A comparative study of fluorescence microscopy and Ziehl-Neelsen techniques in the examination of sputum for the diagnosis of tuberculosis at a middle-income country referral hospital

SETTING: Tuberculosis Reference Laboratory in Cayetano Heredia National Hospital in Lima, Peru. OBJECTIVES: Comparison of auramine fluorescence (FM) and Ziehl-Neelsen (ZN) light microscopy for diagnosis of tuberculosis.

MATERIAL & METHODS: Two direct smears from 2179 sputum specimens from patients with suspected pulmonary tuberculosis were prepared for blinded examination of one smear each by ZN and FM, with Lowenstein-Jensen (LJ) culture as gold standard.

RESULTS: 51 (2.3%) samples were excluded from analysis because of contaminated culture. All excluded samples were negative in both smears. Of the remaining 2 128 specimens, 272 (12.8%) were positive on culture and 151 (7.1%) were positive on at least one microscopic examination. ZN and FM detected 36.0% and 38.6% of 272 LJ culture-positive specimens. Patient-wise there were 8.1% (126 / 1 553) positives on any smear against 12.0% (187 / 1 553) on any culture. ZN and FM smear were positive in respectively 43.3% and 46.5% of culture-proven cases. All differences between microscopy and culture were significant (P<0.001), but not those between microscopy techniques. The considerable numbers of bacilli found in positive direct smears from discordant microscopy sets suggest that repeating smears can improve microscopy sensitivity more than variations of correctly executed technique, provided that overload is avoided. Since FM has the added advantage of being time-saving allowing a large number of sputum specimens to be examined in a given time then it could be particularly useful in laboratories handling large numbers of sputum specimens and could avoid the sensitivity loss associated with high workload.