Diagnosis of pulmonary tuberculosis can be made with certainty only if the presence of mycobacteria is demonstrated in culture or smear microscopy of sputum. The present study was undertaken to compare various techniques for demonstration of mycobacteria (AFB) in smears, prepared by different techniques from sputum of patients suspected to be suffering from pulmonary tuberculosis. Results of culture for mycobacteria were taken as gold standard.
Early morning 100 samples of sputum from as many patients suspected to be suffering from pulmonary tuberculosis, on the basis of history, examination and chest radiographs were collected in sterile containers. The samples were mixed well on vortex mixer. Direct smears for Ziehl-Neelsen stain, cold stain and fluorescent stain were prepared. Part of sample was processed for concentration smear by petroff’s method and culture on L-J medium. Another portion of the sample was put to concentration by floatation method (Krawsnov and Lawrence, 1969). Smears prepared by petroff’s and floatation method were also stained by Z-N stain.
As depicted in table ITable I.
Showing sensitivity, specificity and predictive value (Pv) of positivity of various methods.Direct smearCold StainFluorescent stainConcentration smearFloatation methodSensitivity57.451.862.970.482.5Specificity97.8100.093.595.695.6Efficiency76.074.077.082.090.0Pv of positivity96.8100.091.995.095.8, floatation method proved to be the most reliable in demonstrating the presence of acid fast bacilli in sputum samples.
Concentration by floatation method gives significantly better results as compared to direct smear (p<0.001). The next best was concentration smear by petroff’s method followed by fluorescent stained smears.
Use of floatation method needs to be popularized as this method detects significantly more number of active pulmonary tuberculosis cases and also does not require any specialized equipment, electricity or skilled manpower as is the case with fluorescent microscopy or petroff’s concentration method.
M.S. Parhar, None.