| dc.description.abstract | Introduction: Tuberculosis remains a big burden in sub - Saharan Africa. Despite much progress on how TB infection occurs, its physiology and immunological interactions, definite diagnostic method is still out of reach. It is in this context that we needed to evaluate the usefulness of blood Buffy coat AFB ZN stain.
Methods: In a cross - section study conducted at KNH and Mbagathi hospitals' TB clinics between August and October, 2010, we evaluated the outcome of sputum and blood Buffy coat AFB ZN Stain smears. Patients with clinical symptoms and CXR suggestive of tuberculosis were eligible. 114 patients were recruited. A questionnaire was completed and complete blood count was also done for each patient. The data collected was entered into MS Excel and later analyzed using STATA 10 and SPSS.
Results: Eleven (9.6%) of all patients had positive AFB Buffy coat smear. 40(35%) had positive AFB sputum smear. All Buffy coat positive smears also were sputum smear positive. Significant factors for outcome of Buffy coat smear were age (p=0.0322), sputum smear (p=<0.0001) and HIV status (p=0.01). In two- way table, compared to sputum smear, Buffy coat smear had sensitivity of 27.5%, specificity of 100%, positive predictive value of 100%, and negative predictive value of 71.8%.
Conclusion: Even though Buffy coat smear could diagnose AFB (100% specificity), its sensitivity was low at 27.5% compared to sputum smear. Therefore, it cannot be used with confidence as an alternative for sputum for it will miss most positive sputum smear, It can be used to rule in tuberculosis due to high predictive value. To increase it sensitivity we need to evaluate molecular methods and culture involving blood Buffy coat. | en_US |
