False-positive quantiferon results at a large healthcare institution.

TO THE EDITOR—Eighteen million US healthcare workers (HCWs) undergo mandatory annual screening for latent tuberculosis [1]. Screening is also performed in tuberculosis contacts, immunocompromised patients, and other individuals employed in the public sector. Many institutions have switched from the tuberculin skin test to the Quantiferon (QFT) assay for latent tuberculosis screening. In recent months, we experienced a sharp increase in false-positive QFT results. We want to ensure that other institutions using QFT testing are aware of the potential problem of falsepositive results. Our institution began using the QFT test in 2008, screening approximately 10 000 HCWs annually. Due to the detection of transient increases in the daily positivity rate, the clinical laboratory implemented a surveillance program in 2010 for tracking daily positivity rates. As we have reported previously [2], in November 2011, the proportion of positive results significantly increased, from a mean baseline of 9% to a mean of 31%. The elevated proportion of positive QFTs returned to baseline after switching to a new lot of tuberculosis antigen tubes. Direct comparison of the tubes with tubes from a different lot identified that a lot-specific problem with the tuberculosis antigen tubes was the root cause of false-positive results. The manufacturer reportedly instituted stricter quality assurance measures prior to distribution of the QFT assays. We recently again detected a significant increase in the daily QFT positivity rate following the introduction of lot A121200T to a mean of 16.2% during the period of 17 July to 2 October 2013, compared with a 5.7% baseline rate in January through June 2013 (P < .001, χ test). Throughout August and September, multiple different lots were implemented in hopes of correcting the problem. However, the elevated rate persisted until lot A1302027 was put into use on 3 October 2013 and the average daily positivity rate throughout the remainder of October declined to 5.4% (Figure 1). An investigation into the QFT practices at our institution in conjunction with the manufacturer did not identify any preanalytical or analytical errors. Within-subjects comparison of QFT results (n = 19) with tuberculosis antigen tubes of the suspect lot (A130500S) and the new lot (A1302027) showed a significant increase in the proportion of positive results with the suspect lot (57.9% vs 0%; P < .001, McNemar test). The median tuberculosis antigen minus nil values were significantly higher in the suspect lot (0.52 IU/mL) than in the new lot (0 IU/mL; P < .001). The impact on the occupational health program and HCWs was significant. During the episode described, 544 individuals with positive QFT results were asked to return for retesting. The