Thesis (M Med.(Family Medicine))--University of Limpopo, 2010. / OBJECTIVE: In order to obtain optimal results using sputum smear microscopy for acid-fast bacilli (AFB) detection in the diagnosis of pulmonary tuberculosis (PTB), good quality sputum with an adequate volume of at least 5.0ml is required. An inadequate amount of sputum sample may result cases of active PTB being missed. This study was aimed at showing that a single sputum sample of at least 5.0ml would lower the chances of missing active PTB cases, and increase sputum smear positivity by microscopy thus enabling prompt treatment of PTB, and restricting its transmission. METHODS: An analytical cross sectional study was carried out at Dr George Mukhari Hospital in Pretoria, South Africa. Two sputum samples, one of 5.0ml and the other with 2.0ml were collected from each adult patient suspected of having active PTB. Sputum collection was supervised and patients were given instructions on how to enhance sputum expectoration. Sputum samples were processed using the N-acetyl-L-cysteine (NaLc-NaOH) method and stained with Auramine O. Sputum analysis was done with the aid of fluorescence microscopy. Following microscopy, both the 2.0ml and 5.0ml specimens were sent for culture using the Middlebrook broth medium, and culture results were available after 6-8 weeks. Using the culture results as gold standard, the yield through microscopy of the 2.0ml specimen versus the 5.0ml specimen for each patient were compared and analysed. RESULTS: A Total of 330 sputum samples were analysed of which 77 were found to be culture positive. A sensitivity of 76.6% and specificity of 99.6% was obtained for AFB test in the 5.0ml specimens; while in the 2.0ml samples the sensitivity was 75.3%, with a specificity of 99.2%. The difference in the smear positivity rate of 76.6% obtained using the 5.0ml sputum specimen compared to the 75.3% obtained using the 2.0ml specimen from patients suspected of having TB in this study was statistically insignificant – ascribable to the small sample size. CONCLUSION: In this study, the volume of sputum collected did not determine a better AFB test yield in the diagnosis of pulmonary tuberculosis in patients suspected of having TB.

Tuberculosis (TB) is the most common opportunistic infection in HIV-infected patients and the emergence of drug-resistant tuberculosis (DR-TB) is a growing problem in resource-limited settings (RLS). TB diagnosis in most RLS still depends on smear microscopy for acid-fast bacilli (AFB) while adequate infrastructure for testing drug sensitivity is unavailable. However, molecular diagnostics that detect Mycobacterium tuberculosis (Mtb) DNA and its genetic markers of drug resistance were recently developed. In this thesis I describe the use of a molecular diagnostic, Genotype MTBDRplus, for characterizing DR-TB and patterns of tuberculosis-like infection in two cities in south-west and north-central Nigeria. I found high rates of DR-TB in Nigerian HIV-infected individuals (9.3% for RIF or INH) with significantly different amounts by location (18.18% in south-west vs. 3.91% in north-central Nigeria, p=0.01). RIF resistance, indicative of MDR-TB, was found in 5.52% treatment-naïve patients, far exceeding the WHO predictions (0-4.3%). Furthermore, RIF resistance was genetically distinct, suggesting location-specific transmission of drug resistance (p=0.04). Genotype MTBDRplus correctly identified the drug-resistant samples compared to sequencing in 96.8% of cases. Mtb was confirmed in 56% of patients and was less likely to be found in patients on ART, while controlling for other relevant demographic characteristics (OR 0.29, P=0.02). Only abnormal respiratory findings on auscultation and the direct sputum smear grade greater than 3/100 were significant predictors of Mtb infection (OR 3.28, P=0.03; OR 6.40, p<0.01 respectively). Concentrated sputum smear was not significantly correlated with Mtb infection, except at the highest grades (>2+). Furthermore, in 49% of samples that were not confirmed for Mtb other actinomycetes were found: atypical Mycobacteria (ATM), Rhodococcus spp., Nocardia spp., Corynebacterium spp. I conclude that concentrated sputum AFB smears may misidentify bacteria as Mtb in a subset of HIV-infected patients. These individuals may have a different, even uncharacterized, actinomycete infection in the respiratory tract. Furthermore, total DR-TB in HIV-infection is high and transmission of DR-TB in HIV-infected patients in Nigeria is higher than estimated by the WHO. Molecular diagnostics are a rapid method for identifying Mtb and monitoring DR-TB, and can guide appropriate treatment decisions for respiratory infections in RLS with a high HIV burden.