Globally, coronavirus disease 2019 (COVID-19) has affected how children learn. We evaluated the impact of Test to Stay (TTS) on secondary and tertiary transmission of severe acute respiratory syndrome coronavirus 2 and potential impact on in-person learning in 4 school districts in the United States from September 13 to November 19, 2021.
Implementation of TTS varied across school districts. Data on index cases, school-based close contacts, TTS participation, and testing results were obtained from 4 school districts in diverse geographic regions. Descriptive statistics, secondary and tertiary attack risk, and a theoretical estimate of impact on in-person learning were calculated.
Fifty-one schools in 4 school districts reported 374 coronavirus disease COVID-19 index cases and 2520 school-based close contacts eligible for TTS. The proportion participating in TTS ranged from 22% to 79%. By district, the secondary attack risk and tertiary attack risk among TTS participants ranged between 2.2% to 11.1% and 0% to 17.6%, respectively. Nine clusters were identified among secondary cases and 2 among tertiary cases. The theoretical maximum number of days of in-person learning saved by using TTS was 976 to 4650 days across jurisdictions.
TTS preserves in-person learning. Decisions to participate in TTS may have been influenced by ease of access to testing, communication between schools and families, testing logistics, and school resources. Tertiary attack risk determination became more complicated when numbers of close contacts increased. Minimizing exposure through continued layered prevention strategies is imperative. To ensure adequate resources for TTS, community transmission levels should be considered.
Test to Stay (TTS) allows asymptomatic close contacts to someone with coronavirus disease 2019, who are not fully vaccinated, to remain in school with regular testing. Previous studies suggest that TTS maintains low school-associated severe acute respiratory syndrome coronavirus 2 transmission.
Our study suggests that TTS preserves in-person learning and highlights potential challenges to TTS implementation. If implemented with layered prevention strategies and careful attention to equity, TTS can be an alternative to at-home quarantine.
The coronavirus disease 2019 (COVID-19) pandemic has profoundly impacted school-aged children. Loss of in-person learning because of quarantine after exposure to a COVID-19 case disproportionately impacts underresourced communities, worsens mental health indicators, and decreases access to tools and support mechanisms, such as school-based meals, mental health services, and the Internet.1–3 To mitigate these effects, some school districts have implemented a modified quarantine approach called Test to Stay (TTS). TTS allows school-based close contacts of COVID-19 cases, who are not fully vaccinated, to remain in school if they are asymptomatic, adhere to regular testing, test negative, and follow specified prevention measures.4
In spring 2021, before younger ages were eligible for vaccination, an open-label randomized trial assigned a subset of schools in the United Kingdom to participate in TTS, which allowed students enrolled in secondary schools and institutions of higher education who came into close contact to someone with COVID-19 to continue in-person learning if they underwent daily severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) testing in place of a 10 day home quarantine.5 Although subject to limitations, results suggested that daily testing of school-based close contacts was noninferior to traditional quarantine. A US study looking at elimination of student quarantine after mask-on-mask exposures to COVID-19 did not identify any associated secondary transmission.6 Two studies of TTS in Lake County, Illinois, and Los Angeles, California, found that TTS did not appear to increase SARS-CoV-2 transmission risk in the kindergarten through grade 12 (K-12) setting and identified substantial numbers of in-person learning days saved.7,8
TTS protocols in the United States vary widely,7,8 and the optimal characteristics of TTS are not yet known. In the context of unique TTS protocols and resource constraints, we evaluated secondary and tertiary transmission of SARS-CoV-2 associated with TTS, clusters among TTS participants, and TTS impact on in-person learning days in 4 geographically distinct public school districts in the United States from September 13 to November 19, 2021. This is the first paper combining lessons from TTS implementation in various school districts across the country.
Methods
The Centers for Disease Control and Prevention (CDC) collaborated with 4 states to evaluate TTS impact on SARS-CoV-2 transmission in pre–K-12 settings between September 13, 2021, and November 19, 2021. The investigation included 1 public school district each from Georgia, Illinois, Kentucky, and New Mexico, for a total of 4 school districts. All schools within the districts in Georgia, Illinois, and New Mexico, and a representative sample of schools in Kentucky, representing 27% of elementary and middle schools, implemented TTS, for a total of 51 participating schools.
TTS implementation varied across the school districts by school type, eligibility criteria, SARS-CoV-2 test type, and testing logistics (Table 1). District TTS protocols often mirrored those of existing state guidelines; however, districts occasionally implemented more restrictive TTS guidelines, primarily differing on eligibility, testing cadence, testing location, and participation in extracurricular or after-school programs. In all 4 districts, TTS-eligible close contacts were defined as individuals not fully vaccinated who were within 3 ft of a COVID-19 case, for a cumulative total of ≥15 minutes over 24 hours if masked; 2 districts allowed participation of close contacts exposed within 6 ft if unmasked. Once enrolled in TTS, participants in all districts were required to mask. Students in grades pre–K-12, teachers, and staff were eligible in 3 districts. Kentucky limited eligibility to students in grades kindergarten through grade 8. In all 4 districts, only close contacts exposed in the school setting were eligible for TTS; individuals identified as household contacts of COVID-19 cases were excluded. Siblings of TTS participants testing negative remained in school.
TTS Criteria . | Georgia . | Kentucky . | Illinois . | New Mexico . |
---|---|---|---|---|
Close contact definition | <3 ft for a cumulative total of 15 min over 24 h (if masked); <6 ft for a cumulative total of ≥15 min over 24 h (if unmasked) | <3 ft for a cumulative total of ≥15 min over 24 h (if masked); <6 ft for a cumulative total of ≥15 min over 24 h (if unmasked) | <3 ft for a cumulative total of ≥15 min over 24 h | <3 ft for a cumulative total of ≥15 min over 24 h |
Exposure type | School exposure only | School exposure only | School exposure only | School exposure only |
Mask criteria for TTS | Masked and unmasked exposures eligible | Masked and unmasked exposures eligible | Masked exposure only | Masked exposure only |
Who is eligiblea | Students, teachers, and staff | Elementary and middle school students | Students, teachers, and staff | Students, teachers, and staff |
Close contact testing cadence | Every school d for 7 d | Every other weekday for 7 d | D 1, 3, 5, and 7 | D 1, 3, 5, and 7 |
Test type | Rapid antigen (CareStart) | Rapid RT-PCR (Thermo Fisher Scientific Accula System) | Rapid antigen (BinaxNOW) | Rapid RT-PCR (Thermo Fisher Scientific Accula System) |
Testing location | 1 centralized location | Individual schools | Individual schools | Individual schools |
Participation in school-sponsored, after-school programs (eg, sports and extracurriculars) allowed while in TTS | Yes | Yes | Yes | Yes |
School quarantine policyb | 7 d with test between days 5 and 7 or 10 d if symptom-free | 7 d with test between days 5 and 7 or 10 d if symptom-free | 7 d with test after d 6 or 14 d if symptom-free | 10 d if symptom-free with no testing-out policy |
Evaluation dates | September 13–November 5 (8 wk) | September 27–November 19 (8 wk) | October 29–November 19 (3 wk) | October 28–November 19 (3 wk) |
TTS Criteria . | Georgia . | Kentucky . | Illinois . | New Mexico . |
---|---|---|---|---|
Close contact definition | <3 ft for a cumulative total of 15 min over 24 h (if masked); <6 ft for a cumulative total of ≥15 min over 24 h (if unmasked) | <3 ft for a cumulative total of ≥15 min over 24 h (if masked); <6 ft for a cumulative total of ≥15 min over 24 h (if unmasked) | <3 ft for a cumulative total of ≥15 min over 24 h | <3 ft for a cumulative total of ≥15 min over 24 h |
Exposure type | School exposure only | School exposure only | School exposure only | School exposure only |
Mask criteria for TTS | Masked and unmasked exposures eligible | Masked and unmasked exposures eligible | Masked exposure only | Masked exposure only |
Who is eligiblea | Students, teachers, and staff | Elementary and middle school students | Students, teachers, and staff | Students, teachers, and staff |
Close contact testing cadence | Every school d for 7 d | Every other weekday for 7 d | D 1, 3, 5, and 7 | D 1, 3, 5, and 7 |
Test type | Rapid antigen (CareStart) | Rapid RT-PCR (Thermo Fisher Scientific Accula System) | Rapid antigen (BinaxNOW) | Rapid RT-PCR (Thermo Fisher Scientific Accula System) |
Testing location | 1 centralized location | Individual schools | Individual schools | Individual schools |
Participation in school-sponsored, after-school programs (eg, sports and extracurriculars) allowed while in TTS | Yes | Yes | Yes | Yes |
School quarantine policyb | 7 d with test between days 5 and 7 or 10 d if symptom-free | 7 d with test between days 5 and 7 or 10 d if symptom-free | 7 d with test after d 6 or 14 d if symptom-free | 10 d if symptom-free with no testing-out policy |
Evaluation dates | September 13–November 5 (8 wk) | September 27–November 19 (8 wk) | October 29–November 19 (3 wk) | October 28–November 19 (3 wk) |
RT-PCR, reverse transcription polymerase chain reaction.
Students, teachers, and staff who were fully vaccinated or who received a COVID-19 diagnosis in the 90 days before exposure were exempt from quarantine and, therefore, not identified for TTS participation.
Quarantine period was calculated using date of exposure as quarantine start date.
TTS-eligible individuals with close contact to a COVID-19 case were tested regularly for 7 days after exposure (testing cadence varied by site, Table 1) and could remain in school if tests were negative, they remained asymptomatic, and adhered to prevention measures. Student participation required parental consent. Georgia and Illinois used rapid antigen tests; Kentucky and New Mexico used point-of-care rapid reverse transcription-polymerase chain reaction tests, which provided results in 30 minutes. Test samples were collected at schools in Kentucky, Illinois, and New Mexico, and at a single centralized site in Georgia. TTS participants were instructed to quarantine at home while not at school or not participating in school-based activities during the TTS period. If close contacts were not eligible or did not consent to TTS, districts in Georgia, Kentucky, and New Mexico required close contacts to quarantine at home for 10 days without a test. Georgia and Kentucky also allowed close contacts to quarantine at home for 7 days with a negative test between days 5 and 7. Illinois required close contacts to quarantine at home for 7 days with a negative test after day 6 or for 14 days without a test. Students, teachers, and staff who were fully vaccinated or who received a COVID-19 diagnosis in the 90 days before exposure were exempt from quarantine and not eligible for TTS participation. At the time of this investigation, students and staff were considered fully vaccinated ≥14 days after receipt of the second dose in a 2-dose series (Pfizer-BioNTech or Moderna COVID-19 vaccines) or after 1 dose of the single-dose Janssen (Johnson & Johnson) COVID-19 vaccine.
Community incidence rates during the evaluation period were obtained from the CDC COVID Data Tracker,9–11 and district demographics were obtained from publicly available sources where possible.12–17 School administrators also completed a survey on school demographics and COVID-19 mitigation measures. We obtained data on index cases, school-based close contacts (including exposure, location, and dates), TTS participation, and testing results from the school districts from testing contractors and/or local health departments.
Secondary and tertiary transmission were analyzed in relation to TTS participation. We defined secondary cases as school-based close contacts not fully vaccinated who received a positive SARS-CoV-2 test within 14 days after exposure to an index case, and tertiary cases as school-based close contacts not fully vaccinated who received a positive SARS-CoV-2 test within 14 days after exposure to a TTS participant who tested positive. We interviewed secondary and tertiary cases or their parent/guardian by phone about activities and potential exposures during the TTS period to determine whether school-associated transmission was probable, possible, or unlikely (Supplemental Information).18 If we were unable to contact secondary or tertiary cases or their parent/guardian and information could not be obtained from the school, we considered transmission undetermined and assumed possible school-based transmission. We defined secondary attack risk (SAR) and tertiary attack risk (TAR) as the number of probable or possible school-based secondary or tertiary cases divided by the total number of close contacts of index cases or secondary cases, respectively. We used the Clopper-Pearson method to calculate confidence intervals (CI) at the 95% level. Secondary and tertiary clusters were identified using the Council of State and Territorial Epidemiologists’s definition of at least 3 school-associated COVID-19 cases within a specified core group who develop symptoms or test positive within 14 days of each other.19
We calculated the theoretical maximum number of in-person learning days saved from TTS, assuming 8 missed school days for every 10 day quarantine in 3 school districts and 10 missed school days for every 14 day quarantine in 1 district. Each round of TTS an individual participated in was considered to have saved the maximum number of in-person learning days for that district, 8 and 10, respectively. We did not account for holidays, exposure notification lag time, shortened quarantine based on test results and lack of symptoms, or home quarantine time of those who switched from TTS to home quarantine midway through their TTS period.
All analyses were performed using R (version 4). This activity was reviewed by the CDC and was conducted consistent with applicable federal law and CDC policy (See, for example, 45 C.F.R. part 46.102(l)(2), 21 C.F.R. part 56; 42 U.S.C. §241(d); 5 U.S.C. §552a; 44 U.S.C. §3501 et seq.).
Results
During September and November 2021, 51 public schools in 4 school districts with 28 373 enrolled students and 4404 teachers and staff implemented TTS. Among the 51 schools, there were 3 (6%) early childhood education schools (pre-K), 30 (59%) elementary schools, 12 (24%) middle schools, 3 (6%) high schools, and 3 (6%) alternative schools (Table 2). On review of each district’s home county 7-day rolling average community incidence rates during the evaluation period, rates were lowest in Illinois (median 39.6 per 100 000; range: 24.9–76.4 per 100 000) and highest in Georgia (median 178.6 per 100 000; range: 75.4–550.9 per 100 000).10,11 Nonpharmaceutical interventions required or recommended by each school district and district demographics are outlined in Table 2. Notably, a mask mandate was in place in all 4 school districts at the start of the fall semester, but was lifted in the Georgia school district on October 16, 2021, midway through this investigation.
. | Georgia . | Kentucky . | Illinois . | New Mexico . | ||
---|---|---|---|---|---|---|
Nonpharmaceutical interventions recommended or required | ||||||
Handwashing | Yes | Yes | Yes | Yes | ||
Regular cleaning and disinfection | Yes | Yes | Yes | Yes | ||
Increased ventilation | Yes | Yes | Yes | Varied | ||
Physical barriers in the classroom | No | No | No | No | ||
Student cohortinga | Varied | Varied | No | Varied | ||
Outdoor instruction | No | Yes | No | Varied | ||
Desk spacing >3 ft | Varied | No | Yes | Varied | ||
Desk spacing >6 ft | No | No | No | Varied | ||
School extracurricular activities | Yes | Yes | Yes | Yes | ||
Mask mandate | Mask optional beginning October 16, 2021 | Yes | Yes | Yes | ||
Screening | ||||||
Screening testingb | Yes (began late October) | No | Yes | Yes (for staff not fully vaccinatedc) | ||
Symptom screening | Yes | No | Yes | Yes | ||
Community incidence | ||||||
Median 7-d rolling average community incidence per 100 000 during evaluation period9 | 178.6 (75.4–550.9) | 67.6 (42.9–176.7) | 39.6 (24.9–76.4) | 51.6 (40.1–69.1) | ||
Number of participating schools | ||||||
Early childhood education (Pre-K) | 1 | 0 | 1 | 1 | ||
Elementary (K-5) | 8 | 8 | 5 | 9 | ||
Middle (6–8) | 2 | 5 | 2 | 3 | ||
High school (9–12) | 1 | 0 | 1 | 1 | ||
Alternative schools | 1 | 0 | 1 | 1 | ||
School district demographics13–17 | ||||||
Urban or rurald | Urban | Urban | Urban | Urban | ||
Hispanic students, % | 39 | 19 | 2 | 40 | ||
White, non-Hispanic students, % | 20 | 46 | <1 | 44 | ||
Black, non-Hispanic students, % | 35 | 23 | 97 | 6 | ||
Asian American/Pacific Islander, non-Hispanic students, % | 2 | 5 | <1 | 3 | ||
American Indian/Alaskan Native, non-Hispanic students, % | <1 | <1 | <1 | 1 | ||
Multiple races, non-Hispanic students, % | 4 | 6 | <1 | 6 | ||
Students who qualify for free or reduced lunch, median % (range) | 40 (14–86) | 57 (14–86) | 100 (100–100) | 77 (6–100) |
. | Georgia . | Kentucky . | Illinois . | New Mexico . | ||
---|---|---|---|---|---|---|
Nonpharmaceutical interventions recommended or required | ||||||
Handwashing | Yes | Yes | Yes | Yes | ||
Regular cleaning and disinfection | Yes | Yes | Yes | Yes | ||
Increased ventilation | Yes | Yes | Yes | Varied | ||
Physical barriers in the classroom | No | No | No | No | ||
Student cohortinga | Varied | Varied | No | Varied | ||
Outdoor instruction | No | Yes | No | Varied | ||
Desk spacing >3 ft | Varied | No | Yes | Varied | ||
Desk spacing >6 ft | No | No | No | Varied | ||
School extracurricular activities | Yes | Yes | Yes | Yes | ||
Mask mandate | Mask optional beginning October 16, 2021 | Yes | Yes | Yes | ||
Screening | ||||||
Screening testingb | Yes (began late October) | No | Yes | Yes (for staff not fully vaccinatedc) | ||
Symptom screening | Yes | No | Yes | Yes | ||
Community incidence | ||||||
Median 7-d rolling average community incidence per 100 000 during evaluation period9 | 178.6 (75.4–550.9) | 67.6 (42.9–176.7) | 39.6 (24.9–76.4) | 51.6 (40.1–69.1) | ||
Number of participating schools | ||||||
Early childhood education (Pre-K) | 1 | 0 | 1 | 1 | ||
Elementary (K-5) | 8 | 8 | 5 | 9 | ||
Middle (6–8) | 2 | 5 | 2 | 3 | ||
High school (9–12) | 1 | 0 | 1 | 1 | ||
Alternative schools | 1 | 0 | 1 | 1 | ||
School district demographics13–17 | ||||||
Urban or rurald | Urban | Urban | Urban | Urban | ||
Hispanic students, % | 39 | 19 | 2 | 40 | ||
White, non-Hispanic students, % | 20 | 46 | <1 | 44 | ||
Black, non-Hispanic students, % | 35 | 23 | 97 | 6 | ||
Asian American/Pacific Islander, non-Hispanic students, % | 2 | 5 | <1 | 3 | ||
American Indian/Alaskan Native, non-Hispanic students, % | <1 | <1 | <1 | 1 | ||
Multiple races, non-Hispanic students, % | 4 | 6 | <1 | 6 | ||
Students who qualify for free or reduced lunch, median % (range) | 40 (14–86) | 57 (14–86) | 100 (100–100) | 77 (6–100) |
K-5, kindergarten through grade 5.
Cohorting means keeping people together in a small group and having that group stay together throughout the day. Whether there was cohorting varied by grade level, school, and/or class type.
Opt-in testing (eg, once a week) of all students, teachers, or staff who are asymptomatic and do not have known, suspected, or reported exposure to SARS-CoV-2; includes those who are fully vaccinated unless otherwise stated.
At the time of this investigation, staff were considered fully vaccinated ≥14 days after receipt of the second dose in a 2-dose series (Pfizer-BioNTech or Moderna COVID-19 vaccines) or after 1 dose of the single-dose Janssen (Johnson & Johnson) COVID-19 vaccine.
Urban/rural designation determined using the 2010 US Census Bureau criteria. Urban designation includes both urbanized areas and urban clusters.12
Across the 4 districts, there were 374 index COVID-19 cases (ranging from 56 to 163 per district) and 2520 school-based close contacts eligible for TTS (ranging from 559 to 783 per site) (Table 3). Among the 374 index COVID-19 cases, 52 (13.9%) were teachers or staff and 322 (86.1%) were students, and among the 2520 close contacts, 37 (1.5%) were teachers or staff and 2483 (98.5%) were students. Index cases were identified either through screening testing or other school- or community-based testing (symptomatic testing, routine testing for medical procedures, or other reasons). TTS participation among eligible close contacts was lowest in Georgia (22%) and highest in Illinois (79%). Among TTS participants, secondary COVID-19 cases were identified: 9 in Illinois, 12 in Georgia, 17 in Kentucky, and 30 in New Mexico (Supplemental Table 4). SAR ranged from 2.2% (95% CI 1.0%–4.2%) in Illinois to 11.1% (95% CI 5.9%–18.6%) in Georgia (Table 3). There were 9 clusters among secondary cases: 1 each in Illinois and New Mexico, 2 in Georgia, and 5 in Kentucky. The majority (8 of 9, 88.9%) of secondary clusters were in elementary schools, and 1 (1 of 9, 11.1%) was in a middle school.
. | Georgia . | Kentucky . | Illinois . | New Mexico . |
---|---|---|---|---|
Index cases | 92 | 56 | 63 | 163 |
TTS | ||||
Close contacts identifieda | 732 | 935 | 589b | 779 |
Close contactsc eligible | 559 | 783 | 589 | 589 |
Close contactsd enrolled | 122 | 415 | 465 | 392 |
% participation | 21.8 | 53.0 | 78.9 | 66.6 |
Index case close contactse | 108 | 404 | 405 | 445 |
Secondary cases | 12 | 17 | 9 | 30 |
SARf (95% CI) | 11.1% (5.9%–18.6%) | 4.2% (2.5%–6.7%) | 2.2% (1.0%–4.2%) | 6.7% (4.6%–9.5%) |
Secondary case contactse | 17 | 69 | 148 | 74 |
Tertiary cases | 1 | 1 | 0 | 13 |
TARf (95% CI) | 5.9% (0.1%–28.7%) | 1.4% (0%–7.8%) | 0% (0%–2.5%) | 17.6% (9.7%–28.2%) |
Tertiary case close contactse | 0 | 0 | 0 | 37 |
Clustersg | ||||
Secondary clusters | 2 | 5 | 1 | 1 |
Tertiary clusters | 0 | 0 | 0 | 2 |
Theoretical maximum d of in-person learning savedh | ||||
TTS eligible contacts | 4472 | 6264 | 5890 | 4712 |
TTS enrolled contacts | 976 | 3320 | 4650 | 3136 |
. | Georgia . | Kentucky . | Illinois . | New Mexico . |
---|---|---|---|---|
Index cases | 92 | 56 | 63 | 163 |
TTS | ||||
Close contacts identifieda | 732 | 935 | 589b | 779 |
Close contactsc eligible | 559 | 783 | 589 | 589 |
Close contactsd enrolled | 122 | 415 | 465 | 392 |
% participation | 21.8 | 53.0 | 78.9 | 66.6 |
Index case close contactse | 108 | 404 | 405 | 445 |
Secondary cases | 12 | 17 | 9 | 30 |
SARf (95% CI) | 11.1% (5.9%–18.6%) | 4.2% (2.5%–6.7%) | 2.2% (1.0%–4.2%) | 6.7% (4.6%–9.5%) |
Secondary case contactse | 17 | 69 | 148 | 74 |
Tertiary cases | 1 | 1 | 0 | 13 |
TARf (95% CI) | 5.9% (0.1%–28.7%) | 1.4% (0%–7.8%) | 0% (0%–2.5%) | 17.6% (9.7%–28.2%) |
Tertiary case close contactse | 0 | 0 | 0 | 37 |
Clustersg | ||||
Secondary clusters | 2 | 5 | 1 | 1 |
Tertiary clusters | 0 | 0 | 0 | 2 |
Theoretical maximum d of in-person learning savedh | ||||
TTS eligible contacts | 4472 | 6264 | 5890 | 4712 |
TTS enrolled contacts | 976 | 3320 | 4650 | 3136 |
Includes all close contacts of index, secondary, and tertiary cases.
Data for exempt contacts not collected at this site.
Calculated as number of times an individual was eligible to enroll in TTS or home quarantine. These individuals may have been exposed more than once during a quarantine period; however, eligible exposures were only included if they led to a new quarantine period.
Calculated as number of times an individual enrolled in TTS. These individuals may have been exposed more than once during a quarantine period; however, enrollments are only included if they led to a new quarantine period.
Calculated using the number of times an individual had contact with a unique index, secondary, or tertiary COVID-19 case, respectively. This includes all TTS- or home quarantine-eligible exposures; individuals could have been a contact more than once. Exempt individuals who had contact with a case were excluded.
Calculated using the number of times an individual had contact with a unique index or secondary case, respectively, as the denominator. This denominator includes all TTS- or home quarantine-eligible exposures; individuals could have been a contact more than once. Exempt individuals who had contact with a case were excluded.
The number of times a case (index case or secondary case) infected ≥3 individuals.
The number of days of in-person learning a TTS participant could expect to save in comparison with the full traditional quarantine number of days of home quarantine, assuming exposure date and exposure notification date were the same and assuming the individual missed no days of school while enrolled in TTS. Full traditional quarantine varied across sites as described in Table 1.
Among school-based close contacts of TTS secondary cases, there were no possible or probable school-associated tertiary cases identified in Illinois, whereas there were 1 each in Georgia and Kentucky, and 13 in New Mexico. Six cases from New Mexico who were considered tertiary cases were also within the 14 days of exposure to an index case and may have been secondary cases rather than tertiary cases. In Illinois, there were an additional 6 cases identified who were determined to be unlikely school-associated tertiary cases because of household exposure to COVID-19 (Supplemental Table 4). The estimated school-associated TAR was 0% (95% CI 0%–2.5%) in Illinois, 1.4% (95% CI 0%–7.8%) in Kentucky, 5.9% (95% CI 0.1%–28.7%) in Georgia, and 17.6% (95% CI 9.7%–28.2%) in New Mexico (Table 3). Georgia had the lowest TTS participation rate (22%). No tertiary clusters were observed in Illinois, Kentucky, and Georgia, and 2 possible tertiary clusters were identified in New Mexico, which were both in elementary school special education classes. Assuming a maximum of 8 missed school days for every 10 day quarantine and 10 missed school days for every 14 day quarantine, TTS preserved 976 to 4650 in-person learning days for TTS participants. If all staff and students eligible for TTS had participated, a maximum of 4472 to 6264 in-person learning days could have been saved by TTS participation (Table 3).
Discussion
In 4 geographically diverse, United States school districts implementing a range of TTS protocols, TTS was an alternative to traditional quarantine and saved an estimated 976 to 4650 in-person learning days. Although other recent studies found that TTS maintained no or low school-based transmission,6–8 we found that SAR and TAR varied among school districts. Districts in Illinois and Kentucky had no or low (1%) tertiary transmission, respectively. Georgia had a higher TAR (6%), and New Mexico had the highest TAR (18%) and 2 tertiary clusters. Both tertiary clusters occurred in elementary school special education classes, where students may require closer contact or have more frequent small group interactions.20
Although some districts had a higher TAR, there were a small number of tertiary cases overall, leading to wide CIs. Also, TAR determination became more complicated when numbers of close contacts increased, making it difficult to identify source exposures. Six close contacts in New Mexico who were exposed to both index cases and secondary cases within the same 14 day period were counted as tertiary cases. If all 6 had been considered secondary cases, TAR in New Mexico would decrease to 8.0% (95% CI 3.3%–15.9%) and SAR would increase to 8.1% (95% CI 5.7%–11.0%). Hesitancy by index cases to report close contacts could have led to an overestimate of the TAR. However, there was a risk of school-based tertiary transmission in some districts that should be considered when implementing TTS.
This study was conducted at a time when delta was the predominant SARS-CoV-2 variant in the United States. Emergence of more transmissible variants causing breakthrough infections, such as omicron, will make it more difficult to document all close contacts. For schools without preexisting, robust contact tracing and access to testing resources, TTS required significant resources, leading to increased staff burden and heightened need for testing and logistics planning.7,21,22 As case counts rise, schools may be less likely to maintain contact tracing and will need additional testing resources.23
To ensure safe and equitable TTS implementation, schools require increased testing capacity, on-site testing tools, and coordinated testing communication. Resource allocation may be challenging for underresourced school districts, and care must be taken to ensure that TTS is equally affordable for both well-resourced and underresourced schools. Not only could inequitable resource distribution and support increase disparities related to in-classroom learning,24 but they have the potential to increase disparities related to food security, support for those with disabilities, and other ancillary services affecting social determinants of health impacting students’ lives.2,25
Access and communication likely influenced the lower participation in TTS in some districts. Daily TTS sample collection at 1 centralized location in Georgia may have complicated participation for parents, particularly those with inflexible work schedules and/or without transportation access. Other studies have cited testing location convenience and transportation as important factors impacting testing access.26–28 Raising awareness of the TTS program and how it works may be challenging, and success in communicating with parents may vary by school staffing and time that staff can spend reaching out to parents to explain the program and obtain consent.
COVID-19 vaccines reduce the risk of getting and spreading SARS-CoV-2.29 Fully vaccinated student and staff close contacts (or those who had been diagnosed with COVID-19 in the previous 90 days) were exempt from quarantine and TTS. Some schools did not capture data for exempt or ineligible close contacts, and schools with fewer contact tracing resources and busier staff may have been less likely to capture these close contacts. With the emergence of highly transmissible variants, school districts may consider expanding TTS testing to exempt individuals. However, expansion may only be sustainable in districts with more resources, further contributing to inequitable testing opportunities.
We found that TTS resulted in a theoretical maximum of 976 to 4650 in-person learning days saved, with the opportunity to save more in-person learning days with higher participation. Heterogeneity across sites, lack of attendance records, low testing capacity in 1 district, and lack of standardization and documentation by some testing contractors complicated an estimation of observed in-person learning days saved across sites. Thus, we report a theoretical maximum, consistent with previous studies.7,30 Also, increases in case incidence and other factors can lead to delays in staff and student exposure notification, meaning close contacts might not find out about an exposure until near the end of what would have been their quarantine period. Exposure notification delays likely led to fewer days saved by TTS participation; however, TTS was less disruptive to in-person learning compared with traditional quarantine. Prompt notification of positive cases to schools and contact tracing are important for slowing the spread of COVID-19 and preserving in-person learning.31
A limitation of this investigation is that not all test-positive contacts could be reached to obtain outside-of-school exposure information, and these individuals were grouped with possible school-associated cases. This may have resulted in an overestimate of SAR and TAR because some who were unable to be reached may have had outside-of-school exposures that were unknown to investigators or the school. Another limitation is that we were not able to collect demographic data for all participants, nor present feedback obtained on barriers to TTS implementation in a standardized fashion. We were also unable to comment upon COVID-19 infection trends in classrooms where occupancy varied over time because occupancy data were not collected in real time. Additionally, we do not have detailed data on remote learning options, so we are unable to describe impact on children of various ages. Given heterogeneity of TTS implementation across sites, we were also unable to comment upon any association between testing location and participation rates. Finally, the school districts in our study were a convenience sample and not a representative sample of districts in the United States, and the results may not be generalizable to all populations.
Conclusions
With well-planned, well-resourced implementation, combined with layered mitigation strategies, TTS allows students not fully vaccinated to remain in school after close contact with an individual with SARS-CoV-2. Despite the range of TTS protocols and difficulty measuring the exact number of in-person learning days saved, we estimated that TTS preserved in-person learning. Given the risk of future local outbreaks that may lead some schools to reinstitute TTS, this multisite analysis could be helpful in guiding implementation. Considering community transmission levels and continuing layered prevention are important. Properly wearing a well-fitting mask and vaccination are core components of COVID-19 mitigation. Further research on implementation barriers will help support schools and guide resource allocation for successful and accountable TTS programs. Experiences in these 4 districts highlight TTS implementation variability and underscore the broad nature of challenges faced by schools implementing TTS across the country. TTS implementation is resource-intensive and may put increased demands on staff, contact tracing, and testing. Education and public health agencies should work closely with school districts to ensure resources are available to schools wanting to implement TTS.
Acknowledgments
We thank members of the state and local author group, which include Thomas Massaro, MD, PhD; Lisa Patch, MSN, RN, NCSN; LaVerne Hagler; Angela Clark, RN; Jenna Gettings, DVM, MPH; Rachel Franklin, MPH; Amanda Atkins; LaCreisha Murry; Deborah Boian, MBA; Demetrus Liggins, PhD; Kraig Humbaugh, MD, MPH; and Kathleen Winter, PhD, MPH. State and local partners were involved in project study design and logistical planning. They also participated in review and revision of the manuscript. This paper would not have been possible without their support. We also thank the principals, school district officials, and other state and local health department partners contributing to this investigation, including Jill Keys and Katrina Howard, as well as members of the CDC field deployment and data analysis teams, including Nana Baryeh, David Philpott, Angel Rivera, Colleen Scott, and Katlynn Sifre. Finally, we thank the CDC Health Department Liaison Officers Fija Scipio, Alicia Dunajcik, Phillippa Chadd, Anita Pullani, Jamilla Green, Pam Graydon, and Christina Winfield, as well as Tanya Williams with the Field Deployment Unit.
A complete list of study group members appears in the Acknowledgments.
Dr Lammie worked with coronavirus disease 2019 response leadership to conceptualize the combining of these multiple projects into a single paper, drafted the initial manuscript, and served as a field-deployed medical epidemiologist; Dr Ford worked with coronavirus disease 2019 response leadership to conceptualize the combining of these multiple projects into a single paper, drafted the initial manuscript, served as field team lead, and worked with Mr van Zyl to verify the combined analysis; Ms Swanson was a field-deployed epidemiologist and Ms Guinn and Dr Kamitani were field deployment team leads who all worked to clean and prepare their respective site data for analysis and participated in review and revision of the manuscript; Dr Rose advised on data analysis and participated in review and revision of the manuscript; Mr van Zyl advised on data analysis, participated in review and revision of the manuscript, and worked with Dr Ford and other field team leads to complete and verify the analysis; Ms Marynak and Ms Shields supported all work related to the Georgia and New Mexico projects, respectively, and participated in review and revision of the manuscript; Drs Donovan and Mark-Carew and Ms Holman were project team leads who were involved in study design and field support and also participated in review and revision of the manuscript; Dr Welton was both a field deployment team lead and a project team lead who was involved in study design and field support and also participated in review and revision of the manuscript; Ms Thomas assisted in conceptualization and revision of the manuscript and was involved in project study design and field support for all projects; Mr Neatherlin assisted in conceptualization and revision of the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. Names of specific vendors, manufacturers, or products are included for public health and informational purposes; inclusion does not imply endorsement of the vendors, manufacturers, or products by the Centers for Disease Control and Prevention or the US Department of Health and Human Services.
FUNDING: Funded by the Centers for Disease Control and Prevention.
CONFLICT OF INTEREST DISCLAIMER: The authors have indicated they have no conflicts of interest relevant to this article to disclose.
Comments