Diagnostic accuracy and safety of Diaskintest® compared with the tuberculin skin test for detecting tuberculosis infection in BCG-vaccinated Brazilian adults

Afranio Lineu Kritski¹Carla Conceição¹Martha Oliveira¹Ana Paula Razal Dalvi¹Marcelo Cordeiro-Santos^2,3,4Adriana Da Silva Rezende Moreira¹Aline Lopes¹Rogerio Rufino⁵Luciana Silva Rodrigues⁵Eliene Mesquita⁶Sergian Vianna Cardozo⁷Maria Jose Fernandes Pereira⁸Sumire Sakabe⁹Anete Trajman¹Marcela Bhering^1,10*

• ¹Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil
• ²Universidade do Estado do Amazonas, Manaus, Brazil
• ³Fundacao de Medicina Tropical Doutor Heitor Vieira Dourado, Manaus, Brazil
• ⁴Universidade Nilton Lins, Manaus, Brazil
• ⁵Universidade do Estado do Rio de Janeiro, Rio de Janeiro, Brazil
• ⁶Secretaria Estadual de Saude do Rio de Janeiro, Rio de Janeiro, Brazil
• ⁷Universidade do Grande Rio Professor José de Souza Herdy, Duque de Caxias, Brazil
• ⁸Serviço de atenção especializada em tuberculose, Itaborai, Brazil
• ⁹Centro de Referência e Treinamento DST-AIDS, São Paulo, Brazil
• ¹⁰Oswaldo Cruz Foundation (Fiocruz), Rio de Janeiro, Brazil

Background: Evidence on the accuracy and safety of the ESAT-6/CFP-10–based Diaskintest® in BCG-vaccinated populations outside Eastern Europe remains limited. In Brazil, recurrent shortages of purified protein derivative (PPD) have challenged the implementation of tuberculin skin testing, underscoring the need to evaluate alternative tools for tuberculosis infection (TBI) screening. This trial compared the diagnostic performance and safety of Diaskintest® with the tuberculin skin test (TST) using PPD Rt-23 in Brazilian adults, a predominantly BCG-vaccinated population. Method: A double-blind randomised clinical trial was conducted at eight centres in Brazil between July 2023 and September 2024. Participants were allocated to the TB group (microbiologically confirmed pulmonary TB), in whom sensitivity was estimated, and the control group (healthy, unexposed adults), in whom specificity was estimated using the QuantiFERON-TB Plus® (QFT-Plus) as the reference standard. All participants first underwent QFT-Plus testing, followed by intradermal application of Diaskintest® and TST in opposite arms, with randomised right-left allocation. Induration was measured at 72–96 hours using a prespecified 5 mm cutoff. Secondary outcomes included safety, assessed through active monitoring of adverse events (AEs), included injection site reactions. Results: A total of 337 controls and 136 TB participants were enrolled. TST showed higher sensitivity than of both Diaskintest and QTF-plus (0.84 [95% CI 0.76-0.90] vs 0.68 [95% CI 0.59-0.76], and 0.63 [95% CI 0.54-0.72], respectively). Diaskintest® demonstrated higher specificity than TST (0.93% [95% CI 0.90-0.96] vs 0.75 [95% CI 0.70-0.80]). Injection-site reactions occurred less frequently with Diaskintest® than with TST (1.7% vs 4.9%. RR=0.35 [0.15-0.79]). The most common reactions were phlyctenular reactions and itching. No serious AEs were observed. Conclusion: TST had greater sensitivity than Diaskintest®, whereas Diaskintest® demonstrated higher specificity and fewer local adverse reactions. In this study, specificity was estimated using QFT-Plus as a surrogate reference standard, acknowledging the absence of a true gold standard for tuberculosis infection. These complementary performance profiles highlight a trade-off between false-positive reduction and case detection, suggesting that the choice of test should consider programmatic priorities and local epidemiological context. Trial registration: Trial registry number RBR-7tn2ysw (https://ensaiosclinicos.gov.br/ registered on January 25, 2021).