M1692
and IlI scores), 48 h (Ranson score) and 72 h (CTSI) after admission. 78 patients with early CECT were prospectively enrolled in this study (gallstone-related, 60.2%; alcohol-related, 51%; secondary', 6.4%; idiopathic, 28.2%), 44 patients were classified as mild and 34 as severe AP When assesing AP severity, CTS1 outpertormed the other 3 scores in terms of posinve (PP\O and neganve predictive value (NPV) (89% and 83% respectively, vs 73% and 69% tar APACHE III score, 76% and 71% for APACHE ll score, and 65% and 88% for Ransou criteria), \u assesing development of organ failure, APACHE Ill score achieved the higher PPu (77% vs 50% Ior APACHE II score, 45% for Ranson criteria and 33% for CTSI) and APACHE iI score achieved the higher NPV (92% vs 89% for APACHE 11 score and 88% for Ranson cnteria and CTSI ). l~anson criteria and APACHE lI and IlI scores ~rfbrmed poorly with respect to pancreatic necrosis development prediction (PPV, 44%, 48% and 55% respecnvely) When we compared LRPT, only"that for CTSI (11.21 vs 2.41 tar Rauson score, 4,09 fur APACHE II and 366 tbr APACHE Iil score) was able to generate large and conclusive changes ham pre-test to post-test probability in AP severity prediction. APACHEll and III scmes and Ranson score performed slightly better than CTS1 in predicting organ f~ilme (LRPT:4.06, 3.28 3.03 and 1.79 respecuvely), while when predicting pancreatm necrosis, APACHE Ii and Iil scores perfomled slightly better than Ranson score (LRVr: 2.07, 275 and 1.78 respectwely), in all outcome measures APACHE scores generate small and of similar extent changes in pmbabdity. CTSI is superior to other scoring systems in predicting AP severity and pancreati~ necrosis development, however both Ranson and APACHE scores perfbrm better with respect to organ failure prediction.
Diagnostic Efficacy of Magnetic Resonance Cholangiopancreatography after Secretin Stimulation (MRCP-S) in Pancreatic Disease Isabel Pascual, Jose Soler, Vicent Hern_andez,Jose Vicente kopez, Ramon Anon, Pedro Aimela, Andres Pena, Cirilo Amoros, Vicente Sanchlz, Miguel Minguez, Adolto Benages Objectives: To assess whether the secretin stimulation improves diagnostic efficacy of MRCP in patients with severe chronic pancreatitis and acute pancreatitis Patients and methods: MRCP-S was performed in 9 patients diagnosed of severe chminc pancrearitis by means of ultrasonogmphy and/or computed tomography (group 1), in 32 patients with acute pancreatitis of alcoholic or idiopathic origin (group 2) and in 6 patients without pancreatic or biliary" disease (control group). MR studies were carried out with a Sonata 1.5 T MR imager (Siemens). Images were obtained before secretin administration (I U/Kg) and every minute during the following 10 minntes.The parameters assessed before and after the administration of secretin were the following: diameter of the main pancreatic duct (MPD) before secretin stimulation and at 3 and 10 minutes afterward; duodenal filling at 10 rain; number of segments of the MPD visualized (head, body, tail); visualization of side branches; presence of ductal narrowing; endoluminal filling defects; visualization of accessory duct and presence of pancreas divisum Results: There was a significant increase of the MPD diameter at 3 and 10 min after secretin in the three groups. The baseline MPD diameter, at 3 rain and at 10 min were significantly greater in group 1 than in group 2 and control group; there were no differences between group 2 and control group. The table shows the results of qualitative image analysis before (pre-S) and after secretin (post4) in groups 1 and 2. In control group, there were no differences in the assessed parameters before and alter secretin. Conclusions: The MRCP-S improves the visnafizatmn of pancreatic ducts. The improvement of duct visualization is greater in patients without chronic pancreatiris because the dilation of MPD in chronic pancreatitis patients makes easier its baseline visualization
M1690 The Effect of Gastroenterology Training on Practice Patterns and Outcomes in Acute Uncomplicated Pancreatitis Timothy g Gardner Andrea M. Osborne, Todd iVlackenzie, Douglas L, Robertson
Group I (CP) Paramoteal Pre-S Post-S Side branches 7 (77,8) 8 (88,9) Ductal narrowing 4 (44.4) 4 (44.4) duct 4 (44,4) 6 (66,7) Pencrens divlsum 0 (0) 1 (11.1) No.MPDsegments 25 (9Z6) 27 (100) visualized Endolum~al filling 2 (22.2) 2 (22.2) defects Data in parenthesisare percentages
BACKGROUND: Patients admitted with acute pancreariris may be u~ated primarily by gastmeutemlogists or internists. The degree of vanarion in practice patterns between these groups and the ett)ct of that variation on outcome is unknown. AIM: Examine practice patterns and hospital outcome in a group of patients with acute pancrearitis managed prnnardy by internists or gastroentemlogists. METHODS: Patients admitted to our medical center since 1990 with tbe first diagnosis of acute uncomplicated pancreatitis on the general medicine or gastroenterology serdces were identified using ICD-9 codes. Charts were abstracted for relevant process (feeding, laboratory testing, imaging, use of antibiotics and proeedures) and outcome measures (mortality, length of stay', 30 day readmission). Appmpnale parametric testing liar cominuous (t test) and categorical (chi square) variables, logistic regression and propensity score adjusted analyses were used to assess for between group difterences. RESUI.TS: 163 patients were identified (122 on the general internal medicine service and 41 on the gastroememlogy service). Patients admitted in transfer (n = 35) were excluded leaving 128 tar analysis. Tbe two groups were similar in demographics, modified APACHE-II, Cbarlson scores and etiology of panereatius. Of the processes examined, only" two were significantly' difte:rent: i) Gastroentemlogists were more likdy than internists to ruder admission Ci" scans (55% vs. 24%, p<0.012) and ii) patients treated by"gastroentemlogists were more likely"to have an ERCP (50% vs. 9.3%, p<0.001). There was no significant tonality difference belween gastrocntemloglsts (n=0) and internists (n= 1). There was no significant difi'eren~e in 30 day readmission rates. Mean length of stay* was shorter for gastroenterologtsts than tar internists (6.8 vs. 8.9 day's), however this was not significant (p=0.50). CONCLUSIONS: W~nile there was some subtle variation in practice patterns between gastroenterologists and internists ueating patients hospitalized with a first diagnosis of acnte uncomplicated pancreatiris, these variations did not effect important outcomes.
p NS NS NS NS NS
Pm-S 3 (9,4) 2 (63) 5 (15.6) 2 (6,3) 77 (80.2)
Group z CAP) Post-S 14 (43.8) 6 (18~8) 21 (65.6) 5 (15.6) 96 (100)
p < 0,001 < 0,05 <0.001 <0,05 < 0,001
NS
0 (0)
1 (3,1)
NS
M1693 Incidence of Acute Pancreatitis Due to Antiretroviral Treatment Olivia Hentic, C. Charlois, C. Barennes, T May, J Reynes, A. Waldner, Jm Lang, C. Chene, Philippe Levy, C. keport, Philippe Ruszniewski, Gmupe D'emde Aproco Acute pancreatitis (AP) secondary to antiretmviral (AR\O medications m patients treated with HIV appears frequent and may"lead to treatment discontinuation. Aim: determine the incidence and severity of AP due to ARV in H1V+ patients on triple-therapy containing a pmtease inhibitor. Methods: A cohort of 1281 patients on ARV treatment with a protease inhibitor were followed prospectively. Case files of patients with documented AP on ARV (as validated by Adverse Events Committee) were examined retrospectively. Diagnostic revision criteria of AP were in accm-dance with the French Consensus on AP (2001). AP severity accorded with Atlanta 1992 criteria and a drag cause was imputed as defined by Pancreatox Resuhs: From 1997 to 2001, 14/1281 patients (1%), (10 males, 4 females: mean age 36 yrs) had documented AP (density incidence = 0.4% person-years). ARV treatment, either alone or in combination, most frequently incriminated were: DDI (n = 7), D4T (n = 6) and 3TC (n = 2). Mean delay-between introductinn of drag and AP was 7.3 months (range: 0.1-18). Only"9 patients conformed to the Consensus criteria (pain + enzymes: n=3; CT: n = 5; or autopsy: n = 1). Remaining patients had either asymptomatic elevation of pancreatic enzymes (n = 3) or atypical or poorly defined pain and/or non significant augmentation of enzymes without radiological confirmation (n = 2). In 9 patients with AP a cause other than drags was confirmed in 2 (alcohol n = 1; bihary n = 1). A biliary cause could not be formally excluded as abdominal US to eliminate biliary stones was available in only 6 patients and none had EUS. A drag cause was imputed as possible or probable in 7 other cases. Severity criteria were mulrimrgan failure (n = 4), which was always fatal (decompensated viral hepatitis in 3 cases) and pancreatic or extra-pancreatic necrosis (n = 7). Four of 7 patients wath AP secondary to AVR died. Conclusions: a) the diagnosis of AP due to A~N is evoked in 1% of HIV+ patients on a protease inhibitor b) excessive diagnoses are made in half of cases (e.g., asymptomatic elevation of pancreatic enzymes) c) other causes of AP (alcohol, bihary) should be thoroughly eliminated d) severity of AP in patients on AVR appears higher than in immunocompetent individuals. As therapeutk modifications are frequently imposed in such cases of "AP', standardization of diagnostic criteria as well as prospective evaluation of the clinical significance of isolated enz}~oe elevation in these patients is mandatoD,.
M1691 Acute Pancreatitis in Crohn's Disease: Clinical Features and Outcomes Plcha Moolsintong, Edward V. Lofius Jr., Suresh T. Churl, Laurence J. Egan, William J. rremaine, William J Sandborn BACKGROUND: Acute pancreatins has occasionally been associated with Cmhn's disease (CD) The relative trequeucy of various causes of pancreatitis in CD patients (biliary, medication-related, duodenal CD, autoimmunc) have not been well described. METHODS: A centralized diagnostic index was used to identit}"all CD patients with acute pancrearitis that were evaluated at Mayo Clinic Rochester between 1976 and 2001. Pancrearitis diagnosis was made or confirmed at our institution biochemically (amylase and/or lipase -> 2x normal) or radiographica/ly (CT or US), CD diagnosis was also made or confirmed at our institution by pathology, endoscopy or radiography, Records were absu-acted for demographics, presenting symptoms, diagnostic tests, risk factors of pancmaritis, treatment, and follow-up. RESULTS: Forty
M1694 Clinical Features and Prognostic Factors of Recurrent Acute Biliat3, Pancreatitis Vicent Hemandez Sr., Isabel Pascual, Pedro Almela, Ramon Anon, Cirilo Amoros, Belen Herreros, Francisco Mora, Miguel Minguez, Adolfo Benages Objectives: To assess the prewalence of recurrence of acute biliary pancreatins (ABP), its clinical features and the presence of pmgnosric tactors of recurrence. Patients and methods: Patients admitted with ABP from 1-Jan-2000 to 31-May-2002 were diwided into 2 groups according to if they had presented or not recurrence of ABP (groups A and B, respectively). In both groups, demographic cha~-acterisrics,severity (Atlanta Criteria), presence of choledocholithiasis, analytical parameters, Bafihazar grade and treatment of biliary- lithiasis were studied. In group A, time lapsing until recurrence and its clinical characteristics were
A-399
AGA Abstracts