BACKGROUND: Smear microscopy is relatively insensitive for the diagnosis of TB. The digestion of sputum with household bleach prior to smear preparation has been reported to improve its sensitivity. This method has not been validated. METHODS: Seven hundred and fifty six patients with symptoms suggestive of pulmonary TB (PTB) were asked to submit 3 sputum specimens for direct microscopy. One specimen was selected at random for culture and another specimen was digested to prepare a further smear. The WHO case definition (>or=2 positive smears or one positive smear and positive culture) was used to compare the sensitivity and specificity of the smears. FINDINGS: Four hundred and fifty five (60%) patients were culture-positive. Of these, 235 (31%) had "definite" PTB and 223 (29%) "very likely" PTB (smear-negative, culture-positive). The WHO case definition identified 51% (235/458) of the patients with "definite" or "very likely" PTB. One digested smear detected 219 (93%) of the 235 patients with "definite" PTB and 10 patients with "very likely" PTB (sensitivity (95%CI) 50% (45-55%); specificity 99% (97-100%)). The positive and negative predictive values for one digested smear were 98% (95-99%) and 56% (52-60%) respectively, which were not different (p>0.5) to the WHO case definition (100% and 57%, respectively). INTERPRETATION: One bleach-digested smear is as sensitive and specific as the WHO case definition for the diagnosis of PTB.
BACKGROUND: Smear microscopy is relatively insensitive for the diagnosis of TB. The digestion of sputum with household bleach prior to smear preparation has been reported to improve its sensitivity. This method has not been validated. METHODS: Seven hundred and fifty six patients with symptoms suggestive of pulmonary TB (PTB) were asked to submit 3 sputum specimens for direct microscopy. One specimen was selected at random for culture and another specimen was digested to prepare a further smear. The WHO case definition (>or=2 positive smears or one positive smear and positive culture) was used to compare the sensitivity and specificity of the smears. FINDINGS: Four hundred and fifty five (60%) patients were culture-positive. Of these, 235 (31%) had "definite" PTB and 223 (29%) "very likely" PTB (smear-negative, culture-positive). The WHO case definition identified 51% (235/458) of the patients with "definite" or "very likely" PTB. One digested smear detected 219 (93%) of the 235 patients with "definite" PTB and 10 patients with "very likely" PTB (sensitivity (95%CI) 50% (45-55%); specificity 99% (97-100%)). The positive and negative predictive values for one digested smear were 98% (95-99%) and 56% (52-60%) respectively, which were not different (p>0.5) to the WHO case definition (100% and 57%, respectively). INTERPRETATION: One bleach-digested smear is as sensitive and specific as the WHO case definition for the diagnosis of PTB.