Prospective, observational and longitudinal study of those attending FCB sports facilities, in Barcelona, Spain, throughout the 2020–2021 season (August 1st 2020–June 16th 2021).

Included participants were teen players (12–17 years old), young adult players (18–23 years old) and adult workers. Participants who did not sign the informed consent or unwilling to ensure adequate follow-up were excluded.

Two recruitment pathways (RP) were considered: RP1 included participants who lived, studied and trained in the facilities (a residential and sports area called “Masia”); RP2 included participants who lived in their own homes, and just went to train and study in the facilities. Seven complete teams were tracked during the whole course: Outdoor Football, Men: (1) Senior youth A, (2) Senior youth B, (3) Junior youth A; Outdoor Football, Women: (4) Reserve, (5) Senior youth; Basketball: (6) Reserve, (7) Senior youth. In each of them, there were participants who belonged to RP1 and RP2. In addition to these seven teams, others were studied, although not all their components wanted to participate.

A survey was conducted upon recruitment including clinical and epidemiological data. During follow-up, a biweekly health questionnaire was applied enquiring about symptoms and possible COVID-19 contacts. A nasopharyngeal (NP) PCR was performed every 15 days, together with a weekly saliva sample (stored as backup sample). Serologies were done upon enrollment and every 3 months.

From January to June 2021, nasopharyngeal antigens (NP Ag) replaced PCRs, a change applied in parallel to its wider implementation, facilitating a faster turnaround of results. Considering the possibility of false negatives, testing frequency was increased to a weekly basis. However, if participants reported symptoms, or were a contact of a positive case, an additional NP PCR was performed.

When a positive case was detected in a team (considered stable sport-specific groups, made up of the same participants throughout the course, following the same routines and isolation protocols), all its components were considered close contacts, so they were studied and immediately confined (see more information in the Supplementary file).

According to Spanish legislation, the use of face masks was compulsory, and it was recommended to maintain a minimum interpersonal distance and to wash hands regularly. However, players within the same team were allowed to remove masks during training or in their houses/rooms. Other preventive measures were applied, such as daily temperature control (see more information in the Supplementary file).

The first vaccination campaign against SARS-CoV-2 in Spain began on December 27th 2020. The campaign for the population between 16 and 29 years old started on June 30th 2021. During the study, 23 (36%) adult workers received at least one vaccine.

A positive case was defined as a participant with a positive NP PCR/Ag, in the absence of IgG. At enrollment, cases with positive PCR but also positive IgG were considered past infections and were not included as positive cases. The only evidence of a following seroconversion in the vaccinated participants was not criteria for the definition of a positive case.

Cases with an indeterminate PCR were closely followed up by questionnaires and weekly PCR, and were considered negative unless a subsequent PCR resulted positive.

We could not distinguish between primary and secondary positive cases, due to the participants belonged to “crossed” cohorts: those belonging to RP1 or RP2 vs. each particular team independently of where they live.

The primary outcome was to determine the attack rate (AR) in the different cohorts: RP1, RP2, and the seven complete teams. For each cohort, we defined the AR as the percentage of participants who provided at least one positive test during the study. We chose this method instead of widely used ones such as the secondary attack rate (SAR) due to the impossibility of explicitly defining the primary and secondary cases, given the stated crossed-nature of the cohorts.

$$\mathrm=\frac\;\mathrm\;\mathrm\;\mathrm\;\mathrm\;\mathrm\;\mathrm\;\mathrm}\;\mathrm\;\mathrm\;\mathrm\;\mathrm\;\mathrm}100$$

With the aim of identifying when transmission was significant and led to local outbreaks, we determined the expected number of cases that would be found according to the surrounding incidence (see more information in the Supplementary file).

SARS-CoV-2 PCR in NP and saliva was performed according previous published[21]. A positive result was considered if at least two SARS-CoV-2 genes were detected, considering “indeterminate” when only one gene was detected.

Rapid Antigen Test used was from Roche®. IgG, IgA and IgM antibodies were determined using a Luminex system against the receptor-binding domain of the spike glycoprotein of SARS-CoV-2.

The study was approved by the Institutional Review Board and the Sant Joan de Déu Ethics Committee (PIC-200–20) and followed Helsinki Declaration recommendations. All participants or their legal guardians provided a written informed consent.

A Redcap-designed database was used and participant’s data were pseudo-anonymized. All athletes, except the professional teams, were offered participation, so no formal sample size calculation was conducted.

Chi-squared test and Fisher’s exact test were used for comparisons of categorical data; Student’s t-test and Mann–Whitney’s U-test for quantitative variables. To compare epidemiological and microbiological results at different times, the Wilcoxon signed-rank test was used. Odds ratio, determined with Fisher’s exact test and a 2 × 2 table (Matlab’s function), was used to assess the AR of a specific cohort in relation to all the others.