Data collection

A link to the online questionnaire was sent to all members of the European Society for Clinical Virology (ESCV) and European Non-Polio Enterovirus Network (ENPEN) in order to reach a good selection of clinical and public health virology laboratories on the European territory. One reminder letter was sent. Data were collected using the EU-survey platform. Questionnaire was opened 23rd September and closed 12th October.

Basic participant information was collected; this included whether their laboratory was performing diagnostic SARS-CoV-2 services and whether it was linked to hospital, university and/or national public health institution. The survey focused on the epidemiology of SARS-CoV-2 infections in the respondent’s institution or country, and hence the number of SARS-CoV-2 infections diagnosed, and population size covered by that institution were captured (Figure 1). Data on mitigation and quarantine strategies applied were also collected. Details of serological testing methods applied in respondent’s institution or country, including their details of their validation, use in diagnostic and different seroprevalence studies, was also collected. Data on serological testing performed was collected, including the estimated number of samples tested.

Figure 1.  Survey structure divided by the five sections of the survey, their specific content and the purpose of the section.

Data analyses

A descriptive analysis was undertaken by participant laboratory and country. The geographical distribution of laboratories by their main function in Europe were mapped.

Participant laboratories were asked to report the COVID-19 case definition in use during the study period.

Responders also reported, information on the month when the highest number of confirmed SARS-CoV-2 cases, and this was compared with the timing of lockdown strategy applied. The estimated number of COVID-19 cases was compared to the length of mitigation strategies applied, in order to demonstrate whether to highest number of cases corresponded to the longest period of restrictions applied.

Serological methods reported by participant laboratory were divided into virus neutralisation, commercial and in-house assays. In addition, we collected and analysed the data on targets (ie, immunoglobulin class or antigenic target) and we recorded if the laboratories has performed the validation or verification of their serological assay, and if so, using how many samples. Serological testing was categorised into clinical diagnostics, testing of convalescent plasma donor to determine which donations contained high enough SARS-CoV-2 antibody levels to be used as a treatment for COVID-19 and seroprevalence studies (ie, serology was used to drive or support studies to determine the susceptibility of the population calculating the seroprevalance in a population).

For seroprevalance studies, we collected and analysed data on study population, design and size of the study.

Laboratories performing seroprevalance studies were asked to estimate the seroprevalance as the proportion of SARS-CoV-2 seropositives by the number of inhabitants, and the study population size when possible. The proportions reported were compared, analysing whether any significant differences were observed with χ² test, considering a P  < 0.05 as statistically significant.

Case-definition

Thirty-three out of 36 participating laboratories reported use of the COVID-19 case definition during the study period: in most cases viral RNA detection in nasal-pharyngeal specimen or secretions taken from the respiratory tract was considered sufficient to confirm a case (70%, 23/33), whereas the remaining responders defined a confirmed COVID-19 case only when clinical symptoms consistent with COVID-19 were reported in addition to molecular RNA detection (28%, 9/33). Case definition was not applicable for three responders (ie, private laboratories) and one participant did not describe a case definition. Notably, as the case definition might have changed for some locations during the first wave of the pandemic, coinciding with the study period, the most recent case definition was recorded.

Case numbers

Around 100 000 laboratory confirmed SARS-CoV-2 infections identified between 1st March and 31st June 2020 were reported by the 34 participant laboratories. However, it is important to note that this was an estimate only and the number of reported cases ranged from less than 10 reported by Bulgaria to more than 10 000 cases reported by participant laboratories in Denmark, Portugal, Spain and Sweden ( Table 1 ). The highest number of infections were recorded in March by 11 laboratories from 6 countries and in April by 15 laboratories from 10 countries ( Table 1 ). Only one participating laboratory recorded a peak number of cases in May (Portugal) and five in June. No such data were obtained from 4 laboratories.

Mitigation and quarantine strategies

Mitigation and quarantine strategies were reported by 33 participants. Sweden (reported by both participating laboratories) was the only country which did not apply any mitigation or lockdown during the first pandemic wave, but were instead recommending people to keep social distancing, avoid gathering with more than 50 people and restricting the access to care homes ( Table 1 ). 30 out of the remaining 32 participant laboratories reported a national lockdown undertaken as mitigation strategy, with varying length. For 26 participating laboratories, it was possible to compare the timing of the peak of COVID-19 case detection and implementation of mitigation measures. It was observed that the peak number of cases were recorded one month after the introduction of mitigation measures by 42% of participant laboratories (11/26), by 39% (10/26) during the first month of lockdown, by 15% (4/26) the month after the end of lockdown and by 4% (1/26) during the last month of these actions ( Table 1 ). Severity of mitigation strategies, such as the length of national lockdown was not always proportional to the number of cases reported. For instance, Bulgaria reported less than 10 cases, whereas over 10 000 cases were reported by the Danish participants. Despite the difference in reported number of SARS-CoV-2 infections, both countries applied four months of lockdown ( Table 1 ). Six countries including Belgium, Croatia, Greece, Ireland, and Spain reported the use of local lockdowns in spite of the national lockdown in order to control the increasing numbers of SARS-CoV-2 infections in the affected areas ( Table 1 ).

Most laboratories reported the use of 14-day isolation period for the close contacts of confirmed SARS-CoV-2 case (29/33) and some of them applied these also for the contacts of suspected SARS-CoV-2 case (20/33). All laboratories placed contacts into 14-day quarantine while waiting for the PCR result and that was also applied for foreign travel (17/33; Table 1 ).

Serological assays used

All except one laboratory had introduced at least one SARS-CoV-2 serological assay for routine work during the first pandemic wave (n = 35), while the remaining laboratory was planning to introduce such a test. All of them had introduced commercial assay for SARS-CoV-2 serology, and one focused on live virus neutralisation testing (Kiel; Table 2 ). Among the commercial assays performed, Euroimmun ELISA was the most common (17 participating laboratories from 36), followed by Roche (n = 12), Abbott (n = 10), Diasorin (n = 7) and Wantai (n = 7; Table 2 ). Three laboratories used in-house assays and eight were performing neutralising antibody testing in addition to the commercial methods. These included neutralisation testing using live virus (n = 7; three laboratories in Germany, one in Croatia, Italy and Norway), pseudotype (Ireland) or other surrogate neutralising antibody testing (Spain, Table 2 ).

Each laboratory using commercial or in-house serological assays reported availability to measure IgG response alone (3/36 laboratories) or together with other immunoglobulins (27/36); further testing included IgM (23/36) or IgA antibodies (12/36). Assays targeted most often the spike protein (S, n = 12) but also nucleoprotein (N, n = 10) or receptor binding domain of spike protein (RBD, n = 8). Only two laboratory reported availability to measure antibody responses against all three target proteins (Austria and one Slovenian laboratory, Table 2 ).

Assay validation

29 out of 36 laboratories reported that their method had been validated before introduction into clinical use. Among the remaining seven laboratories, two reported that the internal validation performed by their testing system was sufficient enough for them and hence no further assessment was required. Two laboratories could not identify sufficient control material for validation, one relied on published validation data on the methods selected for use and one decided to introduce the emergency test for convalescent plasma donor screening without validation. The number of known positive and negative samples included in validation work varied from one sample each to over 500 samples. Four laboratories (14%) had included fewer than 15 samples in their validation panel.

Most laboratories reported the use of serum samples only (20 from 34 responses), whereas 12 laboratories also accepted plasma, one saliva and one plasma and saliva in addition to serum. Only one laboratory was limited to using plasma or capillary blood only.

Application of serological testing

Participating laboratories were asked to report the purpose of serological testing in use (see section Serological assays used above). In particular 30/36 laboratories declared to use SARS-CoV-2 serological testing as support for seroprevalance studies ( Table 3 ), 22 laboratories as a diagnostic service, 13 laboratories performed serological testing to support collection of the convalescent plasma and 8 as part of larger studies, including also other methodologies of detection ( Table 4 ).

Within these studies or in diagnostic routine, serological testing was used to confirm a past infection in patients without SARS-CoV-2 diagnosis (26 participants 34), to prove acquired immunity to the infection (n = 15), to confirm past infection in those with a negative diagnosis (n = 15), to estimate the number of asymptomatic infections (n = 13) and to evaluate serological responses in immunocompromised patients (n = 10).

Seroprevalence studies performed used several different designs; the descriptive cross-sectional study design was most commonly used followed by cohort study, convenient or probability sampling ( Table 3 ). Targeted sampling was also used. A total of 59 seroprevalence studies were reported as having been performed so far; most of them were focusing on health care workers (HCW, 20 laboratories out of 30), followed by those on general populations (16 laboratories), schools or other institutions (10/30) and blood donors (8/30). Furthermore, 13 HCW studies included health care workers or other personnel such as cleaners, porters, administrative or laboratory staff in hospitals. Other three HCW studies focused on staff members in contact with SARS-CoV-2 infected patients, 2 studies on a specific ward and other 2 on HCW with a previous laboratory confirmed SARS-CoV-2 infection.

Three laboratories reported the observed seroprevalence among HCW; it was 11% in Belgium, 5% in Denmark and 22% in the Netherlands ( Table 3 ). Five out of the 16 laboratories performing seroprevalence studies on general population reported a seroprevalence ranging from 0.88% in Slovenia to 5.6% in Belgium and 4 of those laboratories reported also the study population size, which varied from 1368 to 18 000 people involved. In particular, Slovenia appeared to have a significantly lower seroprevalence when compared with other countries reporting higher seroprevalences (0.88% vs 2.2 in Denmark, P  = 0.001; 0.88% vs 2.9% in Lisbon 2, P  < 0.0001; 0.88% vs 2.75% in Penafiel, P = 0.0001). This difference was also reflected when compared the estimate number of confirmed COVID-19 positive, that was significantly lower in Slovenia (101-1000 cases) compared with Denmark and Portugal that reported >10 000 cases.

Furthermore, 31 laboratories reported the epidemiological information collected with their serological data ( Table 5-8 Table 6 Table 7 Table 8). Most of them collected sample date (26 laboratories), patient’s age (n = 28), information whether they had a laboratory confirmed SARS-CoV-2 infection and if so, date of diagnosis (n = 26), reported symptoms (n = 22), symptom onset (n = 23) and severity of infection (n = 14) ( Table 5 ). Data collected on severity of infection included whether they were hospitalised (n = 17), the length of their hospitals stay (n = 9) or whether they were admitted to the intensive care (n = 14) ( Table 7 ). Smaller number of laboratories reported data on outcome of the infection (n = 14, Table 6 ) and the presence of underlining medical conditions (n = 14, Table 8 ). Data on long-term outcomes was rarely collected; data on on-going SARS-CoV-2 PCR positivity was collected by 8 laboratories whereas only a few laboratories systematically collected data on long-term respiratory, cardiac or neurological issues (n = 2, 1 and 1, respectively; Table 6 , Table 8 ).