Evaluation of the Ogawa-Kudoh method for tuberculosis isolation in two health units in Mozambique

Background: Mozambique is among the highest tuberculosis, tuberculosis–HIV and multidrug-resistant-tuberculosis burden countries. Although molecular technologies are available in-country, mycobacterial isolation through culture remains an important tool for tuberculosis diagnostics and drug susceptibility testing. Objective: We evaluated the use of the Ogawa-Kudoh (OK) mycobacterial culture, a simple technique, to isolate Mycobacterium tuberculosis in two health units, in Maputo City, Mozambique. Methods: From May to December 2014, 122 patient samples were collected in Chamanculo General Hospital and Polana Caniço General Hospital. The specimens were first tested in the health units using the OK method and afterwards shipped to the National Tuberculosis Reference Laboratory for mycobacterial culture using the NALC-NaOH-Citrate (NALC) decontamination method followed by inoculation in Lowenstein Jensen (LJ) solid media as the reference standard. Results: Among 107 samples with valid results, 98 (91.6%) had concordant results in both methods; 9 (8.4%) had discordant results. The contamination rate was 4.1% (5/122) for the OK and 9.0% (11/122) for the NALC/LJ methods. The sensitivity of OK was 80% (95% confident interval [CI]: 51.4–94.7) and the specificity was 94% (95% CI: 85.8–97.3). The degree of agreement between both methods was moderate (Kappa: 0.68; 95% CI: 0.48–0.89). Conclusion: The OK method showed satisfactory sensitivity and specificity. The method also had a lower contamination rate when compared to the NALC/LJ. Similar to other studies in resource-limited settings, our findings showed that the OK method can effectively be implemented in settings with limited laboratory capacity to isolate tuberculosis bacteria by culture for further testing.