Skip to main content

Assessment of protection against reinfection with SARS-CoV-2 among 4 million PCR-tested individuals in Denmark in 2020: a population-level observational study

Our take —

The study sought to investigate the risk of reinfection among people infected during the first surge and potential subsequent reinfection in the second surge. The study used data from 525,339 people in Denmark who received PCR-testing from March to May 2020, to estimate 80.5% protection of first infection from later reinfection. This was reduced among people aged 65 years and older, though, to 47.1% protection. The primary study limitation was that individuals were not required to have a negative test between their first wave and second wave tests, which may lead to the same persistent infection being classified as reinfection. However, this study represents the largest to date to investigate reinfection, using a country-wide sample with individual-level information.

Study design

Prospective Cohort

Study population and setting

This study sought to describe reinfection among people infected during the first wave of COVID-19 in Denmark. It used individual-level patient data from people tested in Denmark in 2020 from the Danish Microbiology Database. The first COVID-19 epidemic wave was determined to be from March to May 2020, and the second was defined as September to December 2020. Individuals who tested positive during the first surge and died before the second were excluded. At each polymerase chain reaction (PCR) testing site part of TestCenter Denmark, individuals were tested for SARS-CoV-2 infection by nasopharyngeal swab. Individuals included in the analysis must have been tested before June 1, 2020 (either negative or positive), and then they were followed through the second surge to examine different rates of later testing positive. Individuals testing positive from June 1 to August 31 were excluded as well. The rate of infection was defined as the number of positive PCR tests in the second surge, over the number of person-days at risk for infection during the second surge from September 1st through December 31st. Individuals were censored at death. Adjusted rate ratios were generated using Poisson regression, adjusting for sex, age group, and frequency of tests on each person in 2020.

Summary of Main Findings

After excluding individuals who died between surges and those who tested positive between surges, 525,339 people who tested positive during the first surge also had follow-up in the second surge. Overall, 11,068 (2.11%) people tested positive in the first surge, of whom 72 (0.65%, 95% CI: 0.51 – 0.82) tested positive in the second surge. Among those 514,271 individuals testing negative in the first surge, 16,819 (3.27%) went on to test positive in the second surge. This was equivalent to a daily rate of infection of 5.35 positive tests per 100,000 people during the second surge among people who initially tested positive in the first surge vs. 27.06 positive tests per 100,000 people in the second surge among those initially testing negative. This was equivalent to an adjusted RR of 0.20 (95% CI: 0.16, 0.25) among those who initially tested positive vs. those who initially tested negative. The estimated protection against repeat infection after prior infection was 80.5% (95% CI: 75.4% – 84.5%). Similar findings were present in their sensitivity analyses with all testing data, regardless if people were tested in the first or second surge, but individuals age 65 years and older had reduced protection from reinfection (47.1%, 95% CI: 24.7 – 62.8%).

Study Strengths

The study has data from all individuals receiving PCR tests in Denmark during the study period, which reduces the chance of selection bias. The study also had their long-term follow-up and was able to capture second tests during the second surge, with little loss to follow-up. Denmark has large testing capacity and offered free testing without referral early in the pandemic, which contributed to the completeness of the data.


The largest limitation is that behavior and clinical information were not available—individuals who initially tested negative may have had increased risk behaviors compared to individuals who tested positive, or vice versa if individuals initially test positive and then assume subsequent immunity, and therefore engage in more risky behaviors. Therefore, this increase in their infection in the second surge may exaggerate the relative protective effect when used as a reference to compare people who tested positive in the first surge. Additionally, misclassification of reinfection may occur if individuals did not actually clear the initial infection during the first surge, and later tested positive due to the same infection in the second surge. This issue likely affects only a small number of their cases, but the lack of confirmed negative tests between surges may reduce the accuracy in these estimates.

Value added

This is the largest population-based study of reinfection to occur, using country-wide individual-level data.

This review was posted on: 22 April 2021