key: cord-0806647-iynii125 authors: Jones, R. D.; Romer, J. E. title: COVID19 Trends in Florida, August 10 through October 3, 2020 date: 2020-12-07 journal: nan DOI: 10.1101/2020.11.24.20235861 sha: 20b5420ae9875bcba57847cec7198948593541d2 doc_id: 806647 cord_uid: iynii125 Student enrollment across Floridas 67 districts totals more than 2.67 million, with five of the ten most populated districts in the country in the state. Analysis of 3,451 laboratory-confirmed cases of SARSCoV2 in Florida K-12 schools during August 10 through October 3, 2020 might inform decisions about continued in-person and virtual learning access and promotion across the United States, as well as mitigation measures related to reopening policies across districts responsive to model results. These results also provide a baseline for monitoring trends and evaluating mitigation strategies. data in K-12 schools in July 2020, and has since reported more than 200,000 confirmed cases of the virus in the K-12 environment nationwide -the largest date repository for cases in schools. The Covid Monitor's national database allows for analysis of the characteristics and trends of cases in schools over time. The results may inform decisions about continued in-person and virtual learning access and promotion across the United States, as well as mitigation measures related to reopening policies across districts responsive to model results. These results also provide a baseline for monitoring trends and evaluating mitigation strategies. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted December 7, 2020. For Model A, three independent variables were used to model the count of total cases within a district: 1) percent in-person enrollment; 2) cases per 1,000 per week within the county population for the same time period; and 3) total district enrollment and staff (squared-root transformed). The estimated number of total cases and case rates over a 30-day period for a district with 25,000 students ( Figure 1 ) considered community incidence rates ranging from 0.5 to 3.0 cases per 1,000 per week. To adjust for the varying exposure time, an offset term of log (days) was included in the model (see estimated regression coefficients in Table 2 ). . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted December 7, 2020. Cases related to schools reported by the Florida Department of Health (DOH) were used to model the effect of the same three variables on school incidence rate. Unlike the districtprovided data, the DOH data was collected over a period of 28 days, from 9/6/2020 -10/3/2020. The DOH report, which included data for all 67 Florida counties, under-reported the case data provided by 41 of the state's districts by more than 1,238 confirmed cases. The DOH data used here for comparison purposes is not representative of case totals in K-12 schools in the state during the report period, but it is the only data made available by the state. Only districts with known in-person enrollment were included in the state-portion of the analysis, and schools that had not yet begun in-person instruction were excluded (Miami-Dade and Broward). Each school district included in the analysis was open the entire period but for varying lengths of time prior to 9/6/2020. To adjust for this, 'Days Open' is included in the . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted December 7, 2020. ; https://doi.org/10.1101/2020.11.24.20235861 doi: medRxiv preprint P a g e 6 o f 8 model and is calculated as the number of days between 9/6/2020 and the first day of school. The variable for total persons (district enrollment and staff) was log-transformed as opposed to the square-root in the previous model. R (version 4.0.2) was used to conduct all analyses. During August 10 -October 1, 2020, average incidence (cases per 1,000 students enrolled in face-to-face instruction) in Florida high school students (4.5) was nearly twice that of elementary students (2.4). Overall staff rates (5.1) were more than twice the overall student rate (2.3). Staff case rates were highest at the elementary (4.5) and middle-school (4.8) level, and only lower than student incidence at the high-school level (4.2). The proportion of student to staff cases in Florida schools was closest in the elementary setting (59% students -41% staff) compared to the high school setting (80% students -20% staff). A 1 percentage-point increase of in-person students (e.g. 60% to 61%) would have an estimated increase in district-wide student and staff case rates of about 2.1% (Model A, Table 2 ). For each increase in 1 case per 1,000 per week in the community, the average rate within schools is estimated to increase by more than 41%. For each increase of one unit of the squared-root of total persons (staff and in-person students), the expected rate increases by 0.7%. The total number of expected cases in K-12 schools was lower for 90% in-person enrollment at low levels of community infection (0.5 cases per 1,000 per week) than for half that amount (48%) in-person enrollment at much higher community incidence rates (3.0 cases per 1,000 per week). Thus, the level of infection within the community appears to be the primary factor that influences rate of infections in school, with the percent of in-person students by enrollment total a significant secondary influencer on K-12 case incidence. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted December 7, 2020. ; https://doi.org/10.1101/2020.11.24.20235861 doi: medRxiv preprint P a g e 7 o f 8 FIGURE 1: Influence of percent in-person attendance and COVID-19 community case rate on case totals and case incidence in K-12 school. The results of the regression (Model B, Table 2 ) confirm the significant effect of both inperson percentage of students and community case rate on case incidence in schools. Differences in the coefficient estimates are likely due to different observation periods, the counties included in each analysis, and actual differences in the number of confirmed cases by DOH and the districts themselves. Despite the different model estimates, the conclusion is that lower in-person enrollment and community infection are important factors in the number of school cases. Case incidence varies significantly between school grade levels and between students and staff. Staff rates are higher than student rates in all school environments except high schools. The rate of cases within schools is highly correlated with cases within a community, more than the size of the district by total enrollment. Percent enrollment in face-to-face instruction is a secondary influencer of case incidence rates in schools. For example, a 10 percentage-point increase in in-person enrollment would increase the school district's case rate by 22.5%. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted December 7, 2020. ; https://doi.org/10.1101/2020.11.24.20235861 doi: medRxiv preprint P a g e 8 o f 8 These findings can provide a baseline for monitoring national trends. A progressive, stratified reintroduction to in-person learning in communities with low case incidence rates could provide for a significantly lower risk of transmission in students, staff, and communities. Along with the implementation of mitigation methods proposed by the CDC, distance learning until community transmission is largely under control could have a profound impact on COVID-19 in the K-12 environment. Once community transmission reaches advised levels, a phased re-entry to in-person learning, especially at the high school level, could prevent further spread within the school setting. COVID-19 Trends Among School-Aged Children -United States United States Centers for Disease Control. 2020. Indicators for Dynamic School Decision-Making The Covid Monitor provided data related to cases in K-12 school districts in Florida.