Diagnosis of smear-negative tuberculosis is greatly improved by Xpert MTB/RIF

Giulia Lombardi(University of Bologna), Valentina Di Gregori(University of Bologna), Nicolò Girometti(University of Bologna), Marina Tadolini(University of Bologna), Francesco Bisognin(University of Bologna), Paola Dal Monte(University of Bologna)
PLoS ONE
April 21, 2017
Cited by 73Open Access
Full Text

Abstract

BACKGROUND: Diagnosis of pulmonary (PTB) and extra-pulmonary tuberculosis (EPTB) in smear-negative patients can be difficult. We assessed retrospectively the performance of Xpert MTB/RIF system (Xpert, Cepheid) in diagnosing smear-negative tuberculosis (TB), which represents the most common form of TB in a low incidence setting. METHODS: Performance of Xpert was compared to acid-fast microscopic examination using Ziehl-Neelsen (ZN) stain in patients with culture-confirmed TB. RESULTS: 386 Mycobacterium tuberculosis (MTB) culture-positive samples were detected out of 5170 specimens tested with smear microscopy, Xpert and culture: 323 were both culture- and Xpert-positive, and 63 culture-positive only. Of these, 234 (60.6%) were smear-negative. In addition Xpert detected 40 probable TB cases, based on clinical findings, which were culture-negative. Compared to culture, Xpert showed an overall sensitivity of 83.7% and a specificity of 99.1%; sensitivity was higher for respiratory samples (86.5%) than for non-respiratory samples (76.8%). Xpert sensitivity for smear-negative culture-confirmed TB was 73.1% and was not influenced by TB localization. As sensitivity of microscopy alone was poor (39.4%), Xpert improved both diagnosis of pulmonary TB (Δ = 36.5%) and extra-pulmonary TB (Δ = 63.4%). CONCLUSIONS: Xpert MTB/RIF is a sensitive method for rapid diagnosis of TB compared to the conventional ZN staining. Xpert can serve as a sensitive and time-saving diagnostic method for microbiological diagnosis of smear-negative TB in countries with a low TB prevalence.


Related Papers

No related papers found

Powered by citation graph analysis