Skip to main content

Surveillance testing for SARS-COV-2 infection in an asymptomatic athlete population: a prospective cohort study with 123 362 tests and 23 463 paired RT-PCR/antigen samples

Our take —

In this study, researchers compared the performance of antigen tests to RT-PCR tests for detection of SARS-CoV-2 infection among 1,931 athletes from 11 US universities. Antigen tests were performed on days of participation in the sporting activities while RT-PCR tests were performed at least once weekly. There were 346 total cases among the athletes (172 identified through the routine, point-of-care antigen-based testing program). Fifty two percent of these 172 were isolated before they would have been identified by RT-PCR alone. Adjusted sensitivity for paired antigen tests was low: 22.9% (95% CI: 15.1%-33.2%), while specificity was high: 99.8% (95% CI: 99.8% – 99.9%), and it was estimated that the sensitivity of weekly RT-PCR testing was similar to antigen testing 2-3 times a week. This suggested that, in resource limited settings, frequent screening with antigen tests could be combined with additional PCR in cases of contact tracing or symptoms. The high participation among athletes (97.9%) and robust testing were strengths of the study. However, this study took place before the Delta variant circulated widely, and before vaccination among non-priority groups became common in the United States—both factors may have implications for performance of the tests. The findings highlight that while athletic activities could be safely resumed with extensive testing, COVID-19 prevention protocols (e.g., masking, social distancing, and contact tracing) should be observed, as about 14% of cases resulted from athletic-related outbreaks. Overall, there were no reported cases of spread from athletes on one team to opponents.

Study design

Prospective Cohort

Study population and setting

The objective of this study was to compare the utility of daily rapid antigen testing to less frequent RT-PCR tests for detection of SARS-CoV-2 infection. Compared to RT-PCR (the reference for SARS-CoV-2 testing), antigen tests are less expensive, do not require a laboratory analysis, have a quicker processing time, but have lower sensitivity. University athletes from 11 different US universities participating in four sports with high risk of transmission (basketball, football, water polo, wrestling) underwent daily screening tests using Quidel SARS-CoV-2 antigen test on days when they participated in sports. At least once a week, they were also tested with RT-PCR paired with antigen test. Additional RT-PCR tests were performed for initial screening or re-entry after time way from athletics, and if an athlete was symptomatic or identified as a high-risk contact. The researchers assessed prevalence of RT-PCR confirmed tests, sensitivity and specificity of antigen paired with RT-PCR, adjusting for COVID-19 prevalence in the population, and accounting for multiple tests within the same individual (through generalized estimating equations).

Summary of Main Findings

From September 29, 2020, to February 28, 2021, a total of 81,175 antigen and 42,187 RT-PCR tests were performed among the 1,931 athletes from 11 universities. There were 346 cases identified—through antigen screening tests (n=172), symptomatic RT-PCR testing (n=74), RT-PCR initial/re-entry screening (n=32), and RT-PCR contact tracing (n=24). About 52% of the 172 identified through screening tests were identified on non-paired testing days and isolated before they would have been identified by RT-PCR alone, which the authors estimate prevented 234 athlete days of infectiousness. 83 positive RT-PCR tests were reported among tests paired with 23,462 antigen tests. Adjusted sensitivity for antigen tests was low: 22.9% (95% CI: 15.1%-33.2%) while specificity was high: 99.8% (95% CI: 99.8% – 99.9%). In seven cases, RT-PCR was positive, but antigen was negative, but RT-PCR result was available a day after antigen test. Two athletic-related outbreaks (involving 3 or more athletes) occurred at two different schools during the period and accounted for 13.8% of the 346 total cases. Most cases (86%) among the athletes were community-acquired, i.e., took place outside the athletic activities, highlighting the need for multifaceted mitigation measures including masking, social distancing, etc.

Study Strengths

Study Strengths:

This study included a robust and frequent testing program with high participation (97.9% of eligible athletes took part) with a small proportion of missing data due to blank, invalid, and repeat tests that were excluded from analyses (<2% of all tests conducted).


Though RT-PCR is considered the gold standard for SARS-CoV-2- testing, the possibility of false negative and false positive RT-PCR tests cannot be completely ruled out. Across sites, RT-PCR assays with different limits of detection were used, potentially leading to some differences in SARS-CoV-2 detection. This study was conducted before vaccination became available for non-priority groups, and the Delta variant was in wide circulation—both factors may have implications for antigen test performance. In addition, results are specific to the antigen test used in the study. No economic evaluation was completed.

Value added

This comparison of the performance of daily antigen tests with weekly and situation specific RT-PCR for detection of SARS-CoV-2 infection, provides useful insights that can inform and help optimize the frequency of antigen or PCR surveillance testing. This may be particularly informative in settings with constraints (cost and/or access) to more frequent RT-PCR testing.

This review was posted on: 1 September 2021