Suspected pulmonary tuberculosis in rural South Africa--sputum induction as a simple diagnostic tool?

OBJECTIVES: To assess the value of sputum induction (SI) as a diagnostic tool for patients with suspected pulmonary tuberculosis (PTB) who are unable to expectorate or who have a negative sputum smear. DESIGN: Study of an inpatient cohort undergoing SI. SETTING: Mseleni Hospital, a rural district hospital in northern KwaZulu-Natal. SUBJECTS: All adult patients with suspected TB seen at the hospital over a 4-month period. OUTCOME MEASURES: (i) Successful SI; (ii) sputum acid-fast smear result; (iii) change of admission diagnosis as a result of the induction procedure; and (iv) number of patients discharged with a diagnosis other than TB who represented within 4 months with TB. RESULTS: A total of 51 patients (31 female) underwent SI; of these 36 (71%) were able to produce a sputum sample. Fifteen (42%) of those were acid-fast smear-positive (29% of all patients included). The admission diagnosis was changed in 16 (44%) of the patients who were able to give an induced sputum sample as opposed to 4 (27%) who had been unable to expectorate despite an induction attempt (P = 0.38). Three (12.5%) of the 24 patients with a discharge diagnosis other than TB (17 pneumonia, 3 old TB, 2 carcinoma of the lung, 1 bronchiectasis) turned out to have TB within the follow-up period; 2 of those had extrapulmonary TB. CONCLUSION: SI produced a positive smear result in 29% of patients with suspected TB who had previously been smear-negative or unable to expectorate. The method proved an aid to clinical decision making.