OBJECTIVES: To establish the sensitivity, specificity, positive predictive value and negative predictive value of serology (IgG ELISA) as an alternative diagnostic method for Helicobacter pylori infection in patients with gastro-duodenal peptic ulcer and digestive hemorrhage. The diagnosis of Helicobacter pylori infection in these patients is difficult due to the low sensitivity of invasive tests and the need to discontinue treatment with proton pump inhibitors to perform a breath test with urea 13C or the detection of Helicobacter pylori antigens in feces. PATIENTS AND METHODS: We included 214 patients (164 men and 50 women) with an average age of 58 +/- 15 years, who were admitted to hospital due to upper gastro-intestinal bleeding caused by a gastro-duodenal peptic ulcer. The presence of Helicobacter pylori was established by means of gastric biopsy (fast urease test histology and/or culture) and a breath test with 13C-labeled urea. Serology was performed with the ELISA method (Pyloriset EIA-G by Orion Diagnostica). Positive Helicobacter pylori infection was accepted with any positive invasive method or breath test, and no infection was established if all invasive tests performed and the breath test with 13C-labeled urea were negative. We calculated the sensitivity, specificity, positive predictive value and negative predictive value of serology in the global series and in different subgroups of patients according to age (> 60 and < 40 years), recent exposure to non-steroidal anti-inflammatory drugs, type of endoscopic wound and history of gastro-duodenal peptic ulcer. RESULTS: 192 patients (89.7%) showed infection due to Helicobacter pylori. In the global series (n = 214) we obtained a sensitivity, specificity, positive predictive value and negative predictive value of 87.5, 54.5, 94.3 and 33.3%, respectively. Specificity was greater in the group not exposed to non-steroidal anti-inflammatory drugs (n = 110) as compared to the exposed group (n = 104), in the < 40 year old group (n = 28) with respect to the > 60 years group (n = 105), in the duodenal ulcer group (n = 141) with respect to the gastric ulcer group (n = 59), and in the group with a history of gastro-duodenal peptic ulcers (n = 92) as compared to the group without any of these past events (n = 122); nevertheless, no significant statistics were reached. CONCLUSIONS: IgG (ELISA) serology shows low specificity and a low negative predictive value in the diagnosis of Helicobacter pylori infection in patients with gastro-intestinal bleeding due to gastro-duodenal peptic ulcer. The diagnostic value of serology did not improve significantly when age, recent exposure to non-steroidal anti-inflammatory drugs, type of endoscopic wound or history of gastro-duodenal peptic ulcer was taken into consideration. We may consider that serology is not a good diagnostic method for the detection of Helicobacter pylori in patients with digestive hemorrhage caused by gastro-duodenal ulcer.
OBJECTIVES: To establish the sensitivity, specificity, positive predictive value and negative predictive value of serology (IgG ELISA) as an alternative diagnostic method for Helicobacter pylori infection in patients with gastro-duodenal peptic ulcer and digestive hemorrhage. The diagnosis of Helicobacter pylori infection in these patients is difficult due to the low sensitivity of invasive tests and the need to discontinue treatment with proton pump inhibitors to perform a breath test with urea 13C or the detection of Helicobacter pylori antigens in feces. PATIENTS AND METHODS: We included 214 patients (164 men and 50 women) with an average age of 58 +/- 15 years, who were admitted to hospital due to upper gastro-intestinal bleeding caused by a gastro-duodenal peptic ulcer. The presence of Helicobacter pylori was established by means of gastric biopsy (fast urease test histology and/or culture) and a breath test with 13C-labeled urea. Serology was performed with the ELISA method (Pyloriset EIA-G by Orion Diagnostica). Positive Helicobacter pylori infection was accepted with any positive invasive method or breath test, and no infection was established if all invasive tests performed and the breath test with 13C-labeled urea were negative. We calculated the sensitivity, specificity, positive predictive value and negative predictive value of serology in the global series and in different subgroups of patients according to age (> 60 and < 40 years), recent exposure to non-steroidal anti-inflammatory drugs, type of endoscopic wound and history of gastro-duodenal peptic ulcer. RESULTS: 192 patients (89.7%) showed infection due to Helicobacter pylori. In the global series (n = 214) we obtained a sensitivity, specificity, positive predictive value and negative predictive value of 87.5, 54.5, 94.3 and 33.3%, respectively. Specificity was greater in the group not exposed to non-steroidal anti-inflammatory drugs (n = 110) as compared to the exposed group (n = 104), in the < 40 year old group (n = 28) with respect to the > 60 years group (n = 105), in the duodenal ulcer group (n = 141) with respect to the gastric ulcer group (n = 59), and in the group with a history of gastro-duodenal peptic ulcers (n = 92) as compared to the group without any of these past events (n = 122); nevertheless, no significant statistics were reached. CONCLUSIONS: IgG (ELISA) serology shows low specificity and a low negative predictive value in the diagnosis of Helicobacter pylori infection in patients with gastro-intestinal bleeding due to gastro-duodenal peptic ulcer. The diagnostic value of serology did not improve significantly when age, recent exposure to non-steroidal anti-inflammatory drugs, type of endoscopic wound or history of gastro-duodenal peptic ulcer was taken into consideration. We may consider that serology is not a good diagnostic method for the detection of Helicobacter pylori in patients with digestive hemorrhage caused by gastro-duodenal ulcer.