Diagnostic Efficacy of T-SPOT.TB for Active Tuberculosis in Adult: A Retrospective Study.

Abstract

To explore the diagnostic efficacy and optimal diagnosis threshold of T-SPOT.TB for active tuberculosis in adults and to evaluate the influential factors for T-SPOT.TB results. A retrospective study of 1193 adult inpatients from April 2015 to March 2018 in Ruijin Hospital was conducted. All included patients underwent T-SPOT.TB assay, and were divided into two groups, active tuberculosis (ATB) and non-active tuberculosis (non-ATB) groups. Their demographic data, underlying diseases, personal history and laboratory findings were collected to calculate the diagnostic efficacy at different diagnosis thresholds and analyze the impact factors. Symptoms and imaging features of ATB patients were recorded and analyzed. A total of 114 ATB patients and 1079 non-ATB patients were included in the study, and ATB patients had a higher level of T-SPOT.TB than the non-ATB group. Sensitivity and specificity of T-SPOT.TB for diagnosing ATB are 78.95% and 68.58% as the threshold at 6sfu. In the diagnosis accordance curves, ESAT-6, CFP-10, and max (ESAT-6 or CFP-10) reached the plateau at 40sfu, while sum (ESAT-6 and CFP-10) reached the plateau at 70sfu. Multivariate logistic regression analysis showed that obsolescent tuberculosis (p=0.001), smoking history(p=0.005), diabetes(p=0.035) and advanced age (≥65 years old) (p=0.031) were risk factors for false-positive result of T-SPOT.TB. In terms of imaging features, logistic regression analysis suggested that the thin-wall cavitary lesion was the only feature associated with the result of T-SPOT.TB. As for using T-SPOT.TB test to diagnose active tuberculosis, increased threshold could significantly elevate the diagnosis accordance. And we suggest that the threshold of T-SPOT.TB could be increased to 40sfu for diagnosing ATB. Attention should be paid when diagnose ATB in population with obsolescent tuberculosis, smoking history, diabetes and advanced age, for the risk of false-positive.