INTRODUCTION: The purpose of this retrospective study was to outline the practical use and clinical value of the immunodiagnostic interferon-gamma release assay (IGRA) on suspicion of active Mycobacterium tuberculosis (TB) infection. MATERIAL AND METHODS: A retrospective study of all patients (n=91) tested with IGRA (Quantiferon-TB-Gold, Celletis International, Australia) for M. tuberculosis infection from 1 January 2005 to the 31 December 2006 at a Danish regional hospital. RESULTS: In 74 patients suspected of active TB, the sensitivity of IGRA was 80% (8/10), the specificity was 85% (50/59), PPV was 47% (8/17) and NPV was 96% (50/52). The positive likelihood ratio (LR) was 5.3 and the negative LR was 0.24. Due to a suboptimal diagnostic setup, the diagnosis of lung cancer in one patient was delayed significantly by a positive IGRA. CONCLUSION: In the current study, IGRA had a high false-positive rate for diagnosing active TB and had a rather low sensitivity which is in line with recent meta analyses. The physicians' use and judgement of this new test seemed suboptimal, as the overwhelming use of IGRA for suspected active TB was not in concordance with international guidelines. IGRA does not differentiate between active and latent M. tuberculosis infection, and IGRA may seldomly be used as a supplementary evaluative tool (with many false negative and "false positive" results).
INTRODUCTION: The purpose of this retrospective study was to outline the practical use and clinical value of the immunodiagnostic interferon-gamma release assay (IGRA) on suspicion of active Mycobacterium tuberculosis (TB) infection. MATERIAL AND METHODS: A retrospective study of all patients (n=91) tested with IGRA (Quantiferon-TB-Gold, Celletis International, Australia) for M. tuberculosis infection from 1 January 2005 to the 31 December 2006 at a Danish regional hospital. RESULTS: In 74 patients suspected of active TB, the sensitivity of IGRA was 80% (8/10), the specificity was 85% (50/59), PPV was 47% (8/17) and NPV was 96% (50/52). The positive likelihood ratio (LR) was 5.3 and the negative LR was 0.24. Due to a suboptimal diagnostic setup, the diagnosis of lung cancer in one patient was delayed significantly by a positive IGRA. CONCLUSION: In the current study, IGRA had a high false-positive rate for diagnosing active TB and had a rather low sensitivity which is in line with recent meta analyses. The physicians' use and judgement of this new test seemed suboptimal, as the overwhelming use of IGRA for suspected active TB was not in concordance with international guidelines. IGRA does not differentiate between active and latent M. tuberculosis infection, and IGRA may seldomly be used as a supplementary evaluative tool (with many false negative and "false positive" results).