Poster Presentations
Session Title: Category 2b. CIRRHOSIS AND ITS COMPLICATIONS: b. CLINICAL ASPECTS
Presentation Date: Apr 15, 2010
BEDSIDE REAGENT STRIP ANALYSIS OF ASCITES CAN RELIABLY RULE OUT SPONTANEOUS BACTERIAL PERITONITIS AND IS A COST-EFFECTIVE STRATEGY
J.J. Kuiper*, H.R. van Buuren, R.A. de Man
Department of Gastroenterology and Hepatology, Erasmus Medical Center, Rotterdam, The Netherlands. *j.j.kuiper@erasmusmc.nl
Background/aims: Spontaneous bacterial peritonitis (SBP) is a serious complication in patients with decompensated cirrhosis of the liver. Timely diagnosis of SBP is essential to prevent complications like sepsis and hepatorenal syndrome. The golden standard for diagnosis of SBP is an ascitic polymorphonuclear (PMN) count ≥ 250/mm3, but this test is time-consuming and is not uniformly available, especially during after-office hours. Reagent strip testing of ascites for diagnosing SBP may be a promising bedside diagnostic tool. The aim of this study was to assess the reliability of reagent strip testing for diagnosing or excluding SBP and to assess its clinical implication.
Methods: Consecutive samples of ascites obtained from patients with cirrhosis and ascites were included. All samples were tested simultaneously for ascitic PMN-count using standard laboratory methodology as well as reagent strips (Combur-10®, Roche Diagnostics). The reagent strip had 4 outcomes: 0, 25, 100, 500 leukocytes/µl. The strips were independently analyzed visually by a dedicated observer and at the same time by an electronic analyzer (Urisys 1100®, Roche Diagnostics).
Results: 157 samples of ascites were analyzed. According to the laboratory PMN-count SBP was diagnosed in 12 (7.6%) samples. Automated analysis of these samples indicated the presence of ≥100 leukocytes/µl in all cases. Using this cut-off value resulted in the diagnostic test characteristics of the reagent strip as depicted in the table.
| Sensitivity | Specificity | Positive predictive value | Negative predictive value | |
| Visual | 75% (95% CI 42.8-93.3) | 98.6% (95% CI 94.6-99.7) | 81.8% (95% CI 47.8-96.8) | 97.9(95% CI 93.6-99.5) |
| Automated | 100% (95% CI 69.9-100) | 93.1 (95% CI 87.3-96.5) | 54.5% (95% CI 32.6-74.9) | 100%(95% CI96.6-100) |
Conclusion: In this population with a relatively low prevalence of SPB a negative reagent strip result reliably ruled out SBP. A positive test should be followed by conventional work-up (PMN-count and culture). Using an automated analyzer was not superior to standard visual reading of the strips. Using reagent strips would have obviated the need for more laborious and expensive laboratory studies in more than 90 cases. This suggests that using this simple and quick bedside test may result in a significant lowering of costs.
Back