A multicentre evaluation of the laser assisted ratio analyser (LARA): a novel device for measurement of 13 CO 2 in the 13 C-urea breath test for the detection of Helicobacter pylori infection D. R. CAVE*, S. VELDHUYZEN VAN ZANTEN  , E. CARTER à , E. F. HALPERN§, S. KLEIN ± , C. PRATHER**, M. STOLTE   , L. LAINE àà *St Elizabeth's Hospital Medical Center of Boston, Brighton MA;  QE-11 Health Sciences Center, Halifax, Nova Scotia, Canada; àPasco Washington, USA; §Center for Imaging and Pharmaceutical Research, Massachusetts General Hospital, Boston MA; ±Washington University School of Medicine, St Louis MO; **Mayo Clinic. Rochester MN; ààUSC School of Medicine, Los Angeles CA, USA; and   Institute of Pathology, Klinicum Bayreuth, Bayreuth, Germany Accepted for publication 8 January 1999 INTRODUCTION H. pylori is a common infection of the gastric mucosa and is the cause of chronic active gastritis and the majority of peptic ulcers. 1 It is associated strongly with the development of gastric neoplasia. 2, 3 SUMMARY Background: The laser assisted ratio analyser (LARA) was developed as a novel device to measure 13 CO 2 in the urea breath test for the detection of H. pylori infection. The analyser was tested in a prospective multicentre study in 444 patients in North America (Phase 1) followed by second study involving 160 patients (Phase 2). Methods: Patients undergoing endoscopy for clinical indications had antral and gastric biopsies taken for histological examination, culture and CLO test. One hour after endoscopy, a baseline breath sample was obtained, 100 mg of 13 C-urea were ingested and breath samples were obtained at 30 and 60 min post ingestion. Data obtained with the LARA were compared with the results of culture, rapid urease testing and central pathology in two different combinations {reference standards}. The study was conducted in two phases: in Phase 2, a modi®cation was made to the LARA that improved the removal of water vapour from the breath sample. Results: In Phase I, data from 331 patients were analysed using a cut off of (delta) 7.8  0.8, the sensitivity of the method was 91.7% and the speci®city was 86.5%, using the reference standard of 2 of 3 tests (CLO, culture or histology) being positive. Positive and negative predictive values were, respectively, 85.2% and 92.5%. In Phase 2 of the study, 160 patients were enrolled and 141 patients were analysed using the same standards. We used the same reference standards but with a cut off of (delta) 6.1  0.6. The sensitivity and speci®city increased to 96.8% and 98.6%, respectively. Positive and negative predictive values were, respec- tively, 98.4% and 97.3%. The detection rates for H. pylori were similar in patients with peptic ulcer or H. pylori associated gastritis. Conclusions: The LARA provides an accurate non- invasive means of detecting 13 CO 2 in the 13 C-urea breath test for H. pylori in a multicentre clinical environment that compares well with invasive `gold standard' methods. Correspondence to: Dr D. R. Cave, St Elizabeth's Medical Center of Boston, 736 Cambridge Street, Brighton MA O2135, USA. E-mail:dcave@pol.net Aliment Pharmacol Ther 1999; 13: 747±752. Ó 1999 Blackwell Science Ltd 747