 y.o. (31.5 ± 9.6%), excessive alcohol intake more than 300g/wk ethanol (31.9 ± 9.7%) or liver dysfunction (AST> 33, ALT> 35 or GGT> 65 for male, 27 for female) (30.3 ± 9.4%) than control groups (23.5 ± 5.4%, 24.0 ± 3.9% and 23.9 ± 4.6%, respectively) (p<0.05). Impaired glucose metabolism and cardiac diseases had lower prevalence of GERD (21.8 ± 5.3% and 17.3 ± 7.4%, respectively) than control groups (29.6 ± 7.7% and 26.3 ± 3.0%, respectively) (p< 0.05). Young age (OR 1.6, 95% CI 1.2-2.2), excessive alcohol intake (OR 1.5, 95% CI 1.1-2.0) and liver dysfunction (OR 1.3, 95%CI 1.0-1.7) were suggested possible predictive factors for GERD. FSSG scores were significantly decreased after treatments (4.4 ± 0.8 score) compared with before treatments (11.5 ± 0.8 score) (p< 0.01). Conclusions: Quarter of people was found to suffer from symptoms of GERD without attending hospital. Young age, excessive alcohol intake and liver dysfunction were suggested to predict the appearance of GERD. Low sensitivity level of FSSG suggested a considerable prevalence of asymptomatic GERD. FSSG was sufficient to diagnose and assess the therapeutic effects for GERD.

S1931 Yield of Combined Impedance-pH Monitoring in Patients with Refractory Reflux Symptoms George Karamanolis, Zacharias Tsiamoulos, Asimina Gaglia, Theodora Kalli, Dimitrios Polymeros, Konstantinos Triantafyllou, Theodoros Liakakos, Spiros D. Ladas In patients with gastroesophageal reflux disease (GERD), persistent symptoms on PPI therapy may be due to residual acid or non-acid reflux. Combined impedance-pH (MII-pH) technology has been suggested to be superior to pH-monitoring alone in the management of these refractory to PPI patients. However, the utility of implementation of this technique in routine clinical practice is still unknown. Aim: To prospectively investigate the outcomes of patients studied with combined MII-pH monitoring in our unit and to evaluate the yield of additional MII monitoring over pH alone. Methods: From July 2007-Septemebr 2008, 55 consecutive outpatients (32 men; mean age: 48yrs) underwent MII-pH studies on PPI therapy (35% full and 65% double dose) for persistent typical (70%) or atypical (30%) symptoms. Results: Pathological pH monitoring on PPI was found in 12 (22%) patients; in 16% of those on double dose vs. 33% on full dose. In patients with normal pH monitoring, combined MIIpH monitoring showed pathological reflux in 21 (49%) patients (weakly acidic in 22% and non-acidic in 78%). Adding MII increased the rate of abnormal results over pH monitoring alone from 33% to 48% on full dose and from 16% to 33% on double dose PPI. Conclusions: The rate of abnormal pH monitoring in patients with refractory GERD symptoms while on PPI is relatively low especially in those on double-dose. The use of combined MII-pH substantially increased the diagnostic yield of pH-monitoring alone: pathological reflux could be found in about half of patients without reflux on pH-monitoring.

S1934 To BRAVO or Not? Wireless pH Monitoring Versus Conventional Nasopharyngeal pH Catheter Monitoring in the Evaluation of Gastroesophageal Reflux Disease in a Multiracial Asian Cohort Daphne Ang, Tiing Leong Ang, Kieron B. Lim, Teo Eng Kiong, Kwong Ming Fock Background: Traditional catheter-based oesophageal pH monitoring is limited by patient discomfort and inconvenience. The wireless Bravo pH system is a validated technique which overcomes these problems,but its diagnostic accuracy is uncertain with limited data on its use in Asians. Aims:1.Determine the efficacy of Bravo monitoring in patients referred for evaluation of gastroesophageal reflux disease(GERD)in a multiracial Asian cohort. 2.Compare the prevalence of abnormal acid exposure and symptom marker based positive symptom index (SI)or symptom association probability(SAP) in subjects who underwent wireless and conventional pH catheter. Methods: The records of consecutive patients with typical or atypical GERD symptoms who underwent pH monitoring between January 2004 and October 2008 were reviewed. The Bravo capsule was positioned 6 cm above the endoscopically localised squamocolumnar junction. Manometry was performed to identify the high pressure

S1932 A Bayesian Method for Analyzing Relationships Between Esophageal Reflux Episodes and Symptoms Jerry D. Gardner Introduction. To analyze the temporal relationship between symptoms and gastroesophageal reflux, the 24-hour esophageal pH recording is typically divided into intervals of equal time, such as 2 minutes. Currently there are 3 methods for analyzing the data: (1) the symptom index (SI) is the percentage of intervals with a symptom that are immediately preceded by

A-295

AGA Abstracts

AGA Abstracts

patients where the prevalence of GERD was 2-3 times lower than that of the Western population. Methods. One hundred patients with dyspepsia by Rome II criteria and age > 18 years who had been scheduled for endoscopy were prospectively recruited from May 2007 to January 2008. After obtaining informed consents, all patients completed the translated Thai version of CDQ (validated by back-translation from Thai to English) before undergoing endoscopic examination of upper gastrointestinal tract. Those with negative endoscopic examination and score of CDQ equal to or greater than 4 were scheduled to have 24-hour pH monitoring done. Result. There were 68 female and 32 male patients with a mean age of 45.6 + 12.4 years and a range of 25-78 years. There were 44 patients with score of CDQ > 4 and 3 of these had endoscopic LA grade B esophagitis. None of the patients with CDQ score less than 4 had endoscopic esophagitis. The remaining 41 patients with CDQ score > 4 had normal endoscopy and underwent 24- hour pH monitoring and only 8 of these 41 patients had positive 24-hour pH test (defined as total acid exposure time >5% in 24 hours). The diagnosis of GERD was confirmed by endoscopy or pH monitoring in 11 patients out of 44 patients (25%) with CDQ suggesting GERD. Conclusion. Symptoms-base diagnosis by structural Carlsson-Dent questionnaire in Thai population with low prevalence of GERD over diagnosed GERD as compared with endoscopy and 24 hr pH monitoring. Symptomsbase approach may not be appropriate in this setting.

AGA Abstracts

zone of the lower oesophageal sphincter before pH catheter insertion. The prevalence of abnormal total acid exposure time(AET)and positive symptom correlation using SI and SAP were compared using chi-square and Student t-test. Results:55 patients(20 Male,43.6 ± 13.9 years)and 47 patients(20Male, 47.7 ± 13.9 years)underwent wireless and pH catheter respectively. The 2 groups did not differ significantly in age, racial distribution and study indications. All patients completed the wireless study,with a mean recording time of 45.5 ± 2.7 hours. Four patients(8.5%)failed completion of pH catheter recording due to intolerance. Abnormal AET(% total time pH<4 for >4.2%)using the wireless and pH catheter recording occurred in 20 of 55 patients(36%)and 16 of 47 patients(34%)(pNS)respectively. A positive reflux-symptom association despite normal AET occurred in 7(12.7%) and 9(19.1%)patients who underwent wireless and pH catheter respectively(pNS). Two of 7(29%)subjects recorded a positive symptom association only on day 2 of the Bravo study. Conclusion: The wireless capsule was well tolerated in our multiracial Asian cohort.The prevalence of abnormal acid exposure and positive reflux-symptom association using both modalities was similar. Our findings suggest that the wireless device is a safe and useful alternative investigative modality in evaluating GERD patients without compromising diagnostic accuracy. Table 1. Summary of study indications for oesophageal pH monitoring

S1935 Failed BRAVO pH Monitoring: An Autoimmune Phenomenon? Lauren B. Gerson Background: Bravo pH probes provide 2-day esophageal pH monitoring. However, failure of probe attachment and/or early dislodgement can occur. The mechanism behind individual study failure remains unclear. Methods: We performed a retrospective study of Bravo studies performed between 2005-2008. Patients with failed or incomplete studies were invited to undergo evaluation for possible rheumatologic disorders with testing of ESR, ANA, CRP, antismooth muscle Ab, SSA/SSB Ab, and anti-RNP. The control group were patients with successful Bravo studies. Results: 229 patients received Bravo pH probes including 137 (60%) via conscious sedation (CS), and 92 without CS. Failed Bravo studies occured in 24 (10.5%) patients. (Table 1) There were no significant differences compared to the control group regarding mean patient age (p=0.2), gender (p=0.8), patients with spastic motility disorders (p=0.8), year of placement (p=0.4), or probes placed with CS (p=0.08). All patients tested negative for CRP, anti-SM, SSA, SSB, RNP antibodies; 1 patient had an isolated ANA level≥1:320 of unclear significance. Conclusions: Failure of Bravo attachment or early dislodgement occurred in 10% of studies. Patients with failed Bravo recordings did not appear to have any risk factors for this occurence compared to the control group. Patients with Failed Bravo Studies

AGA Abstracts

* Underwent subsequent successful Bravo. +CS=with conscious sedation; -CS=without sedation N/A = not obtained. S1936 Non Acid Gastroesophageal Reflux in Children with Esophagitis Marina Orsi, Judith Cohen Sabban, Gabriela Donato Bertoldi, Veronica Busoni, Laureana Olleta, Daniel E. DAgostino, Silvia B. Christiansen The possible mechanisms involved in the development of esophagitis is still controversial. It has been postulated that non acid reflux could exert a more deleterious effect on the esophagus This mechanism could be important in Barrett's occurrence. The Multichannel Intraluminal Impedance -pH monitoring (MII-pH) is able to detect acid or non acid gastroesophageal reflux (GER) as well as the bolus clearance and thus determine the different impacts of these on the mucosa. Aim : To evaluate esophagitis in relation to the presence of acid or non acid exposure, bolus clereance and height of the column in children with gastroesophageal reflux. Materials and Methods: Since May 2005 to November 2008, a prospective study was conducted in forty eight children (32 boys /16 girls) with 8.5 years median age (r 1-18yrs) suspected of GER. All of them were evaluated at the Gastroenterology Unit of Hospital Italiano-Buenos Aires. First, an upper endoscopy with biopsies was performed and subsequently a 24 hr MII-pH study with a Sleuth Monitoring Recorder using catheters (ZIN or ZPN S61CO1E) with 7 impedance sensors and one pH probe at the distal end. The biopsies were informed by two different pathologists in a blinded manner. Patients with congenital anomalies, mental retardation or on medications were excluded. Results: There was no significant difference in: number of acid episodes nor with clearance bolus in children with or without esophagitis. In this group of patients, the esophageal damage seems to be more related with non acid reflux than with acid exposure. Conclusion: In this study esphagitis was more related to the presence of an increased non acid to acid pattern, instead poor correlation was seen with the number of episodes or with bolus clearance. It is the quality of the reflux material and not the time of exposure which appears to be more harmful to the esophagus. Further studies are necessary to confirm or not these observations and thus help prevent dangerous mechanisms in longstanding disease. The combination of impedance with endoscopy may improve our knowledge on this frequent and complex entity. Results

A-296