Study population and setting
Between April 27 and May 11, 2020, a total of 61,075 participants in Spain (~75% of individuals from 35,883 contacted households) participated in a seroprevalence study. The study included a questionnaire regarding symptoms and risk factors related to COVID-19, a point-of-care (POC) antibody test, and, for those who agreed, a blood sample. Seroprevalence was calculated using the POC test and by immunoassay. A range of seroprevalence was calculated based on participants being positive for both tests or positive for either test. Seroprevalence was adjusted for non-response rates in certain groups.
Summary of Main Findings
Based on results from the POC test, seroprevalence was 5.0% (95% CI 4.7-5.4) and 4.6% (95%CI 4.3-5.0) based on immunoassay. The range was 3.7% (95% CI 3.3-4.0%), testing positive by both tests, to 6.2% (95% CI 5.8-6.6%), testing positive on either test. Children younger than 10 years had a lower seroprevalence, a higher prevalence was observed in Madrid (>10%) and lower in coastal areas (3%). Among those with a positive PCR more than 14 days before the study (n=195), seroprevalence ranged from 87.6% (both tests positive) to 91.8% (either test positive). Seroprevalence was less than 20% (n=7,273) in individuals who experienced symptoms, and between 21.9% and 35.8% of seropositive individuals were asymptomatic. When comparing the two tests, the POC test resulted in similar epidemiological interpretations despite having poorer performance characteristics when compared to the immunoassay.
This study represents a large, random, population-based sample that utilized two-stage sampling stratified by province and municipality size. About one quarter of individuals recruited declined to participate, but the authors attempted to account for non-response biases. Another strength of this study is that seropositivity was assessed both through POC test and through immunoassay, allowing for less reliance on a single test’s performance and an assessment of the sensitivity of the POC test compared to that of the immunoassay.
Recall bias and subsequent misclassification cannot be completely ruled out regarding whether or not participants experienced any COVID-like symptoms. Given that recruitment was done by household, certain groups of people were excluded from analyses, including young adults registered at their parents’ home but living somewhere else, and care-home residents. Race was not captured in this study, and therefore disparities by race could not be examined. The performance characteristics of the antibody tests were determined in a population of symptomatic patients, and therefore may be different in a population that includes asymptomatic individuals.
This study provides nationally representative seroprevalence data that can be stratified by sex, age, and geographic region. This information can highlight areas that need further resources, and can be used for planning and decision-making. It also demonstrates the utility of a rapid POC test for epidemiologic studies and surveillance.
This review was posted on: 18 July 2020