RESUMO
BACKGROUND: The T-SPOT.TB, an interferon-gamma release assay, is an indirect test of Mycobacterium tuberculosis infection. Due to sparse and conflicting evidence, the use of interferon-gamma release assay is limited in young and HIV-infected children. We determined the prevalence of invalid, borderline, positive and negative results and associations with key demographic variables during routine pediatric use in a low tuberculosis burden setting. METHODS: For pediatric samples received at Oxford Diagnostic Laboratories between 2010 and 2015, the associations between initial test outcome and demographics were estimated by bivariate analysis and logistic regression. RESULTS: A total of 44,289 samples (median age 12.5 years; interquartile range 7.7-15.5), including 5057 samples (11.6%) from children under 5 years old, were received from 46 U.S. states, Washington, DC and Puerto Rico. A total of 592 samples (1.3%) could not be tested. T-SPOT.TB positivity was strongly correlated (r = 0.60; P < 0.0001) with state TB incidence. Compared with negative results, positive results were more likely in samples from older children (P < 0.0001), public health clinics (P < 0.0001) and rural locations (P = 0.005). Although infrequent (0.6%), invalid results were more common in samples collected at HIV clinics (odds ratio = 2.5, 95% confidence interval: 1.3-4.9) and from younger children (P = 0.03). These invalid results were more likely due to a robust nil (negative) control response rather than a weak mitogen (positive) control response. CONCLUSIONS: The T-SPOT.TB test correlated strongly with well-recognized risk factors for tuberculosis infection and provided evaluable results in 98% of children. To optimize the impact of testing on clinical decision making and patient outcomes, local epidemiology and individual patient risk should be considered when incorporating IGRAs into pediatric guidelines.