Reported COVID-19 Incidence in Wisconsin High School Athletes in Fall 2020

All procedures were deemed exempt by the Institutional Review Board of the University of Wisconsin. In collaboration with the Wisconsin Interscholastic Athletics Association, surveys were distributed to all high school athletic directors on October 1, 2020. In addition to the school name and location, athletic directors were asked whether they had restarted participation in sports since the initial COVID-19 restrictions in the spring of 2020. Those athletic directors who reported reinitiating sports were asked to provide the specific sports and the dates of restarting, number of athletes, number of practices and games, and number of COVID-19 cases among athletes in each sport, as well as the self-reported sources of infections (if known) during September 2020. They were also asked about their type of instruction during September (virtual or in person) and whether they required the use of face masks for athletes while playing. Schools were included if they had any sport that had restarted participation during September 2020.

Data were initially evaluated using descriptive statistics, including estimates of central tendency (mean, median) and variability (SD, interquartile range, range) for continuous variables, and counts and percentages for categorical variables. Reported COVID-19 case rates were expressed as the number of reported cases per 100 000 athletes (cases/total number of athletes × 100 000) overall and for each sport. The duration of participation for each sport at each school was determined as the difference in days between the date of restarting and October 1, 2020, and player-days were determined as the product of the number of participating athletes and duration. Reported COVID-19 incidence rates were expressed as the number of reported cases per 100 000 player-days (cases/total number of player-days × 100 000) overall and for each sport, with CIs calculated using an exact method.

In addition, the number of cases, total population, case rate, and incidence rate during September were determined for the general population for each county in which a respondent high school was located, using publicly available online information from the respective county's health authority. To determine whether local county COVID-19 case rates were associated with reported COVID-19 case rates among high school athletes, the total number of high school athletes and reported COVID-19 cases were aggregated by county. For those counties with more than 100 athletes, the relationship between COVID-19 case rates among high school athletes and the general population was evaluated with a linear model weighted for the total population of each county.

For those sports with data from 50 or more schools, the relative risk of each sport was evaluated using a mixed-effects negative binomial regression model to predict the number of COVID-19 cases for each team with the local county incidence, school instructional delivery type (virtual versus in person), and sport as fixed effects, the log of player-days as an offset (to model the rate of cases when sampling time differed), and school as a random effect, yielding an incidence rate ratio (IRR) with cheer or dance as the reference (because this represented the median unadjusted incidence rate). The local county incidence was included to account for the background incidence of COVID-19 in the community, and school was included as a random variable to account for repeated measures among different sports from the same school. To evaluate the relationship between the reported COVID-19 incidence and sport characteristics, a multivariable negative binomial regression model was calculated to predict the number of cases, with the local county incidence, sport location (indoor, outdoor), sport contact (contact, noncontact), sport type (team, individual), and school instructional delivery type as covariates, and the log of player-days as an offset.

To evaluate the association between overall COVID-19 incidence and reported face mask use, incidence rates and 95% CIs were calculated for each sport with >50 schools for those reporting face mask use or no use. Separate multivariable negative binomial regression models were then developed to predict the number of cases, with the local county incidence, instructional delivery type, and face mask use (yes or no) as covariates, and the log of player-days as an offset. Coefficients from the models were exponentiated to represent the IRRs for binary variables, and Wald CIs were generated. Significance level was determined a priori at the .05 level, and all tests were 2 tailed. All statistical analyses were performed in R (version 4.0.3).

Of 831 high schools in Wisconsin, 244 submitted complete survey responses. A total of 207 of 415 schools that had restarted a fall sport, per the Wisconsin Interscholastic Athletic Association, responded to our survey and were included in the study, for a 50% response rate. These schools represented 30 074 student-athletes who had participated in 16 898 practices and 4378 games. In-person instruction was in place at 187 schools (90.3%) during September 2020. Among the 207 schools that had restarted sports participation, 270 cases of COVID-19 were reported, yielding a case rate of 898 cases per 100 000 athletes and an incidence rate of 32.6 (95% CI = 28.9, 36.8) cases per 100 000 player-days. From September 6 to October 4, 2020, 2390 cases of COVID-19 were reported among 14- to 17-year-olds in Wisconsin, for a case rate of 1067 cases per 100 000 people and an incidence of 38.1 cases per 100 000 person-days.¹⁹  Of the 209 cases with a self-reported known source, 115 (55%) were attributed to household contact, followed by community contact outside sport or school (n = 85, 41%), school contact (n = 5, 2.4%), sport contact (n = 1, 0.5%), and other (n = 3, 1.4%). For those sports with >50 participating schools, the incidence rate ranged from 13.3 (girls' tennis) to 45.2 cases per 100 000 player-days (football), as shown in Figure 1 (full data shown in Supplemental Table 1, available online at http://dx.doi.org/10.4085/1062-6050-0185.21.S1).

When aggregated by county, athletes' case rates demonstrated a moderate positive association with the case rates for their county's general population (β = 1.14 ± 0.20, r = 0.60, P < .001; Figure 2). The IRRs for specific sports, adjusted for the state COVID-19 incidence, instruction delivery type, and repeated measures from the same school are provided in Figure 3. The IRRs for school instructional delivery and sport characteristics are presented in Table 1. A total of 173 schools (84%) reported face mask use by players while playing sports. The unadjusted incidences for teams with or without reported face mask use in each sport with >50 respondent schools are shown in Figure 4. After adjusting for local county COVID-19 incidence and school instructional delivery, face mask use was not associated with a decreased COVID-19 incidence in cross-country, football, boys' soccer, or girls' volleyball (Table 2).

In this statewide survey study of high school athletic directors representing their high school athletes, we did not identify a statistically significant association between the COVID-19 incidence and sport type after adjusting for local virus incidence and school instructional delivery. Girls' tennis, girls' golf, and cross-country displayed the lowest adjusted incidence rates of COVID-19, while football had the highest rate. Nonetheless, the CIs around these estimates were wide, perhaps due to the relatively low incidence of COVID-19 in Wisconsin during this 1-month study period. No independent, statistically significant differences existed in the reported COVID-19 incidences between indoor versus outdoor sports, contact versus noncontact sports, or individual versus team sports. However, the values for indoor and contact-sport classification approached significance, suggesting that this relationship may become clearer with a larger sample during a longer study period or a period of higher COVID-19 incidence. In fact, a recent preprint publication¹²  involving high school athletes nationwide demonstrated this relationship over a longer period when background COVID-19 rates were higher throughout the country.

We found that case rates of reported COVID-19 in high school athletes were significantly related to the case rates of their county's general population. Although direct comparisons between our data and public health data are limited, the case and incidence rates among high school athletes in this study appeared similar to the overall COVID-19 incidence rate among 14- to 17-year-olds in Wisconsin during roughly the same timeframe. In addition, 96% of the cases among athletes had a reported source of infection from household and community contacts, with only 1 (0.5%) from a known sport contact. Thus, the background local COVID-19 incidence may have a greater effect on overall COVID-19 incidence among high school athletes than participation in a specific sport or type of sport.

Reported face mask use among the sports with the largest number of respondent schools (cross-country, football, boys' soccer, and girls' volleyball,) did not have a significant relationship with COVID-19 incidence. This finding was consistent with results among outdoor sports in a recent nationwide sample of high school athletes,¹²  yet it contrasted with those authors' observation that face mask use was associated with a decreased COVID-19 incidence among indoor sports. The overwhelming majority of athletic directors at the respondent schools in our investigation (84%) reported face mask use while on the field or court. It is possible that the imbalanced group sizes and relatively low number of cases among the included athletes may have limited our ability to identify a true relationship between face mask use and reported COVID-19 incidence. Although not statistically significant, the incidence rates for football and boys' soccer were higher among those that reported face mask use than those that did not. It is unclear whether this is attributable to the reasons stated earlier or to confounding characteristics of the individuals participating in these sports that were not accounted for in the data collected.

Similarly, we did not find a relationship between the COVID-19 incidence and the type of instructional delivery (in person or virtual). Of the cases with a reported known source, only 2.5% were attributed to a school contact. This was consistent with prior indications that schools were not significant contributors to the spread of COVID-19.²⁰  However, 90% of our respondent schools described in-person instruction, making it difficult to fully evaluate the role of in-person school instruction in the COVID-19 incidence among high school athletes. Nonetheless, we included school instruction type within our adjusted models to account for this potential confounder. However, it is important to recognize that we could not account for the transmission or incidence of COVID-19 among attendees at high school sporting events beyond the athletes. This route represents a potentially significant contributor to the community COVID-19 risk, and risk mitigation procedures should continue to be prioritized to protect both athletes and attendees.

Our study had several limitations. We were unable to verify the information provided by the athletic directors via a separate independent source. Local county-level daily COVID-19 case data were frequently unavailable for adolescents or children, so our adjusted models could account only for the general population-level, background incidence in each county. Nonetheless, we found that reported case rates in our sample and case rates from the county general populations were positively related. It is difficult to fully interpret comparisons of COVID-19 cases reported by athletic directors and those collected by public health agencies, but we included public health data to add context for our findings and to adjust our incidence models. We did not ask whether respondent high schools were public or private, and we could not account for differences in socioeconomic status among schools that could have influenced outcomes. In addition, we were unable to consider individual-level differences between athletes in different sports or between teams that did or did not report face mask use that might have influenced our results. As mentioned earlier, the incidence of COVID-19 was relatively low during September 2020 in Wisconsin, and this may have limited our ability to detect statistically significant associations in certain situations. Reported sources of infection were provided by the schools themselves, and whether these represented the results of formal contact tracing by local health authorities is unknown. Finally, these data reflect information regarding athletes in a single state and may not be generalizable to other populations.

After adjusting for local county COVID-19 incidence, we identified no statistically significant differences in reported COVID-19 incidence among sports or sport types among Wisconsin high school athletes during September 2020. Reported COVID-19 case rates among athletes were highly correlated with case rates for the general population for their respective counties, and incidence rates among athletes were very similar to those of 14- to 17-year-olds in Wisconsin in general. Most cases among athletes were attributed to household and community contact, with very few attributed to school or sport contacts. Further research is warranted to better define the risk factors for COVID-19 transmission during adolescent sport participation and the relative benefits of different risk mitigation strategies.

Supplemental Table.

Found at DOI: http://dx.doi.org/10.4085/1062-6050-0185.21.S1