What if They Don't Have Tuberculosis? The Consequences and Trade-offs Involved in False-positive Diagnoses of Tuberculosis

Abstract

To find the millions of missed tuberculosis (TB) cases, national TB programs are under pressure to expand TB disease screening and to target populations with lower disease prevalence. Together with imperfect performance and application of existing diagnostic tools, including empirical diagnosis, broader screening risks placing individuals without TB on prolonged treatment. These false-positive diagnoses have profound consequences for TB patients and prevention efforts, yet are usually overlooked in policy decision making. In this article we describe the pathways to a false-positive TB diagnosis, including trade-offs involved in the development and application of diagnostic algorithms. We then consider the wide range of potential consequences for individuals, households, health systems, and reliability of surveillance data. Finally, we suggest practical steps that the TB community can take to reduce the frequency and potential harms of false-positive TB diagnosis and to more explicitly assess the trade-offs involved in the screening and diagnostic process.

Figure 1.

Screening and diagnostic pathway for tuberculosis (TB). From a general population, a screening population is formed from individuals with (orange) and without (green) TB. The diagnostic algorithm is applied to the screening population, categorizing individuals into those recommended for TB treatment (following a true-positive or false-positive diagnosis) or not. The contribution of false-positive TB diagnoses is mostly driven by the prevalence of TB in the screening population and the specificity of the diagnostic algorithm (see Table 1). The dashed arrows on the right highlight the 2 processes that new screening or diagnostic strategies aim to achieve (orange = convert false-negative diagnoses into true-positive diagnoses; green = convert false-positive diagnoses into true-negative diagnoses).

Figure 2.

Change in positive predictive value by varying sensitivity (A) or specificity (B). Figures show relationship between positive predictive value (% of individuals with tuberculosis [TB] diagnosis that actually have TB disease) and prevalence of disease in screening population for combinations of sensitivity and specificity. A, Lines show how relationship changes if specificity for algorithm 1 (see Table 1) remains constant at 97% but sensitivity increases. B, Lines show how relationship changes if sensitivity remains constant at 53% but specificity increases or decreases.