Settings: A general hospital and three peripheral health institutions (PHI) in Tiruvallur District, south India.
Objective: To validate the case detection strategy for diagnosis of pulmonary tuberculosis by smear microscopy of two sputum specimens versus three.
Methods: In the Revised National Tuberculosis Control Programme (RNTCP), three smears from every symptomatic patient attending the PHI were examined for acid-fast bacilli (AFB) by Ziehl-Neelsen (ZN) microscopy. The data from the Tuberculosis Laboratory Registers from the above centres were analysed.
Results: Of 7843 chest symptomatics who had provided three sputum specimens for examination, 895 (11.4%) were smear-positive on at least two specimens. Examination of two specimens--first spot and early morning specimens (837, 93.5%) or early morning and second spot specimens (843, 94.2%)--yielded the highest number of cases.
Conclusion: The examination of two sputum smears (one spot and an early morning) is as effective as examination of three smears.