TB diagnostic tests: How do we figure out their costs?

Hojoon Sohn, Jessica Minion, Heidi Albert, Keertan Dheda, Madhukar Pai

Research output: Contribution to journalReview articlepeer-review


Tuberculosis continues to be a major global health problem. Lack of accurate, rapid and cost-effective diagnostic tests poses a huge obstacle to global TB control. While several new diagnostic tools are being developed and evaluated for TB, it is important that new tools are introduced for widespread use only after careful validation of accuracy, impact as well as cost-effectiveness in real-world settings. While there are large numbers of studies on the accuracy of TB diagnostic tests, there are few studies that are focused on cost and cost-effectiveness. There are currently no widely accepted standards on how to evaluate costs of a TB test. In this review, we describe the basic approach for computing the costs of TB diagnostic tests, and provide templates for various data elements and parameters that go into the costing ana lysis. We hope this will pave the way for a standardized methodology for costing of TB diagnostic tests. Such a tool would enable improved and more generalizable costing analyses that can provide a strong foundation for more sophisticated economic analyses that evaluate the full economic and epidemiological impact resulting from the implementation and routine use of performance-verified new and innovative diagnostic tools. This, in turn, will facilitate evidence-based adoption and use of new diagnostics, especially in resource-limited settings.

Original languageEnglish (US)
Pages (from-to)723-733
Number of pages11
JournalExpert Review of Anti-Infective Therapy
Issue number6
StatePublished - Aug 2009
Externally publishedYes


  • Cost
  • Cost-effectiveness
  • Diagnostic tests
  • TB

ASJC Scopus subject areas

  • Microbiology
  • Microbiology (medical)
  • Virology
  • Infectious Diseases


Dive into the research topics of 'TB diagnostic tests: How do we figure out their costs?'. Together they form a unique fingerprint.

Cite this