SETTINGS: Tuberculosis (TB) diagnostic laboratories in Latin America. OBJECTIVES: Evaluation of thin-layer agar (TLA) compared to Löwenstein-Jensen (LJ) culture for the diagnosis of TB. DESIGN: Phase II prospective study in six laboratories. Samples included sputum and extra-pulmonary specimens from patients with a clinical diagnosis of TB. Respiratory samples were decontaminated using NaOH/ NALC; all samples were centrifuged, stained with Ziehl-Neelsen for acid-fast bacilli (AFB), cultured on LJ and TLA and identified according to recommended procedures. Sensitivity and likelihood ratios (LR), growth detection time and contamination rate were calculated for both media. RESULTS: A total of 1118 clinical specimens were studied. Cultures detected Mycobacterium tuberculosis in all AFB-positive samples, whereas for AFB-negative specimens LJ detected 3.2% and TLA 4.4%. Sensitivity was 92.6% (95%CI 87.9-95.9) and 84.7% (95%CI 78.8-89.0) for TLA and LJ, respectively. Positive and negative LRs were similar. Contamination was 5.1% for TLA and 3.0% for LJ. Median time to detection of a positive culture was 11.5 days (95%CI 9.3-15.0) for TLA and 30.5 days (95%CI 26.9-39.0) for LJ (P < 0.0001). CONCLUSION: Difference in the characteristics of the participating laboratories, the disease prevalence and the number and type of specimens processed did not affect the overall performance of TLA as compared to LJ, supporting the robustness of the method and its feasibility in different laboratory settings.
SETTINGS: Tuberculosis (TB) diagnostic laboratories in Latin America. OBJECTIVES: Evaluation of thin-layer agar (TLA) compared to Löwenstein-Jensen (LJ) culture for the diagnosis of TB. DESIGN: Phase II prospective study in six laboratories. Samples included sputum and extra-pulmonary specimens from patients with a clinical diagnosis of TB. Respiratory samples were decontaminated using NaOH/ NALC; all samples were centrifuged, stained with Ziehl-Neelsen for acid-fast bacilli (AFB), cultured on LJ and TLA and identified according to recommended procedures. Sensitivity and likelihood ratios (LR), growth detection time and contamination rate were calculated for both media. RESULTS: A total of 1118 clinical specimens were studied. Cultures detected Mycobacterium tuberculosis in all AFB-positive samples, whereas for AFB-negative specimens LJ detected 3.2% and TLA 4.4%. Sensitivity was 92.6% (95%CI 87.9-95.9) and 84.7% (95%CI 78.8-89.0) for TLA and LJ, respectively. Positive and negative LRs were similar. Contamination was 5.1% for TLA and 3.0% for LJ. Median time to detection of a positive culture was 11.5 days (95%CI 9.3-15.0) for TLA and 30.5 days (95%CI 26.9-39.0) for LJ (P < 0.0001). CONCLUSION: Difference in the characteristics of the participating laboratories, the disease prevalence and the number and type of specimens processed did not affect the overall performance of TLA as compared to LJ, supporting the robustness of the method and its feasibility in different laboratory settings.
Authors: Giovanni Battista Migliori; Alberto Matteelli; Daniela Cirillo; Madhukar Pai Journal: Can J Infect Dis Med Microbiol Date: 2008-03 Impact factor: 2.471
Authors: Luis E Cuevas; Renee Browning; Patrick Bossuyt; Martina Casenghi; Mark F Cotton; Andrea T Cruz; Lori E Dodd; Francis Drobniewski; Marianne Gale; Stephen M Graham; Malgosia Grzemska; Norbert Heinrich; Anneke C Hesseling; Robin Huebner; Patrick Jean-Philippe; Sushil Kumar Kabra; Beate Kampmann; Deborah Lewinsohn; Meijuan Li; Christian Lienhardt; Anna M Mandalakas; Ben J Marais; Heather J Menzies; Grace Montepiedra; Charles Mwansambo; Richard Oberhelman; Paul Palumbo; Estelle Russek-Cohen; David E Shapiro; Betsy Smith; Giselle Soto-Castellares; Jeffrey R Starke; Soumya Swaminathan; Claire Wingfield; Carol Worrell Journal: J Infect Dis Date: 2012-04-03 Impact factor: 5.226
Authors: Helena Huerga; Francis Varaine; Eric Okwaro; Mathieu Bastard; Elisa Ardizzoni; Joseph Sitienei; Jeremiah Chakaya; Maryline Bonnet Journal: PLoS One Date: 2012-12-19 Impact factor: 3.240
Authors: N Alvarez; E Zapata; G I Mejía; T Realpe; P Araque; C Peláez; F Rouzaud; J Robledo Journal: Biomed Res Int Date: 2014-04-28 Impact factor: 3.411