We appreciate Revollo and Llibre's comments1 on the letter recently published by our group on an outbreak of SARS-CoV-2 infection in Figueras prison (Girona)2. As a reminder, infection was detected by rapid antigen test (RAT) in three mildly symptomatic inmates between 23 and 25 December. As a result, in the afternoon of 25 December, the 81 remaining inmates of that prison block were screened using RAT and nine positive results were identified. On 28 December, the 72 inmates who tested negative by RAT underwent rt-PCR testing, 27 (37.5%) of which were positive. The sensitivity of the RAT in this scenario was very low at just 25%, which is why we reported it.
For reasons of brevity, we did not include any information about the population studied in our original letter, which, according to Revollo and Llibre's comments, could be relevant. Since 1 July 2020, new prisoners in Catalonia have been screened by rt-PCR. In total, 46.2% of those infected by the outbreak had been incarcerated after that date and had a prior negative rt-PCR test. The rest of the infected inmates had been in prison for many months and had not been diagnosed with SARS-CoV-2 infection nor monitored due to close contact with an infected individual. As such, the risk of there being a persistently positive or residual rt-PCR result in an infected inmate, as raised by Revollo and Llibre, we consider to be extremely small. Regarding the use of rt-cycle thresholds (Ct) that Revollo and Llibre also discuss, their use in initial phases of infection is low as the values vary over time3. In fact, we only use them very rarely, almost exclusively to assess infection risk in cases with persistently positive PCR results that require prolonged isolation, as "discharge" without knowing whether or not the subject is infectious is a risk in a confined environment.
We agree with Revollo and Llibre's assessment of RATs' high sensitivity for detecting symptomatic cases with a high viral load and transmission potential, typically in the first five days. However, current data are not as conclusive when it comes to their use in pre-symptomatic or asymptomatic patients. The Centers for Disease Control and Prevention (CDC) suggest that negative RAT results should sometimes be considered "presumptive", and in some circumstances (contact with an infected person or high prevalence of infection in the community), it is advisable to confirm the result with a SARS-CoV-2 nucleic acid amplification test (NAAT)4. Other organisations, such as Cochrane, have also confirmed that RATs are generally less sensitive in asymptomatic patients and more sensitive in settings with a high prevalence of infection5. Although it is true that RATs have shown high sensitivity in infected subjects with Ct <25, as pointed out by Revollo and Llibre, the cycle thresholds, as has already been mentioned, are dynamic and vary over time. Moreover, cases with Ct <25 may not include all potential risk cases.
In addition, the diagnostic strategy in a scenario with low or no viral circulation (scenario A) cannot be similar to a scenario with high viral circulation and localised outbreaks (scenario B) where infection of asymptomatic patients can be 70% or higher6. The specificity of rapid antigen tests is high (close to 100%) and they may be suitable for screening populations in scenario A, even assuming that they entail defined and potentially acceptable risks in certain circumstances5. However, what may be acceptable in scenario A, such as the Barcelona Love of Lesbian pilot concert that Revollo and Llibre participated in7, is not acceptable in the context of an outbreak, and even less so in a confined environment like a prison. Exception to this rule is when RAT screening negative results are subsequently confirmed by rt-PCR, as currently recommended by the guidelines and protocols of Spain's Ministry of Health8, the European Centre for Disease Prevention and Control (ECDC)9 and the CDC4.
ConfidentialityThe protocols governing the publication of patient data of our place of work have been followed.
FundingNone.
Conflicts of interestNone.
Please cite this article as: Marco A, Solé C, Abdo IJ, Turu E. Respuesta a «Test rápidos antigénicos o PCR en tiempo real para SARS-CoV-2, ¿qué test usar y por qué?». Enferm Infecc Microbiol Clin. 2021;39:532–533.
