THE zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA AMERICAN JOURNAL OFGASTROENTEROLCGY Vol. 93, No. 1, 1998 Copyright 0 1998 by Am. Coil. of Gastroenterology ISSN 0002-9270/98/$19.00 Published by Elsevier Science Inc. PI1 SOOO2-9270(97WOO3-8 The Chemtrak Hp Chek Fingerstick Whole Blood Serology Test for the Detection of zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED Helicobacter pyZori Infection W. D. Chey, M.D., U. K. Murthy, M.D., W. Linscheer, M.D., C. Barish, M.D., D. Riff, M.D., H. Rubin, M.D., M. Safdi, M.D., H. Schwartz, M.D., U. Shah, M.D., L. Wruble, M.D., and H. M. T. El-Zimaity, M.D. University of Michigan, Ann Arbor, Michigan; VA Medical Center, Syracuse, New York; Raleigh, North Carolina; Anaheim, Cul~f~wnicr;Beverly Hills, California; Cincinnati, Ohio: Miami, Florida; Leonardstown, Maryland; Memphis, Tennessee; and VA Medical Center, Houston, Texas Objective: To evaluate a new whole blood serology test (Hp Chek; ChemTrak) that detects IgG antibodies to Helicobacter pylori. Methods: The study was conducted at 10 sites within the United States. Patients undergoing upper endoscopy for dyspepsia were recruited for en- rollment. Those treated for H. pylon’ infection within a year of endoscopy and those who had regularly used proton pump inhibitors, bismuth compounds, or antibi- otics within a month of endoscopy were not eligible. During endoscopy, specimens were obtained from the corpus and antrum for histological examination, which was performed by a single experienced pathologist. The Hp Chek was tested using whole blood and serum. Se- rum was also tested with a reference enzyme-linked immunosorbent assay (ELISA) at a centralized location. Test characteristics for the Hp Chek and ELISA were calculated using histology as the “gold standard.” Re- sults: Two hundred eighty-seven patients (140 women and 147 men; mean age 53 + 6 yr) were enrolled. The Hp Chek was easy to perform and yielded results 9 min after inoculation of the test cassette with whole blood or se- rum. When the Hp Chek used with whole blood was compared with histology as the gold standard, the sen- sitivity was 88%, specificity 85%. positive predictive value 83%, negative predictive value 90%, and percent agreement 86%. There were no statistically significant differences among the results obtained with the Hp Chek using whole blood, the Hp Chek using serum, or refer- ence ELISA. Conclusions: The Hp Chek whole blood serology test was easy to perform and rapid and yielded performance characteristics comparable to those of a reference ELISA or the Hp Chek used with serum. (Am J Gastroenterol 1998;93:16-19. 0 1998 by Am. Coil. of Gastroenterology) INTRODUCTION Diagnostic tests for Helicohacter pylori can be divided into those that require endoscopy with gastric mucosal bi- Received Max 20. 1997; accepted Aq. 24. 1997. 16 opsy (rapid urease testing, histology, culture) and those that can be performed independently of endoscopy (serology, urea breath testing) (1). Serology is currently the most cost-effective means of testing for H. p$ori infection (1). The role of H. pylori infection in the pathogenesis of dys- peptic symptoms in patients with no evidence of peptic ulcer disease remains poorly defined (2). However, preliminary work using decision analytical modeling suggested that it may be cost-effective to treat patients for H. pylori infection without evaluation with upper endoscopy (3). For this rea- son, it is anticipated that serology testing will become in- creasingly important in the management of dyspeptic pa- tients, particularly in the primary care setting (4). The majority of currently available serology tests for H. pylori rely upon the identification of specific immunoglob- ulin G (IgG) antibodies that can be found in saliva (5) and blood (6). Tests for the serodiagnosis of H. pylori in blood can be classified as quantitative (enzyme-linked immu- nosorbent assay [ELBA]) or qualitative. A number of qual- itative tests that are easy to perform and inexpensive are now commercially available (7, 8). These tests were origi- nally developed for use with serum, but recently, several companies have introduced products intended for use with whole blood. Unfortunately, the sensitivity of the current generation of whole blood serology tests has been inconsis- tent (9-12). We conducted a multicenter trial to evaluate the test characteristics of a new whole blood serology test manu- factured by ChemTrak (Hp Chek, formerly the “Accume- ter”; Sunnyvale. CA). Results obtained with the Hp Chek were compared with histology and with a reference ELBA. MATERIALS AND METHODS Patient population This study was conducted at 10 geographically diverse sites in the United States. Study sites included seven com- munity gastroenterologists’ offices, two university medical centers, and one VA Medical Center. Patients undergoing upper endoscopy for dyspeptic symptoms were recruited for enrollment. Those who had received treatment for H. pylori