Helicobacter pylori testing in the primary care setting:: which diagnostic test should be used?

Aim: To identify the most accurate and efficient test for diagnosing Helicobacter pylori infection in primary care patients. Study design: A whole blood test, an ELISA, and carbon(13) urea breath test (CUBT) were evaluated in a primary care setting and validated against two different gold standards that used gastric biopsies. Population: Primary care patients who had dyspeptic complaints lasting at least 2 weeks and were referred for endoscopy. Outcomes measured: Positive and negative predictive values, sensitivity and specificity were determined for all three noninvasive H. pylori tests. Results: Data from the three non-invasive H. pylori tests were available for 136 primary care dyspeptic patients referred for endoscopy. They were compared with data from the gold standards. The positive predictive value of the whole blood test was in the range 71-75%, the ELISA 83-86%, and the CUBT 88-92%, while the negative predictive values were in the ranges 72-77%, 96-100%, and 95-98%, respectively. The sensitivity of the whole blood test was in the range 36-42%, the ELISA 93-100%, and the CUBT 92-97%, while the specificities were in the ranges 92-93%, 90-91% and 93-95%, respectively. The positive predictive value of the ELISA dropped significantly at lower H. pylori infection rates. Discussion: Both the ELISA and CUBT are effective in the primary care setting, while the whole blood tests produces inferior results. ELISA might, however, be less suitable for detecting H. pylori infection in a population with a low rate of infection.