Abstract

Background: Smear negative and culture positive results in pulmonary tuberculosis are not uncommon yet the underlying factors are not well established. Objective: To determine factors associated with smear negative and culture positive results. Methods: Pulmonary tuberculosis patients were consecutively recruited for 12 months from five health facilities in Mwanza region, Tanzania. Sputum examinations were done at the recruitment centre and at the TB reference laboratory using Auramine O technique. Culture was done at the TB reference laboratory using Lowenstein Jensen solid media. A post-hoc analysis compared patients who had a smear negative culture positive result (case) with patients who had not (controls). Results: A total of 655 pulmonary TB patients were recruited, 18 had no culture results and were excluded from the analy- sis. Of the remaining 637 patients, 127 (19.9%) had three negative smears at the recruitment centre and 34 patients were a case. Current smoking was strongly associated with being a case, especially in women. Of the 127 patients who had three negative smears at the recruitment center, 104 (81.9%) also had a negative smear at the reference laboratory. Of these, 13 (12.5%) were still culture positive. Conclusion: The frequencies of smear-negative culture-positive results differ between health facilities, indicating possible difference in quality of laboratory procedures. Strengthening of laboratory capacity is needed both for optimizing smear microscopy techniques, and for performing sputum cultures for diagnosing TB when there is a high rate of suspicion. The association between smoking and smear negative culture positive TB needs to be assessed in adequately large studies.