Advertisement

More Favorable Outcomes with Peptic Ulcer Bleeding Due to Helicobacter Pylori

  • Rebecca D. Chason
    Affiliations
    Division of Digestive and Liver Diseases, Department of Internal Medicine, University of Texas Southwestern Medical Center and Parkland Memorial Hospital, Dallas
    Search for articles by this author
  • Joan S. Reisch
    Affiliations
    Division of Biostatistics, Department of Clinical Sciences, University of Texas Southwestern Medical Center and Parkland Memorial Hospital, Dallas
    Search for articles by this author
  • Don C. Rockey
    Correspondence
    Requests for reprints should be addressed to Don C. Rockey, MD, Department of Internal Medicine, Medical University of South Carolina, 96 Jonathan Lucas Street, Suite 803, MSC 623, Charleston, SC 29425.
    Affiliations
    Department of Internal Medicine, Medical University of South Carolina, Charleston
    Search for articles by this author

      Abstract

      Background

      Acute upper gastrointestinal bleeding is a common complication of peptic ulcer disease, often caused by Helicobacter pylori and nonsteroidal anti-inflammatory drug (NSAID) use. The purpose of this study was to determine whether the cause and biologic behavior of ulcers associated with acute upper gastrointestinal bleeding might lead to divergent patient outcomes.

      Methods

      In this Institutional Review Board-approved study, we compared clinical features and outcomes of patients with acute upper gastrointestinal bleeding due to ulcers categorized into 4 groups: Helicobacter pylori positive or negative combined with NSAID usage positive or negative. Likelihood chi-squared analyses were utilized for group comparisons and stepwise multiple logistic regression models were utilized to determine which factors were related to bleeding outcomes.

      Results

      Of 2242 patients with upper gastrointestinal bleeding, 575 (26%) had gastroduodenal ulcer disease, and of those with appropriate diagnostic testing, approximately half (228, 10% overall) had evidence of Helicobacter pylori infection and half (216, 10% overall) had no evidence of Helicobacter pylori infection. Patients without Helicobacter pylori infection had significantly more comorbid conditions than those with Helicobacter pylori and higher Charlson Index comorbidity scores (2.6 ± 2.6 [mean and SD] vs 1.9 ± 2.3, P = .003). Hospital length of stay was significantly longer for Helicobacter pylori-negative patients (mean 11.4 ± 21.7 vs 6 ± 8.5 days and median 5.5 vs 3 days, P <.001 and <.001, respectively). Rebleeding events within 30 days were more frequent in Helicobacter pylori-negative patients than Helicobacter pylori-positive patients (11% vs 5%, P = .009). Rebleeding was most frequent in patients without Helicobacter pylori and with no reported use of NSAIDS (18%, P = .01).

      Conclusions

      Helicobacter pylori-negative ulcers were associated with poorer outcomes regardless of use of NSAIDs. Patients with ulcers negative for Helicobacter pylori and no history of NSAID use had the worst outcomes and had more severe systemic disease.

      Keywords

      Acute upper gastrointestinal bleeding is a common medical condition, often requiring hospital admission, that affects approximately 400,000 individuals per year in the US.
      • Gralnek I.M.
      • Barkun A.N.
      • Bardou M.
      Management of acute bleeding from a peptic ulcer.
      • Laine L.
      • Jensen D.M.
      Management of patients with ulcer bleeding.
      Peptic ulcer disease is currently the most common cause of upper gastrointestinal bleeding, accounting for an estimated 25%-30% of all cases.
      • Laine L.
      • Jensen D.M.
      Management of patients with ulcer bleeding.
      • Sung J.J.
      Management of nonsteroidal anti-inflammatory drug-related peptic ulcer bleeding.
      • Balderas V.
      • Bhore R.
      • Lara L.F.
      • Spesivtseva J.
      • Rockey D.C.
      The hematocrit level in upper gastrointestinal hemorrhage: safety of endoscopy and outcomes.
      • Lau J.Y.
      • Sung J.
      • Hill C.
      • Henderson C.
      • Howden C.W.
      • Metz D.C.
      Systematic review of the epidemiology of complicated peptic ulcer disease: incidence, recurrence, risk factors and mortality.
      • van Leerdam M.E.
      Epidemiology of acute upper gastrointestinal bleeding.
      The 2 major underlying causes of peptic ulcer disease (defined as duodenal or gastric ulcer) are Helicobacter pylori infection and nonsteroidal anti-inflammatory drug (NSAID) use.
      • Sung J.J.
      Management of nonsteroidal anti-inflammatory drug-related peptic ulcer bleeding.
      • Lau J.Y.
      • Sung J.
      • Hill C.
      • Henderson C.
      • Howden C.W.
      • Metz D.C.
      Systematic review of the epidemiology of complicated peptic ulcer disease: incidence, recurrence, risk factors and mortality.
      • van Leerdam M.E.
      Epidemiology of acute upper gastrointestinal bleeding.
      • Wong G.L.
      • Wong V.W.
      • Chan Y.
      • et al.
      High incidence of mortality and recurrent bleeding in patients with Helicobacter pylori-negative idiopathic bleeding ulcers.
      • Gisbert J.P.
      • Calvet X.
      • Cosme A.
      • et al.
      Long-term follow-up of 1,000 patients cured of Helicobacter pylori infection following an episode of peptic ulcer bleeding.
      • Fujinami H.
      • Kudo T.
      • Hosokawsa A.
      • et al.
      A study of the changes in the cause of peptic ulcer bleeding.
      • Lee I.
      • Cryer B.
      Epidemiology and role of nonsteroidal antiinflammatory drugs in causing gastrointestinal bleeding.
      • Sostres C.
      • Lanas A.
      Epidemiology and demographics of upper gastrointestinal bleeding: prevalence, incidence, and mortality.
      • Liu N.J.
      • Lee C.S.
      • Tang J.H.
      • et al.
      Outcomes of bleeding peptic ulcers: a prospective study.
      • Ootani H.
      • Iwakiri R.
      • Shimoda R.
      • et al.
      Role of Helicobacter pylori infection and nonsteroidal anti-inflammatory drug use in bleeding peptic ulcers in Japan.
      • Zapata-Colindres J.C.
      • Zepeda-Gomez S.
      • Montano-Loza A.
      • Vazquez-Ballesteros E.
      • de Jesus Villalobos J.
      • Valdovinos-Andraca F.
      The association of Helicobacter pylori infection and nonsteroidal anti-inflammatory drugs in peptic ulcer disease.
      • Chan H.L.
      • Wu J.C.
      • Chan F.K.
      • et al.
      Is non-Helicobacter pylori, non-NSAID peptic ulcer a common cause of upper GI bleeding? A prospective study of 977 patients.
      • Sung J.J.
      • Kuipers E.J.
      • El-Serag H.B.
      Systematic review: the global incidence and prevalence of peptic ulcer disease.
      • Boltin D.
      • Halpern M.
      • Levi Z.
      • et al.
      Gastric mucin expression in Helicobacter pylori-related, nonsteroidal anti-inflammatory drug-related and idiopathic ulcers.
      Recent studies have suggested that some peptic ulcers are unrelated to either NSAID use or Helicobacter pylori infection and have been termed “idiopathic ulcers.”
      • Wong G.L.
      • Wong V.W.
      • Chan Y.
      • et al.
      High incidence of mortality and recurrent bleeding in patients with Helicobacter pylori-negative idiopathic bleeding ulcers.
      • Boltin D.
      • Halpern M.
      • Levi Z.
      • et al.
      Gastric mucin expression in Helicobacter pylori-related, nonsteroidal anti-inflammatory drug-related and idiopathic ulcers.
      • Gisbert J.P.
      • Calvet X.
      Review article: Helicobacter pylori-negative duodenal ulcer disease.
      • Adamopoulos A.B.
      • Efstathiou S.P.
      • Tsioulos D.I.
      • et al.
      Bleeding duodenal ulcer: comparison between Helicobacter pylori positive and Helicobacter pylori negative bleeders.
      • Niv Y.
      H. pylori/NSAID-negative peptic ulcer—the mucin theory.
      • Helicobacter pylori-negative ulcers are associated with poorer outcomes regardless of use of nonsteroidal anti-inflammatory drugs.
      • Peptic ulcer rebleeding is less likely in women and patients positive for Helicobacter pylori.
      • Patients with ulcer bleeding should have early and aggressive testing for Helicobacter pylori to help triage optimal care.
      Helicobacter pylori are gram-negative, flagellated bacteria that colonize in the gastric mucosa. In the US, the reported prevalence of Helicobacter pylori varies from <10% to nearly 80%, with the highest rates found in patients from lower socioeconomic living conditions.
      • Goh K.L.
      • Chan W.K.
      • Shiota S.
      • Yamaoka Y.
      Epidemiology of Helicobacter pylori infection and public health implications.
      • Malaty H.M.
      Epidemiology of Helicobacter pylori infection.
      Clinically, infection presents most often as peptic ulcer disease; however, it should be noted that only 20% of infected individuals develop peptic ulcer disease and only a quarter of those develop ulcer complications such as upper gastrointestinal bleeding.
      • Malaty H.M.
      Epidemiology of Helicobacter pylori infection.
      NSAIDs, including aspirin, are widely used medications for analgesia, as anti-inflammatory agents, and for their antiplatelet effects.
      • Sostres C.
      • Lanas A.
      Epidemiology and demographics of upper gastrointestinal bleeding: prevalence, incidence, and mortality.
      Although their use is often underreported, it is estimated that over 60 million Americans take them on a routine basis. Further, they are one of the most frequently used medications worldwide.
      • Lee I.
      • Cryer B.
      Epidemiology and role of nonsteroidal antiinflammatory drugs in causing gastrointestinal bleeding.
      • Sostres C.
      • Lanas A.
      Epidemiology and demographics of upper gastrointestinal bleeding: prevalence, incidence, and mortality.
      NSAIDs have been implicated in the mucosal injury of both the upper and lower gastrointestinal tract, where both topical and systemic effects may lead to cellular injury, ulceration, and bleeding.
      • Lee I.
      • Cryer B.
      Epidemiology and role of nonsteroidal antiinflammatory drugs in causing gastrointestinal bleeding.
      • Sostres C.
      • Lanas A.
      Epidemiology and demographics of upper gastrointestinal bleeding: prevalence, incidence, and mortality.
      Based on our own clinical experience and previous data suggesting that the underlying etiology of peptic ulcers may play a role in their clinical course,
      • Lau J.Y.
      • Sung J.
      • Hill C.
      • Henderson C.
      • Howden C.W.
      • Metz D.C.
      Systematic review of the epidemiology of complicated peptic ulcer disease: incidence, recurrence, risk factors and mortality.
      • van Leerdam M.E.
      Epidemiology of acute upper gastrointestinal bleeding.
      • Wong G.L.
      • Wong V.W.
      • Chan Y.
      • et al.
      High incidence of mortality and recurrent bleeding in patients with Helicobacter pylori-negative idiopathic bleeding ulcers.
      • Chan H.L.
      • Wu J.C.
      • Chan F.K.
      • et al.
      Is non-Helicobacter pylori, non-NSAID peptic ulcer a common cause of upper GI bleeding? A prospective study of 977 patients.
      we hypothesized that patients with Helicobacter pylori-negative ulcers may have different outcomes than those positive for Helicobacter pylori. Therefore, the purpose of this study was to determine whether the inherent differences in the cause and biologic behavior of ulcers associated with acute upper gastrointestinal bleeding may be associated with different outcomes.

      Methods

      This study was approved by the University of Texas Southwestern Medical Center Institutional Review Board and met all criteria for good clinical research.
      • Tognoni G.
      The new Helsinki declaration.
      This was a retrospective analysis of all patients who presented to Parkland Memorial Hospital (a University of Texas Southwestern teaching hospital, Dallas, Tex) between January 1, 2006 and February 27, 2012 with acute upper gastrointestinal bleeding, who underwent diagnostic or therapeutic endoscopy by a dedicated gastroenterologist within 24 hours of admission, were found to have peptic ulcer disease as the bleed etiology, and were tested for Helicobacter pylori. Patients with gastrointestinal bleeding were identified and data pertaining to their hospital admission were entered prospectively into our institution's Gastrointestinal Health Care registry (Microsoft Access, Microsoft Corporation, Redmond, Wash). Demographic data included age, sex, and ethnicity, as well as clinical data including symptoms, medical and social history, medications, admission physical examination findings, laboratory data including Helicobacter pylori diagnostic testing, and endoscopic procedure findings (diagnosis, stigmata of recent or active bleeding, and therapeutic intervention). Thirty-day rebleeding events and 30-day mortality also were collected. Patients with documented Helicobacter pylori diagnostic testing (see below) were initially grouped and compared by Helicobacter pylori diagnosis (positive or negative) and then were subdivided and compared by use of any NSAID (positive or negative). Ulcers associated with malignant lesions, those with equivocal diagnostic results, or those without diagnostic testing were excluded from analysis.

      Definitions

      Acute upper gastrointestinal bleeding was defined as bleeding proximal to the ligament of Treitz
      • Gralnek I.M.
      • Barkun A.N.
      • Bardou M.
      Management of acute bleeding from a peptic ulcer.
      presenting as reported or witnessed hematemesis, melena, or hematochezia within 7 days of admission, with a decrease in hematocrit from baseline ≥4 points.
      • Rockey D.C.
      Gastrointestinal bleeding.
      NSAID use was defined as use of either aspirin or any known NSAID more than 4 times in the month before admission. Helicobacter pylori status was determined by employing one or more of the following techniques: 1) serological testing for Helicobacter pylori antibodies, 2) stool samples for antigens, 3) random biopsies taken during endoscopy, and 4) the Campylobacter-like organism test or Rapid urease test. The majority of peptic ulcer patients had more than one diagnostic test performed. If any of the tests were positive for Helicobacter pylori, the patient was presumed to have a Helicobacter pylori-mediated ulcer. Thus, all diagnostic tests performed on any patient had to be negative for the patient to be classified as “Helicobacter pylori negative.” Patients with equivocal results for serological testing or evidence of malignancy postbiopsy were excluded from analysis.
      By design, we differentiated patients into the following 4 groups:
      • 1.
        Helicobacter pylori negative/NSAID negative (HP−/NSAID−) (also idiopathic),
      • 2.
        Helicobacter pylori negative/NSAID positive (HP−/NSAID+),
      • 3.
        Helicobacter pylori positive/NSAID negative (HP+/NSAID−),
      • 4.
        Helicobacter pylori positive/NSAID positive (HP+/NSAID+).
      Comorbid conditions collected were as follows and included: diabetes mellitus, hypertension, coronary artery disease, congestive heart failure, atrial fibrillation, myocardial infarction, asthma, chronic obstructive pulmonary disease, end-stage renal disease, cerebrovascular accident, peptic ulcer disease, gastroesophageal reflux disease, cirrhosis, neoplasia, acquired immune deficiency syndrome, and hepatitis B or C (or both).
      Endoscopic therapy was performed as per the standard of care for ulcers with stigmata of bleeding.
      • Gralnek I.M.
      • Barkun A.N.
      • Bardou M.
      Management of acute bleeding from a peptic ulcer.
      • Laine L.
      • Jensen D.M.
      Management of patients with ulcer bleeding.
      • Barkun A.N.
      • Martel M.
      • Toubouti Y.
      • Rahme E.
      • Bardou M.
      Endoscopic hemostasis in peptic ulcer bleeding for patients with high-risk lesions: a series of meta-analyses.
      Endoscopic therapies include epinephrine injection, thermal coagulation, and hemoclip placement.
      • Gralnek I.M.
      • Barkun A.N.
      • Bardou M.
      Management of acute bleeding from a peptic ulcer.
      • Laine L.
      • Jensen D.M.
      Management of patients with ulcer bleeding.
      The standard practice at our institution is to administer combination therapy, which may include injection of epinephrine and thermal coagulation, injection and hemoclip placement, or injection with thermal coagulation and hemoclip placement to achieve endoscopic hemostasis.
      Causes of death for patients were classified into 8 categories as follows: gastrointestinal bleeding, cardio or respiratory failure, renal failure, liver failure or cirrhosis complications, sepsis, multiorgan dysfunction syndrome, terminal malignancy, or other. The study team adjudicates the cause of death in a blinded fashion; death was considered to be due to bleeding when the patient either died while actively bleeding or when the bleeding event led to a subsequent event that caused death (eg, surgery).

      Statistical Analysis

      Analyses for this research project were carried out utilizing SAS, version 9.2 (SAS Institute Inc., Cary, NC). Data were compiled for 444 subjects. Of those, 216 patients were Helicobacter pylori negative and 228 were Helicobacter pylori positive. Multiple demographic and clinical variables were examined, including age; sex; ethnicity; aspirin and NSAID use; smoking and alcohol use; comorbidities; Charlson Index; presence of hematemesis, melena or hematochezia; vital signs; and multiple laboratory values. Subjects were categorized into 4 groups: Helicobacter pylori positive or negative combined with NSAID usage positive or negative.
      Categorical data items were summarized utilizing frequency counts and percentages while means and SDs were calculated for the numerical measurements. In order to gain insight for the multivariate analyses, the individual measurements were each compared, grouped according to the outcome measure. Likelihood chi-squared analyses were utilized for group comparisons of each of the categorical measurements and one-way analysis of variance for the numerical measurements.
      Stepwise multiple logistic regression models were utilized to determine which of the demographic and clinical risk factors were statistically related to the bleeding outcome. Of the 444 subjects, 429 (96.6%) had complete data. The model entry criteria were selected at 5%. The Hosmer Lemeshow technique was utilized to assess the model fit.

      Results

      Approximately 25% of patients with upper gastrointestinal bleeding had ulcer disease, and of those with appropriate diagnostic testing, approximately half had evidence of Helicobacter pylori infection and half did not (Figure). Of those positive for Helicobacter pylori, 125 (55%) reported use of NSAIDS, while of those negative for Helicobacter pylori, 143 (66%) reported use of NSAIDS (P = .018) (Supplemental Table 1). The epidemiology of ulcers was notably different: more men than women were Helicobacter pylori positive, and patients of Hispanic or black ethnicity were significantly more likely to be Helicobacter pylori positive than white patients (Table 1). Further, patients without Helicobacter pylori infection had significantly more comorbid conditions and higher Charlson Index comorbidity scores
      • Charlson M.E.
      • Pompei P.
      • Ales K.L.
      • MacKenzie C.R.
      A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.
      —with and without adjustment for age—than those with Helicobacter pylori (P = .001) (Table 1).
      Figure thumbnail gr1
      FigurePatient groups. The numbers of patients in the 4 different groups, as defined by Helicobacter pylori status and nonsteroidal anti-inflammatory drug (NSAID) use, is shown.
      Table 1Patient Characteristics
      H. pylori (−)

      n = 216
      H. pylori (+)

      n = 228
      P Value
      NSAID (−)

      n = 73

      n, Mean (%, SD)
      NSAID (+)

      n = 143

      n, Mean (%, SD)
      NSAID (−)

      n = 103

      n, Mean (%, SD)
      NSAID (+)

      n = 125

      n, Mean (%, SD)
      Age in years53 ± 1455 ± 1351 ± 1355 ± 14.077
       Range23–8818–8420–8522–93
      Sex.001
       Female27 (37%)58 (41%)18 (17%)38 (30%)
       Male46 (63%)85 (59%)85 (83%)87 (70%)
      Ethnicity<.0001
       Hispanic18 (25%)32 (22%)41 (40%)54 (43%)
       African American23 (32%)41 (29%)42 (41%)50 (40%)
       Caucasian27 (39%)59 (41%)17 (17%)15 (12%)
       Other
      Other ethnicities included Asian/Pacific Islander/East Indian and Native American.
      5 (7%)11 (8%)3 (3%)6 (5%)
      Smoking31 (42%)90 (63%)53 (51%)71 (57%).030
      Alcohol45 (62%)74 (52%)60 (58%)83 (66%).103
      Charlson Index2.4 ± 2.72.7 ± 2.61.9 ± 2.41.9 ± 2.3.021
      Charlson Age-adjusted Index3.2 ± 3.03.7 ± 3.22.5 ± 2.92.7 ± 2.9.006
      Number of comorbidities1.9 ± 1.32.4 ± 1.71.7 ± 1.51.9 ± 1.5.003
      NSAID = nonsteroidal anti-inflammatory drug.
      Other ethnicities included Asian/Pacific Islander/East Indian and Native American.
      Clinical features suggested that patients without Helicobacter pylori had more severe bleeding than those with Helicobacter pylori, including more prominent hematochezia (12 [16%], P = .01) (Table 2), slightly lower blood pressure (in HP–/NSAID+ patients), and more frequent admission to the intensive care unit (Table 2). Pre-endoscopy Rockall and American Society of Anesthesiologists (ASA) scores also were higher in Helicobacter pylori-negative patients compared with positive patients (Supplemental Table 1). Blood urea nitrogen levels were slightly higher in Helicobacter pylori-negative patients (39 ± 33 vs 34 ± 27, P = .05), consistent with more substantial bleeding (Supplemental Table 1).
      Table 2Clinical Features
      H. pylori (−)

      n = 216
      H. pylori (+)

      n = 228
      P Value
      NSAID (−)

      n = 73

      n, Mean (%, SD)
      NSAID (+)

      n = 143

      n, Mean (%, SD)
      NSAID (−)

      n = 103

      n, Mean (%, SD)
      NSAID (+)

      n = 125

      n, Mean (%, SD)
      Hematemesis44 (60%)62 (43%)51 (50%)61 (49%).134
      Melena43 (59%)118 (83%)85 (83%)97 (78%).001
      Hematochezia12 (16%)16 (11%)10 (10%)4 (3%).010
      ICU at admission24 (33%)51 (36%)22 (21%)28 (22%).027
      Admission vital signs
       Systolic blood pressure125 ± 29122 ± 28127 ± 25126 ± 22.471
       Diastolic blood pressure70 ± 2067 ± 1674 ± 1771 ± 16.022
       Pulse93 ± 2092 ± 2394 ± 2193 ± 20.945
      Rockall (ER)2.9 ± 1.32.9 ± 1.22.4 ± 1.42.6 ± 1.2.012
      ASA score2.4 ± 0.72.5 ± 0.72.2 ± 0.62.3 ± 0.6<.0001
      Admission laboratory values
       Hematocrit (%)30.0 ± 8.827.6 ± 8.629.7 ± 9.428.0 ± 8.4.122
       Platelets (×109/L)261 ± 147254 ± 126233 ± 99251 ± 113.434
       INR1.3 ± 0.61.5 ± 1.41.2 ± 0.41.2 ± 0.4.047
       BUN (mg/dL)36.4 ± 33.339.7 ± 33.030.8 ± 27.936.4 ± 25.8.153
      ASA = American Society of Anesthesiologists; BUN = blood urea nitrogen; ICU = intensive care unit; INR = international normalized ratio; NSAID = nonsteroidal anti-inflammatory drug.
      Helicobacter pylori-positive ulcers were located in the duodenum more often than Helicobacter pylori-negative ulcers (43% vs 30%, P = .016) (Table 3). Additionally, patients with Helicobacter pylori-negative ulcers more frequently had >3 ulcers compared with those that were Helicobacter pylori positive. However, there did not appear to be differences in the frequency of stigmata of recent bleeding or need for endoscopic therapy. All patients were treated with an intravenous or oral proton pump inhibitor during their hospital stay and discharged with prescriptions for proton pump inhibitors.
      Table 3Ulcer Characteristics and Treatment
      H. pylori (−)

      n = 216
      H. pylori (+)

      n = 228
      P Value
      NSAID (−)

      n = 73

      n (%)
      NSAID (+)

      n = 143

      n (%)
      NSAID (−)

      n = 103

      n (%)
      NSAID (+)

      n = 125

      n (%)
      Location of ulcers.016
       Stomach46/73 (63%)83/143 (58%)43/103 (42%)73/125 (58%)
       Duodenum22/73 (30%)44/143 (31%)52/103 (50%)46/125 (37%)
       Stomach + duodenum5/73 (7%)16/143 (11%)8/103 (8%)6/125 (5%)
      Number of ulcers.006
       148/73 (66%)77/143 (54%)71/103 (69%)80/125 (64%)
       27/73 (10%)30/143 (21%)20/103 (19%)27/125 (22%)
       33/73 (4%)10/143 (7%)7/103 (7%)8/125 (6%)
       >315/73 (21%)26/143 (18%)5/103 (5%)10/125 (8%)
      Stigmata present27/73 (37%)61/143 (43%)40/103 (39%)49/125 (39%).853
       Active bleeding10/27 (37%)20/61 (33%)14/40 (35%)10/49 (21%)
       Visible vessel9/27 (33%)28/61 (46%)20/40 (50%)31/49 (63%)
       Adherent clot8/27 (30%)13/61 (21%)6/40 (15%)8/49 (26%)
      Stigmata absent46/73 (63%)82/143 (57%)63/103 (61%)76/125 (61%)
       Pigmented spot8/46 (17%)18/82 (22%)12/63 (19%)6/76 (8%)
       Clean base38/46 (83%)64/82 (78%)51/63 (81%)70/76 (92%)
      Endoscopic therapy26/73 (36%)60/143 (42%)42/103 (41%)50/125 (40%).840
       Epinephrine injection7/26 (27%)5/60 (8%)3/42 (7%)4/50 (8%)
       Thermal coagulation1/26 (4%)3/60 (5%)3/42 (7%)4/50 (8%)
       Hemoclips5/26 (19%)5/60 (8%)0/42 (0%)6/50 (12%)
       Combination therapy13/26 (50%)47/60 (79%)36/42 (86%)36/50 (72%)
      NSAID = nonsteroidal anti-inflammatory drug.
      Duodenal ulcers were more often Helicobacter pylori positive than gastric ulcers (60% vs 47%, P = .014) (Supplemental Table 2). Ulcer characteristics were similar in the 2 locations; however, duodenal ulcers appeared to have a slightly higher frequency of having stigmata of recent bleeding and need for endoscopic therapy, although these differences were not statistically significantly different.
      The length of hospital stay was significantly longer for Helicobacter pylori-negative patients (Table 4, Supplemental Table 1). HP−/NSAID− patients had the longest median length of stay (7 days), and HP+/NSAID+ and HP+/NSAID− stayed the shortest amount of time (3 days). In addition, rebleeding events within 30 days were more frequent in Helicobacter pylori-negative patients (P = .01) (Table 4, Supplemental Table 1); HP−/NSAID− ulcer patients had the highest rate of rebleeding at 18%. Mortality at 30 days was similar for all 4 groups, with the most common cause of death being multi-organ dysfunction syndrome and respiratory/cardiac failure. There were 3 deaths specifically due to gastrointestinal bleeding; 2 of which were in Helicobacter pylori-negative patients.
      Table 4Outcomes
      H. pylori (−)

      n = 216
      H. pylori (+)

      n = 228
      P Value
      NSAID (−)

      n = 73

      n, Mean (%, SD)
      NSAID (+)

      n = 143

      n, Mean (%, SD)
      NSAID (−)

      n = 103

      n, Mean (%, SD)
      NSAID (+)

      n = 125

      n, Mean (%, SD)
      RBC transfused47 (64%)100 (70%)68 (66%)76 (61%).472
       Units4.5 ± 5.14.2 ± 3.43.6 ± 3.43.8 ± 2.5.543
      Platelets transfused2 (3%)3 (2%)1 (1%)2 (2%).827
       Units2.0 ± 1.41.3 ± 0.66.01.0 ± 0.0.026
      FFP transfused9 (12%)11 (8%)4 (4%)6 (5%).136
       Units4.0 ± 3.32.6 ± 1.23.5 ± 2.44.8 ± 3.8.427
      ICU transfer6 (8%)8 (6%)4 (4%)6 (5%).659
      Mechanical ventilation7 (10%)7 (5%)1 (1%)7 (7%).049
      Angiography2 (3%)0 (0%)2 (2%)1 (1%).165
      Surgery5 (7%)4 (3%)3 (3%)1 (1%).128
      Length of stay (mean, days)11.7 ± 12.511.3 ± 25.26.5 ± 9.85.6 ± 7.4.007
      Length of stay (median, days)7 (1, 53)5 (1, 238)3 (1, 62)3 (1, 41)<.0001
      Rebleed within 30 days13 (18%)12 (8%)4 (4%)7 (6%).010
      Mortality within 30 days4 (6%)5 (4%)4 (4%)4 (3%).881
      FFP = fresh frozen plasma; ICU = intensive care unit; NSAID = nonsteroidal anti-inflammatory drug; RBC = red blood cells.
      Regression analysis to evaluate predictors of rebleeding revealed that being Helicobacter pylori positive was associated with a decreased likelihood of rebleeding (Table 5). Additionally, as expected, endoscopic therapy was associated with an increased risk of rebleeding, although combination therapy of all types reduced the risk of rebleeding compared with single-modality therapy.
      Table 5Variables Associated with Rebleeding
      VariableOdds Ratio95% Confidence Interval
      Sex (female)0.2820.106-0.747
      Helicobacter pylori positive0.3930.178-0.868
      Hematochezia2.5931.029-6.536
      Number of comorbidities1.5001.189-1.892
      Therapy—epinephrine injection only6.9971.989-24.609
      Therapy—thermal coagulation only6.8401.419-32.963
      Therapy—combination2.4391.076-5.518
      The table includes data from a logistic regression analysis designed to evaluate factors associated with rebleeding; the Hosmer Lemeshow test for the goodness of model fit indicated a reasonable fit with chi-squared = 8.13 (df = 8) and P = .421.

      Discussion

      In this study, we have documented a number of demographic and clinical differences in patients with peptic ulcer disease, depending on Helicobacter pylori and NSAID status. We also have shown that patients with Helicobacter pylori-negative ulcers appear to have a generally greater degree of morbidity than those with Helicobacter pylori-positive ulcers. Patients with Helicobacter pylori-negative ulcers also exhibited differences in certain outcomes such as a longer hospital stay and more frequent rebleeding events at 30 days. Finally, the data suggest that Helicobacter pylori infection status may be more critical than use of NSAIDs for predicting poorer outcomes.
      We found that female sex and Helicobacter pylori infection were negative predictors of rebleeding. This is consistent with other data showing that patients positive for Helicobacter pylori are typically treated with antibiotics and cured, thus decreasing their rate of rebleeding.
      • Gisbert J.P.
      • Calvet X.
      • Cosme A.
      • et al.
      Long-term follow-up of 1,000 patients cured of Helicobacter pylori infection following an episode of peptic ulcer bleeding.
      • Malfertheiner P.
      • Megraud F.
      • O'Morain C.A.
      • et al.
      Management of Helicobacter pylori infection—the Maastricht IV/ Florence Consensus Report.
      • Graham D.Y.
      • Fischbach L.
      Helicobacter pylori treatment in the era of increasing antibiotic resistance.
      Endoscopic therapy was a significant positive predictor of rebleeding. This is consistent with the concept that high-risk lesions typically are selected for endoscopic therapy.
      • Gralnek I.M.
      • Barkun A.N.
      • Bardou M.
      Management of acute bleeding from a peptic ulcer.
      • Laine L.
      • Jensen D.M.
      Management of patients with ulcer bleeding.
      • Barkun A.N.
      • Martel M.
      • Toubouti Y.
      • Rahme E.
      • Bardou M.
      Endoscopic hemostasis in peptic ulcer bleeding for patients with high-risk lesions: a series of meta-analyses.
      Consistent with previous data, our findings suggest that combination therapy is more effective than use of monotherapy for treatment of ulcers.
      We identified 73 patients with no reported history of NSAID use and negative diagnostic results for Helicobacter pylori, otherwise known as idiopathic ulcers. Bleeding peptic ulcers may interfere with Helicobacter pylori diagnostic testing, therefore false negatives for Helicobacter pylori must be considered as a possible etiology, as well as unreported use of NSAIDs and aspirin.
      • Ootani H.
      • Iwakiri R.
      • Shimoda R.
      • et al.
      Role of Helicobacter pylori infection and nonsteroidal anti-inflammatory drug use in bleeding peptic ulcers in Japan.
      • Gisbert J.P.
      • Calvet X.
      Review article: Helicobacter pylori-negative duodenal ulcer disease.
      • Niv Y.
      H. pylori/NSAID-negative peptic ulcer—the mucin theory.
      Causes of true idiopathic ulcers include older age, tobacco use, Zolinger Ellison Syndrome, systemic mastocytosis, or other underlying systemic diseases.
      • Boltin D.
      • Halpern M.
      • Levi Z.
      • et al.
      Gastric mucin expression in Helicobacter pylori-related, nonsteroidal anti-inflammatory drug-related and idiopathic ulcers.
      • Niv Y.
      H. pylori/NSAID-negative peptic ulcer—the mucin theory.
      Of the 73 patients identified with idiopathic ulcers, the presence of these conditions was not found or is unknown. This group had high Charlson scores and a large number of comorbidities, as well as high ASA and ER Rockall scores, which is consistent with underlying conditions likely contributing to poorer outcomes. This group had the longest median length of hospital stay (7 days) and greatest incidence of rebleeding (18%). Of note, differences in mortality were not observed among the 4 groups, perhaps related to our sample size. We speculate that inclusion of very large numbers of patients may demonstrate differences in mortality. Nonetheless, our findings are consistent with other literature examining bleeding idiopathic peptic ulcers that have reported more adverse outcomes for this group.
      • Wong G.L.
      • Wong V.W.
      • Chan Y.
      • et al.
      High incidence of mortality and recurrent bleeding in patients with Helicobacter pylori-negative idiopathic bleeding ulcers.
      • Niv Y.
      H. pylori/NSAID-negative peptic ulcer—the mucin theory.
      • Wong G.L.
      • Au K.W.
      • Lo A.O.
      • et al.
      Gastroprotective therapy does not improve outcomes of patients with Helicobacter pylori-negative idiopathic bleeding ulcers.
      There is a robust body of literature examining clinical features and outcomes in patients with ulcer bleeding from Asian, European, and South American populations.
      • Sung J.J.
      Management of nonsteroidal anti-inflammatory drug-related peptic ulcer bleeding.
      • Gisbert J.P.
      • Calvet X.
      • Cosme A.
      • et al.
      Long-term follow-up of 1,000 patients cured of Helicobacter pylori infection following an episode of peptic ulcer bleeding.
      • Fujinami H.
      • Kudo T.
      • Hosokawsa A.
      • et al.
      A study of the changes in the cause of peptic ulcer bleeding.
      • Sostres C.
      • Lanas A.
      Epidemiology and demographics of upper gastrointestinal bleeding: prevalence, incidence, and mortality.
      • Liu N.J.
      • Lee C.S.
      • Tang J.H.
      • et al.
      Outcomes of bleeding peptic ulcers: a prospective study.
      • Ootani H.
      • Iwakiri R.
      • Shimoda R.
      • et al.
      Role of Helicobacter pylori infection and nonsteroidal anti-inflammatory drug use in bleeding peptic ulcers in Japan.
      • Zapata-Colindres J.C.
      • Zepeda-Gomez S.
      • Montano-Loza A.
      • Vazquez-Ballesteros E.
      • de Jesus Villalobos J.
      • Valdovinos-Andraca F.
      The association of Helicobacter pylori infection and nonsteroidal anti-inflammatory drugs in peptic ulcer disease.
      • Chan H.L.
      • Wu J.C.
      • Chan F.K.
      • et al.
      Is non-Helicobacter pylori, non-NSAID peptic ulcer a common cause of upper GI bleeding? A prospective study of 977 patients.
      • Sung J.J.
      • Kuipers E.J.
      • El-Serag H.B.
      Systematic review: the global incidence and prevalence of peptic ulcer disease.
      • Gisbert J.P.
      • Calvet X.
      Review article: Helicobacter pylori-negative duodenal ulcer disease.
      • Wong G.L.
      • Au K.W.
      • Lo A.O.
      • et al.
      Gastroprotective therapy does not improve outcomes of patients with Helicobacter pylori-negative idiopathic bleeding ulcers.
      • Kamada T.
      • Hata J.
      • Kusunoki H.
      • et al.
      Endoscopic characteristics and Helicobacter pylori infection in NSAID-associated gastric ulcer.
      • Sung J.J.
      • Tsoi K.K.
      • Ma T.K.
      • Yung M.Y.
      • Lau J.Y.
      • Chiu P.W.
      Causes of mortality in patients with peptic ulcer bleeding: a prospective cohort study of 10,428 cases.
      However, there are very few data on this subject in US populations. Conclusions drawn about the etiology, incidence, severity, treatment, and outcomes of peptic ulcer bleeding in previous studies may or may not be generalizable to an American population due to variations in the epidemiology of Helicobacter pylori and physician practice patterns worldwide.
      • Goh K.L.
      • Chan W.K.
      • Shiota S.
      • Yamaoka Y.
      Epidemiology of Helicobacter pylori infection and public health implications.
      • Malaty H.M.
      Epidemiology of Helicobacter pylori infection.
      • Tang S.J.
      • Lee S.Y.
      • Hynan L.S.
      • et al.
      Endoscopic hemostasis in nonvariceal upper gastrointestinal bleeding: comparison of physician practice in the East and the West.
      A Greek study comparing bleeding duodenal ulcers with and without Helicobacter pylori shows those patients not infected with Helicobacter pylori to have more severe index bleeding and higher rebleeding rates.
      • Adamopoulos A.B.
      • Efstathiou S.P.
      • Tsioulos D.I.
      • et al.
      Bleeding duodenal ulcer: comparison between Helicobacter pylori positive and Helicobacter pylori negative bleeders.
      Results from Asian studies have shown true idiopathic ulcers to be at high risk of recurrent bleeding and mortality.
      • Wong G.L.
      • Wong V.W.
      • Chan Y.
      • et al.
      High incidence of mortality and recurrent bleeding in patients with Helicobacter pylori-negative idiopathic bleeding ulcers.
      • Liu N.J.
      • Lee C.S.
      • Tang J.H.
      • et al.
      Outcomes of bleeding peptic ulcers: a prospective study.
      Thus, our findings are in agreement with these studies, but differ from a German study that reported similar rebleeding rates in peptic ulcers with and without Helicobacter pylori.
      • Schilling D.
      • Demel A.
      • Nusse T.
      • Weidmann E.
      • Riemann J.F.
      Helicobacter pylori infection does not affect the early rebleeding rate in patients with peptic ulcer bleeding after successful endoscopic hemostasis: a prospective single-center trial.
      Thus, an important advantage with the current study is the analysis of a large and heterogeneous US population; implicit in this point is that the data are likely to be generalizable to Western populations.
      We recognize strengths and limitations of our study. First, our study was single centered, and this could limit its generalizability. However, we would emphasize that we examined a very heterogeneous population of patient ethnicities, and this point is likely a strength. Second, as highlighted above, our sample size may not have been large enough to detect differences in mortality, and we cannot exclude the possibility that mortality is increased in patients with Helicobacter pylori and NSAID negative ulcers. Next, a number of patients were excluded due to lack of diagnostic testing for Helicobacter pylori. Although data were collected prospectively, we did not prescribe specific Helicobacter pylori diagnostic testing. While we doubt selection bias on this basis, we cannot exclude the possibility that patients in one group or the other were selectively not evaluated.
      In summary, our findings indicate that patients with Helicobacter pylori-mediated ulcers have different clinical features and outcomes than patients with NSAID or idiopathic ulcers. Specifically, patients with Helicobacter pylori have a more benign clinical course and more favorable outcomes with peptic ulcer bleeding. These findings are significant because they suggest that care for patients with upper gastrointestinal bleeding due to peptic ulcer disease may be triaged, in part, by ascertainment of Helicobacter pylori status. We suggest that all ulcer disease patients have early and aggressive testing for Helicobacter pylori. Patients with Helicobacter pylori can be started on standard-of-care triple antibiotic therapy to facilitate ulcer healing, thus decreasing recurrent ulcer complications.
      • Laine L.
      • Jensen D.M.
      Management of patients with ulcer bleeding.
      • Gisbert J.P.
      • Calvet X.
      • Cosme A.
      • et al.
      Long-term follow-up of 1,000 patients cured of Helicobacter pylori infection following an episode of peptic ulcer bleeding.
      Helicobacter pylori-negative ulcers are associated with poor outcomes regardless of use of NSAIDs and therefore, patients without Helicobacter pylori should be supported medically as appropriate to improve their underlying medical status, and should be treated as aggressively as possible with proton pump inhibitors.

      Appendix

      Supplemental Table 1Comparison of Groups Based on Helicobacter pylori Status
      H. pylori (−)

      n = 216

      n, Mean (%, SD)
      H. pylori (+)

      n = 228

      n, Mean (%, SD)
      P Value
      NSAIDS or aspirin143 (66%)125 (55%).018
      Charlson Index2.6 ± 2.61.9 ± 2.3.003
      Charlson Age-adjusted Index3.5 ± 3.12.6 ± 2.9.001
      Number of comorbidities2.2 ± 1.61.8 ± 1.5.006
      ICU at admission75 (35%)50 (22%).003
      Admission vitals
       Systolic blood pressure123 ± 28126 ± 28.237
       Diastolic blood pressure68 ± 1872 ± 16.012
       Pulse93 ± 2293 ± 20.736
      Rockall (ER)2.89 ± 1.22.51 ± 1.3.002
      ASA score2.55 ± 0.682.27 ± 0.56<.001
      Laboratory values
       Hematocrit (%)28.4 ± 8.728.7 ± 8.9.710
       Platelets (×109/L)256 ± 133243 ± 107.247
       INR1.4 ± 1.191.2 ± 0.40.013
       BUN (mg/dL)39 ± 3334 ± 27.05
      Length of stay (mean, days)11.4 ± 21.76 ± 8.5<.001
      Length of stay (median, days)5.5 (1, 238)3 (1, 62)<.001
      Rebleed within 30 days25 (11%)11 (5%).009
      Mortality within 30 days9 (4%)8 (4%).718
      ASA = American Society of Anesthesiologists; BUN = blood urea nitrogen; ICU = intensive care unit; INR = international normalized ratio; NSAID = nonsteroidal anti-inflammatory drug.
      Supplemental Table 2Comparison of Gastric and Duodenal Ulcers
      Gastric Ulcer

      n = 245
      Duodenal Ulcer

      n = 164
      P Value
      Helicobacter pylori.014
       Positive116 (47%)98 (60%)
       Negative129 (53%)66 (40%)
      NSAID156 (64%)90 (55%).080
      Stigmata present91 (37%)74 (45%).107
       Active bleeding23 (25%)26 (35%)
       Visible vessel46 (51%)37 (50%)
       Adherent clot22 (24%)11 (15%)
      Stigmata absent154 (63%)90 (55%)
       Pigmented spot19 (12%)9 (10%)
       Clean base135 (88%)81 (90%)
      Endoscopic therapy91 (37%)73 (45%).136
       Epinephrine injection11 (12%)8 (11%)
       Thermal coagulation4 (5%)5 (7%)
       Hemoclip(s)11 (12%)5 (7%)
       Combination therapy65 (71%)55 (75%)
      NSAID = nonsteroidal anti-inflammatory drug.

      References

        • Gralnek I.M.
        • Barkun A.N.
        • Bardou M.
        Management of acute bleeding from a peptic ulcer.
        N Engl J Med. 2008; 359: 928-937
        • Laine L.
        • Jensen D.M.
        Management of patients with ulcer bleeding.
        Am J Gastroenterol. 2012; 107: 345-360
        • Sung J.J.
        Management of nonsteroidal anti-inflammatory drug-related peptic ulcer bleeding.
        Am J Med. 2001; 110: 29S-32S
        • Balderas V.
        • Bhore R.
        • Lara L.F.
        • Spesivtseva J.
        • Rockey D.C.
        The hematocrit level in upper gastrointestinal hemorrhage: safety of endoscopy and outcomes.
        Am J Med. 2011; 124: 970-976
        • Lau J.Y.
        • Sung J.
        • Hill C.
        • Henderson C.
        • Howden C.W.
        • Metz D.C.
        Systematic review of the epidemiology of complicated peptic ulcer disease: incidence, recurrence, risk factors and mortality.
        Digestion. 2011; 84: 102-113
        • van Leerdam M.E.
        Epidemiology of acute upper gastrointestinal bleeding.
        Best Pract Res Clin Gastroenterol. 2008; 22: 209-224
        • Wong G.L.
        • Wong V.W.
        • Chan Y.
        • et al.
        High incidence of mortality and recurrent bleeding in patients with Helicobacter pylori-negative idiopathic bleeding ulcers.
        Gastroenterology. 2009; 137: 525-531
        • Gisbert J.P.
        • Calvet X.
        • Cosme A.
        • et al.
        Long-term follow-up of 1,000 patients cured of Helicobacter pylori infection following an episode of peptic ulcer bleeding.
        Am J Gastroenterol. 2012; 107: 1197-1204
        • Fujinami H.
        • Kudo T.
        • Hosokawsa A.
        • et al.
        A study of the changes in the cause of peptic ulcer bleeding.
        World J Gastrointest Endosc. 2012; 4: 323-327
        • Lee I.
        • Cryer B.
        Epidemiology and role of nonsteroidal antiinflammatory drugs in causing gastrointestinal bleeding.
        Gastrointest Endosc Clin N Am. 2011; 21: 597-612
        • Sostres C.
        • Lanas A.
        Epidemiology and demographics of upper gastrointestinal bleeding: prevalence, incidence, and mortality.
        Gastrointest Endosc Clin N Am. 2011; 21: 567-581
        • Liu N.J.
        • Lee C.S.
        • Tang J.H.
        • et al.
        Outcomes of bleeding peptic ulcers: a prospective study.
        J Gastroenterol Hepatol. 2008; 23: e340-e347
        • Ootani H.
        • Iwakiri R.
        • Shimoda R.
        • et al.
        Role of Helicobacter pylori infection and nonsteroidal anti-inflammatory drug use in bleeding peptic ulcers in Japan.
        J Gastroenterol. 2006; 41: 41-46
        • Zapata-Colindres J.C.
        • Zepeda-Gomez S.
        • Montano-Loza A.
        • Vazquez-Ballesteros E.
        • de Jesus Villalobos J.
        • Valdovinos-Andraca F.
        The association of Helicobacter pylori infection and nonsteroidal anti-inflammatory drugs in peptic ulcer disease.
        Can J Gastroenterol. 2006; 20: 277-280
        • Chan H.L.
        • Wu J.C.
        • Chan F.K.
        • et al.
        Is non-Helicobacter pylori, non-NSAID peptic ulcer a common cause of upper GI bleeding? A prospective study of 977 patients.
        Gastrointest Endosc. 2001; 53: 438-442
        • Sung J.J.
        • Kuipers E.J.
        • El-Serag H.B.
        Systematic review: the global incidence and prevalence of peptic ulcer disease.
        Aliment Pharmacol Ther. 2009; 29: 938-946
        • Boltin D.
        • Halpern M.
        • Levi Z.
        • et al.
        Gastric mucin expression in Helicobacter pylori-related, nonsteroidal anti-inflammatory drug-related and idiopathic ulcers.
        World J Gastroenterol. 2012; 18: 4597-4603
        • Gisbert J.P.
        • Calvet X.
        Review article: Helicobacter pylori-negative duodenal ulcer disease.
        Aliment Pharmacol Ther. 2009; 30: 791-815
        • Adamopoulos A.B.
        • Efstathiou S.P.
        • Tsioulos D.I.
        • et al.
        Bleeding duodenal ulcer: comparison between Helicobacter pylori positive and Helicobacter pylori negative bleeders.
        Dig Liver Dis. 2004; 36: 13-20
        • Niv Y.
        H. pylori/NSAID-negative peptic ulcer—the mucin theory.
        Med Hypotheses. 2010; 75: 433-435
        • Goh K.L.
        • Chan W.K.
        • Shiota S.
        • Yamaoka Y.
        Epidemiology of Helicobacter pylori infection and public health implications.
        Helicobacter. 2011; 16: 1-9
        • Malaty H.M.
        Epidemiology of Helicobacter pylori infection.
        Best Pract Res Clin Gastroenterol. 2007; 21: 205-214
        • Tognoni G.
        The new Helsinki declaration.
        Assist Inferm Ric. 2009; 28 ([Italian]): 49-51
        • Rockey D.C.
        Gastrointestinal bleeding.
        Gastroenterol Clin North Am. 2005; 34: 581-588
        • Barkun A.N.
        • Martel M.
        • Toubouti Y.
        • Rahme E.
        • Bardou M.
        Endoscopic hemostasis in peptic ulcer bleeding for patients with high-risk lesions: a series of meta-analyses.
        Gastrointest Endosc. 2009; 69: 786-799
        • Charlson M.E.
        • Pompei P.
        • Ales K.L.
        • MacKenzie C.R.
        A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.
        J Chronic Dis. 1987; 40: 373-383
        • Malfertheiner P.
        • Megraud F.
        • O'Morain C.A.
        • et al.
        Management of Helicobacter pylori infection—the Maastricht IV/ Florence Consensus Report.
        Gut. 2012; 61: 646-664
        • Graham D.Y.
        • Fischbach L.
        Helicobacter pylori treatment in the era of increasing antibiotic resistance.
        Gut. 2010; 59: 1143-1153
        • Wong G.L.
        • Au K.W.
        • Lo A.O.
        • et al.
        Gastroprotective therapy does not improve outcomes of patients with Helicobacter pylori-negative idiopathic bleeding ulcers.
        Clin Gastroenterol Hepatol. 2012; 10: 1124-1129
        • Kamada T.
        • Hata J.
        • Kusunoki H.
        • et al.
        Endoscopic characteristics and Helicobacter pylori infection in NSAID-associated gastric ulcer.
        J Gastroenterol Hepatol. 2006; 21: 98-102
        • Sung J.J.
        • Tsoi K.K.
        • Ma T.K.
        • Yung M.Y.
        • Lau J.Y.
        • Chiu P.W.
        Causes of mortality in patients with peptic ulcer bleeding: a prospective cohort study of 10,428 cases.
        Am J Gastroenterol. 2010; 105: 84-89
        • Tang S.J.
        • Lee S.Y.
        • Hynan L.S.
        • et al.
        Endoscopic hemostasis in nonvariceal upper gastrointestinal bleeding: comparison of physician practice in the East and the West.
        Dig Dis Sci. 2009; 54: 2418-2426
        • Schilling D.
        • Demel A.
        • Nusse T.
        • Weidmann E.
        • Riemann J.F.
        Helicobacter pylori infection does not affect the early rebleeding rate in patients with peptic ulcer bleeding after successful endoscopic hemostasis: a prospective single-center trial.
        Endoscopy. 2003; 35: 393-396