The influence of size or number of biopsies on rapid urease test results: a prospective evaluation.

BACKGROUND

The optimal number or size of endoscopic biopsies for use in rapid urease testing has not been established. Postulating that increasing the amount of tissue sampled would improve diagnostic yield and hasten development of a positive test, we compared urease testing with one regular biopsy, two regular biopsies, and one "jumbo" forceps biopsy.

METHODS

One hundred fifty patients undergoing endoscopy had three sets of prepyloric biopsies placed in a CLOtest: one regular forceps biopsy, two regular forceps biopsy, and one large-channel jumbo forceps biopsy. Biopsies were then taken for two independent histologic examinations. Disagreements were resolved by another examiner.

RESULTS

Eighty-nine patients (59%) had Helicobacter pylori infection by histology; interobserver agreement was 90% with kappa = 0.78. The mean time to a positive test was 5.3 +/- 0.9 hours for one regular biopsy, 3.2 +/- 0.7 hours for two regular biopsies, and 3.8 +/- 0.8 hours for one jumbo biopsy (p < 0.01 for two regular, one jumbo vs. one regular biopsy). Compared to one regular biopsy, the urease test was positive at least 30 minutes earlier in 56% of the patients with two regular biopsies and 54% with one jumbo biopsy. Sensitivities for one regular versus two regular biopsies were 1 hour, 19% versus 33% (p = 0.059); 2 hours, 38% versus 49% (p = 0.17); 3 hours, 48% versus 60% (p = 0.18); and 24 hours, 75% versus 79% (p > 0.20).

CONCLUSIONS

Doubling the amount of tissue in the CLOtest hastens the development of a positive test by approximately 1 1/2 to 2 hours; tests become positive at least 30 minutes earlier in over 50% of the patients. Low cost, ease, and excellent specificity make the rapid urease test a valuable diagnostic tool. Nevertheless, if used as a "rapid" diagnostic test (read within 3 hours of biopsy), it is associated with a false negative rate of approximately 40%.