Performance of QuantiFERON tests for detecting latent tuberculosis infections: A Meta-analysis

Abstract

Early detection of latent tuberculosis (TB) is crucial for prevention. However, there is a debate over the effectiveness of QuantiFERON tests versus tuberculin skin tests. This meta-review critically evaluated QuantiFERON (QFT) tests for latent TB detection. This diagnostic meta-analysis reviewed studies from January 2015 to July 22, 2024, sourced from Medline/PubMed. Rigorous selection and data extraction ensured robust results. Diagnostic test results were organised into 2 × 2 tables, and statistical analysis was performed using Stata versions 14 and 16. Statistical significance was set at p < 0.05. Data from 35 studies with 23,383 participants were analysed. Males represented 46.7% of the sample. QFT tests were positive in 22.5% and negative in 77.5% of cases. The pooled sensitivity was 49% (range: 3% to 92%), indicating a risk of false negatives, and its pooled specificity was 86% (range: 31% to 98%), reflecting good performance in ruling out non-TB cases. Sensitivity was moderate in HIV (60%) and HCWs (55%), low in TB contacts (48%), children (43%), and RA (39%), and lowest in IBD (15%). Specificity was highest in IBD (94%), children (91%), HCWs (84%), RA (82%), TB contacts (80%), and moderate in HIV (74%). The positive predictive value (PPV) was 61% (range: 5% to 97%) and the negative predictive value (NPV) was 72% (range:13% to 98%). PPV was moderate in TB contacts (70%), RA (64%), HCWs (63%), HIV (51%), children (50%), and low in IBD (42%), and NPV was high in children (88%), HIV (84%), IBD (79%), HCWs (73%), RA (72%), and moderate in TB contacts (63%), with substantial heterogeneity across populations. Despite high specificity, the lower sensitivity suggests QFT tests may miss some latent TB infections. The diagnostic odds ratio was 5.73 (SE ± 1), indicating strong yet highly variable performance (I² = 96.2%). This highlighted the need to combine QFT tests with other diagnostics. QFT tests had high specificity and NPV, ensuring reliable negative results. However, lower sensitivity reduces PPV, making positive results less definitive. The results also revealed moderate sensitivity in HIV, lowest in IBD, highest specificity in IBD, lowest in HIV, highlighting heterogeneous performances.

Keywords: Global accuracy; Predictive values; QuantiFERON; Sensitivity; Specificity.