Current issues of ACP Journal Club are published in Annals of Internal Medicine


A negative serum enzyme immunoassay test result ruled out Helicobacter pylori infection

ACP J Club. 1999 May-June;130:76. doi:10.7326/ACPJC-1999-130-3-076

Source Citation

Embil JM, Choudhri SH, Smart G, et al. Comparison of salivary and serum enzyme immunoassays for the diagnosis of Helicobacter pylori infection. Can J Infect Dis. 1998 Sep/Oct; 9:277-80.



What are the diagnostic test characteristics of salivary and serum enzyme immunoassays (EIAs) for Helicobacter pylori infection detected by histologic examination?


Blinded comparison of the test results with histologic examination for the diagnosis of H. pylori infection.


A university hospital in Winnipeg, Manitoba, Canada.


104 adults (mean age 51 y, 50% men) who were being evaluated by flexible upper gastrointestinal endoscopy. 93% had symptoms of peptic ulcer disease.

Description of tests and diagnostic standard

Blood and serum samples were taken before endoscopy and were analyzed using commercial EIA kits and processors. Saliva was collected using the Omni-Sal collection device (MML Diagnostic Packaging, Inc., Oregon, U.S.A.) and analyzed using the HeliSal Salivary EIA kit (Axcan Pharma, Quebec, Canada). Serum was analyzed using the Pylori Stat Serum EIA kit (Bio-Whittaker, Walkersville, Maryland, U.S.A.). The technician reading the samples was blinded to the clinical status of patients. Histologic examination was considered to be the diagnostic standard: 2 prepyloric greater curvature biopsies were obtained and were evaluated by an experienced gastrointestinal pathologist who was blinded to clinical and other data.

Main outcome measures

Sensitivity and specificity of the salivary and serum EIAs.

Main results

50 patients (48%) were positive for H. pylori infection; 23 patients (22%) had endoscopically proven peptic ulcer disease (32% of patients who were positive for H. pylori, and 13% who were negative for H. pylori ). Salivary EIA had low sensitivity and specificity. Serum EIA had high sensitivity and low specificity (Table), so that a negative test result effectively ruled out H. pylori infection.


A negative serum enzyme immunoassay test result ruled out Helicobacter pylori infection in adults with symptoms of peptic ulcer disease. Other test characteristics of salivary and serum enzyme immunoassays were not useful for detection of H. pylori infection.

Source of funding: In part, Axcan Pharma (HeliSal test kits).

For correspondence: Dr. C.N. Bernstein, Division of Gastroenterology, GB443 Health Sciences Centre, 820 Sherbrook Street, Winnipeg, Manitoba R3A 1R9, Canada. FAX 204-787-4826.

Table. Sensitivity, specificity, and likelihood ratios (LRs) for the diagnosis of Helicobacter pylori infection done using salivary and serum enzyme immunoassays (EIAs) compared with histologic examination*

Diagnostic test Sensitivity (95% CI) Specificity (CI) +LR -LR
Salivary EIA 66.0% (51.2 to 78.8) 66.7% (52.1 to 79.2) 2.0 0.51
Serum EIA 97.9% (88.9 to 99.9) 48.1% (34.0 to 62.3) 1.9 0.04

*Abbreviations defined in Glossary; LRs and CIs calculated from data in article.


Noninvasive diagnosis of H. pylori is an admirable goal for clinicians, but if tests are to be used to direct therapy, they need to be more than 90% accurate. Sadly, the study by Embil and colleagues and other studies summarized in their report show that the HeliSal salivary IgG test does not provide this accuracy. Sensitivity was 66%, and specificity was 67%. Embil and colleagues discuss the difficulties of salivary antibody tests, and it is possible that pre-endoscopy anxiety reduced salivary flow. The Pylori Stat serum IgG had a sensitivity of 98%, but specificity was only 48% (although it had been 94% in a previous study). As the authors discuss, this may be because antibodies persisted after eradication of H. pylori or because recent therapy gave false-negative results on antral histologic examination, which was the diagnostic standard. During the previous month, 24% of the patients in the group who were negative for H. pylori had taken antibiotics and 67% had taken antiulcer medications. When uninhibited by antibiotics, the germ proceeds proximally from the antrum when acid is suppressed (1). Infection with H. pylori is most accurately diagnosed by using more than one biopsy method (2) or by a urea breath test. Antibiotic treatment and acid suppression need to be stopped before testing for periods that are ill defined but are conventionally considered to be 4 and 2 weeks, respectively (2). The study by Embil and colleagues nicely shows what happens in the real world.

John Calam, MD
Hammersmith HospitalLondon, England, UK


1. Logan RP, Walker MM, Misiewicz JJ, et al. Changes in the intragastric distribution of Helicobacter pylori during treatment with omeprazole. Gut. 1995;36:12-6.

2. Megraud F. How should Helicobacter pylori infection be diagnosed? Gastroenterology. 1997;113:S93-8.