How many sputum specimens are necessary to diagnose pulmonary tuberculosis?

Am J Infect Control. 2005 Feb;33(1):58-61. doi: 10.1016/j.ajic.2004.08.003.

Abstract

The definitive diagnosis of pulmonary tuberculosis (PTB) relies on identifying or culturing Mycobacterium tuberculosis from respiratory specimens. National guidelines have recommended obtaining 3 sputum specimens from patients with suspected tuberculosis, but there has been little data on the number of specimens actually needed to support a diagnosis. We retrospectively reviewed all patients diagnosed with PTB at a public inner-city hospital and assessed the sensitivity of the acid-fast bacilli (AFB) smear and the number of smears needed to establish the diagnosis. Between January 1, 1997 and October 1, 2000, 425 patients were diagnosed with culture-proven PTB. Acid-fast bacilli (AFB) smears and cultures were performed on 951 respiratory specimens from 425 patients. The overall sensitivity of a positive AFB smear increased from 67% with 1 sputum collected to 71% and 72%, respectively, with the second and third specimens. The sensitivity of smears from 239 HIV-negative patients was 75%, 79%, and 80% with 1, 2, and 3 smears, respectively, collected compared with 57%, 61%, and 62%, respectively, for 142 HIV-positive patients. In summary, 2 respiratory specimens proved adequate in establishing a diagnosis of tuberculosis, and the third specimen added little additional diagnostic value.

Publication types

  • Evaluation Study

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Child
  • Child, Preschool
  • Female
  • Humans
  • Infant
  • Male
  • Middle Aged
  • Odds Ratio
  • Retrospective Studies
  • Sensitivity and Specificity
  • Sputum / microbiology*
  • Tuberculosis, Pulmonary / diagnosis*