Diagnosing HIV-associated tuberculosis: reducing costs and diagnostic delay

Int J Tuberc Lung Dis. 2000 Mar;4(3):240-5.

Abstract

Setting: University-affiliated hospital in South Africa.

Objective: To assess the time to diagnosis and the yield and laboratory cost of diagnostic procedures in human immunodeficiency virus (HIV) associated tuberculosis.

Design: Cohort study.

Patients: Adult HIV-infected patients with newly-diagnosed tuberculosis admitted over a 2-year period.

Results: A total of 141 admissions fulfilled the case definition. Sputum smear yield (43% overall) correlated strongly with chest radiograph appearance but not with CD4+ lymphocyte count. Sputum smear yield was approximately 40% per sample sent, resulting in a high cumulative yield when > or = three samples were sent. Smear of sputum or wide needle lymph node aspirates were the most cost-effective diagnostic methods. Significant diagnostic delay occurred in sputum smear-negative patients. Most patients with sputum smear-negative tuberculosis had either pleural effusions or lymphadenopathy. Lymph node biopsy had a high diagnostic yield even in patients with symmetrical nodes, but was under-utilised in this group. There was unnecessary expenditure on cultures, with many patients having several positive cultures.

Conclusion: Repeated sputum smear examination produces a high cumulative yield in HIV-associated tuberculosis. Considerable savings in laboratory utilisation and bed occupancy would have been made if a streamlined diagnostic approach with greater use of lymph node aspirate and early pleural or lymph node biopsy had been followed.

MeSH terms

  • Adult
  • Bacteriological Techniques / economics
  • Cost of Illness
  • Costs and Cost Analysis
  • HIV Infections / complications*
  • Humans
  • South Africa
  • Sputum / microbiology*
  • Tuberculosis / complications*
  • Tuberculosis / diagnosis*
  • Tuberculosis / economics