Introduction Policies promoting access to syphilis diagnosis have led to the widespread use of rapid tests (RT) in health facilities of different complexities throughout Brazil. To monitor the quality of the testing procedures, the Ministry of Health introduced the National Program of External Quality Assessment for RT (EQA-RT). This involves healthcare professionals registering on an online platform to receive a panel with four dried tube specimens with unknown reactivity for syphilis. The respondents test the samples as if it they were from their own patients and submit the results online for subsequent analysis.

Methods We analysed reports from eight EQA-RT rounds conducted in 2014–2016. The quality of the professionals´ performance was assessed by examining the degree of concordance between the results from the panel samples that were expected and those actually reported. A certificate of approval is issued in the event of 70% accuracy. Any professional receiving less than 70% approval rating is sent a report suggesting possible causes and solutions for issues that might have been responsible to prevent the quality of testing.

Results The average number of participants per round was 843 (545 in 2014, 909 in 2015 and 1.075 in 2016) distributed in 347 mainly primary healthcare units. It was interesting to note that many higher complexity services, despite registering in the program, failed to report the results (e.g. a 55% abstention rate in one of the rounds). Regarding the performance on EQA-RT, an average of 93% of respondents were approved per round, with 88% achieving 100% accuracy. The most common failures identified so far have been incorrect pipetted sample volumes and result reading times

Conclusion The majority of healthcare professionals currently participating in Brazil´s EQA-RT program are maintaining a high level of RT quality. While the number of participants doubled in 2014–2016, there is still a need for education work to encourage more adherence to the Program aimed at ensuring the reliability and credibility of the TR results.