Significance of sputum volume production in tuberculosis patients in Kampala, Uganda

Abstract

Background: Traditionally, the concentration of mycobacteria in the sputum has been used as a marker of severity as well as a measure of treatment response in pulmonary tuberculosis (TB), however, because of its poor sensitivity for the prediction of TB treatment outcomes (particularly in HIV co-infected patients) more effective surrogate markers are needed. Sputum volume itself has not been previously studied as a potential determinant of treatment outcome. Objectives: To assess the association between sputum volume and clinical and radiographic markers of pulmonary TB as well as treatment outcomes among a sample of patients treated for pulmonary tuberculosis at the Mulago Hospital Uganda National Tuberculosis and Leprosy Programme (NTLP) Clinic in Kampala, Uganda. Methods and Findings: Patients admitted to NTLP between July 2003 to January 2007 were enrolled into a prospective cohort study to assess treatment outcomes during and after a course of antituberculous therapy. From this cohort, a secondary analysis of 391 study participants (with a mean follow-up time of 55 months) was conducted to assess the relationship between baseline sputum volume and TB treatment outcome. Of 296 subjects with at least one spot sputum at baseline, a multivariable analysis revealed that having higher volume sputum (≥ 3 ml) at baseline was significantly associated with the presence of cavitary disease (odds ratio (OR) 2.06, 95% confidence interval (CI) 1.2-3.54) and male gender (OR 2.1, 95% CI 1.22-3.62). Among 196 participants with spot sputum specimens who had sufficient microbiologic or demographic data to make a determination regarding their TB treatment outcome, a multivariable logistic regression revealed that having higher volume sputum (≥ 3 ml) at baseline (OR 5.76, 95% CI 1.6-28.56, p = 0.03), increased duration of pre-treatment TB symptoms(≥ 12 weeks; OR 8.07, 95% (11.69-38.59, p = 0.009) and receipt of post-secondary school education (OR 21.03, 95% CI 2.49-177.55, p = 0.005) had significant associations with an unsuccessful TB treatment outcome. Conclusions: Sputum volume produced at the time of pulmonary TB diagnosis is associated with gender and cavitary disease. Of interest, there is a statistically significant increased risk for unsuccessful TB treatment outcomes among those with higher baseline sputum spot volume (≥ 3 ml) in comparison to those with lower sputum volume (<3 mL).