Cross-Sectional, Prospective Cohort
Study population and setting
The study objective was to identify the seroprevalence of SARS-CoV-2 antibodies in Barcelona, Spain from April 21 to 24, 2020 among a random, age-stratified sample of 311 persons in the community who did not have a suspected case of COVID-19 and previously had not tested positive, and from April 29 to May 5, 2020, among 634 patients with COVID-19 symptoms being followed by a general physician. Patients from the community sample were randomly selected from a total of 19,899 individuals registered at a single primary healthcare center in Barcelona. The symptomatic patients were derived from a sample of 743 that had all consulted a primary care physician with COVID-19 symptoms between March 2 and April 24, 2020 either in-person or remotely, and were followed with routine phone calls to assess symptoms until they resolved. The study also included a standardized questionnaire in both groups about symptoms, visits to the emergency department or hospitalization, and participant gender.
Summary of Main Findings
The study found that, of the 311 asymptomatic participants, 17 were seropositive for IgM and/or IgG (5.47%, 95% CI: 3.44 – 8.58%). Of 634 symptomatic participants, 244 (38.49%, 95% CI: 34.78 – 42.33%) were seropositive. During follow-up, 21.1% of those symptomatic went to the ED and 11.8% were hospitalized. 41 patients had received a SARS-CoV-2 PCR test prior to the antibody test, and 25 (61.0%) were positive. Only 16 out of these 25 (64.0%) tested positive for IgG antibodies during the study. The study also found that positive cases for SARS-CoV-2 IgM/IgG had statistically significant odds of having had contact with other positive cases, having attended the ED, or being hospitalized, compared to those asymptomatic. Among those with symptoms, those with a positive antibody test were also significantly more likely to have had any symptoms (88.1%) compared to those with a negative test (81.0%), though the mean number of symptoms among those testing positive was only about 1 symptom higher than those testing negative.
The study assessed symptoms and tested for antibodies among both those who contacted a physician and those who were asymptomatic in the community. They used a standardized questionnaire for symptom assessment, which increases comparability to other studies. They used a randomly selected sample from the community to reduce selection bias and thus attempt to maximize the representativeness of the sample.
Despite attempts to increase representativeness, there was a reduced response rate among the asymptomatic sample (response rate 52%) which could have impacted results if those who agreed to participate differed from those who did not. Major reasons noted for non-response included being contacted via phone, stopping quarantine in order to attend the study, and concerns around comorbidities, which may result in a healthier (and less exposed) sample included in the study than they would have otherwise had. There was also notable disagreement between participants who received a positive PCR test in the past, versus those who received a positive antibody test in this study, which may have been the result of reporting bias, issues with validity of the specific antibody test the study team used or potential natural variation in the antibody response. Additionally, potential limitations related to the performance diagnostics of the antibody test were not explored, and the definition of an asymptomatic individual, including how many symptoms that they may have presented with and still been considered asymptomatic, was not adequately described potentially resulting in misclassification of asymptomatic cases.
This study assessed both a symptomatic and asymptomatic population in Barcelona, Spain, an area that was significant impacted by the pandemic, and assessed SARS-CoV-2 antibodies to look at the population prevalence of past infection.
This review was posted on: 31 July 2020