School closures affected more than 55 million students across the United States when
implemented as a strategy to prevent the transmission of SARS-CoV-2, the virus that
causes COVID-19 (
1
). Reopening schools requires balancing the risks for SARS-CoV-2 infection to students
and staff members against the benefits of in-person learning (
2
). During December 3, 2020–January 31, 2021, CDC investigated SARS-CoV-2 transmission
in 20 elementary schools (kindergarten through grade 6) that had reopened in Salt
Lake County, Utah. The 7-day cumulative number of new COVID-19 cases in Salt Lake
County during this time ranged from 290 to 670 cases per 100,000 persons.
†
Susceptible
§
school contacts
¶
(students and staff members exposed to SARS-CoV-2 in school) of 51 index patients**
(40 students and 11 staff members) were offered SARS-CoV-2 reverse transcription–polymerase
chain reaction (RT-PCR) testing. Among 1,041 susceptible school contacts, 735 (70.6%)
were tested, and five of 12 cases identified were classified as school-associated;
the secondary attack rate among tested susceptible school contacts was 0.7%. Mask
use among students was high (86%), and the median distance between students’ seats
in classrooms was 3 ft. Despite high community incidence and an inability to maintain
≥6 ft of distance between students at all times, SARS-CoV-2 transmission was low in
these elementary schools. The results from this investigation add to the increasing
evidence that in-person learning can be achieved with minimal SARS-CoV-2 transmission
risk when multiple measures to prevent transmission are implemented (
3
,
4
).
On August 24, 2020, a school district in Salt Lake County, Utah, reopened schools
for in-person learning.
††
Elementary schools restricted school-related extracurricular activities and large
group gatherings, placed students in cohorts by classroom, and implemented other COVID-19
strategies to limit spread.
§§
During December 3, 2020–January 31, 2021, CDC was invited by the Utah Department of
Health to investigate SARS-CoV-2 transmission in a convenience sample of 20 elementary
schools in partnership with the school district, the University of Utah’s Health and
Economic Recovery Outreach (HERO) Project,
¶¶
Utah Department of Health, and Salt Lake County Health Department.
School contacts of identified index patients completed a questionnaire about symptoms
and exposures and received SARS-CoV-2 testing. Written consent was provided by participants
(or by a parent or guardian for minors). Persons not susceptible to SARS-CoV-2 infection
were excluded. Saliva samples (or nasal swabs if saliva was unobtainable) were collected
for SARS-CoV-2 RT-PCR testing 5–10 days postexposure; turnaround time for results
was typically 1–2 days. Household members of school contacts with a positive SARS-CoV-2
test result were interviewed and offered SARS-CoV-2 RT-PCR testing. The Utah Public
Health Laboratory performed whole genome sequencing (WGS) for available positive specimens.
A school contact who received a positive test result was considered not to have a
school-associated case of COVID-19 when one of the following occurred: 1) illness
onset preceded the first date of school exposure, 2) a household member had illness
onset during the 14 days preceding the school contact’s illness onset (for symptomatic
school contacts) or before the last date of school exposure (for asymptomatic school
contacts), or 3) WGS demonstrated that the lineage of the index patient’s isolate
differed from that of the school contact.*** To understand school mitigation measures
and classroom characteristics, principals and teachers of each index patient were
surveyed. Classroom seat distances between students and between the teacher and nearest
student were measured. SAS (version 9.4; SAS Institute) was used for descriptive statistics.
This activity was reviewed by CDC and was conducted consistent with applicable federal
law and CDC policy.
†††
The 20 elementary schools included 1,214 staff members and 10,171 students, 81% of
whom attended school in person and 56% of whom were eligible for free or reduced-price
meal programs. Among the student population, 53% were non-Hispanic White persons,
31% were Hispanic or Latino persons, 5% were Asian persons, 5% were Native Hawaiian
or Other Pacific Islander persons, and 4% were Black or African American persons.
Fifty-one index patients (40 students, median age = 9.5 years [range = 5–12 years]
and 11 staff members, median age = 50 years [range = 26–62 years]) were identified
from 48 classrooms (Table 1). These index patients were infectious at school for a
median of 2 days (range = 1–4 days), and 16 (31%) were asymptomatic. A total of 1,083
school contacts (943 students and 140 staff members) were identified; 42 (4%) were
not susceptible to SARS-CoV-2 infection.
§§§
Among the 1,041 susceptible school contacts (student median age = 9 years [range =
5–18 years]; staff member median age = 39.5 years [range = 19–83 years]), 144 (14%)
were quarantined (Table 2). Among the 735 (71%) tested school contacts (participation
range = 44%–100% across schools), testing was completed a median of 8 days after the
school exposure (range = 6–15 days). Overall, 103 of 133 (77%) staff member contacts
and 632 of 908 (70%) student contacts were tested; among 303 Hispanic or Latino contacts
and 566 non-Hispanic White contacts, 237 (78%) and 382 (67%) respectively, were tested.
TABLE 1
Characteristics of index and school-associated patients with laboratory-confirmed
COVID-19 in 20 elementary schools — Salt Lake County, Utah, December 3, 2020–January
31, 2021
Characteristic
No. (%) of persons with COVID-19
Index (n = 51)*
School-associated (n = 5)†
Cases per school, median (range)
2 (1–9)
0 (0–2)
School contacts, median (range)
20 (5–53)
—
§
Close contacts, median (range)
6 (0–23)
—
Other school contacts, median (range)
13 (0–52)
—
Median age, yrs (range)
Students (index: n = 40; school-associated: n = 4)
9.5 (5–12)
10.5 (10–12)
Staff members (index: n = 11; school-associated: n = 1)
50 (26–62)
43 (43–43)
Sex
Male
24 (47.1)
2 (40.0)
Female
27 (52.9)
3 (60.0)
Race/Ethnicity
White, non-Hispanic
30 (58.8)
1 (20.0)
Hispanic/Latino
15 (29.4)
2 (40.0)
Black/African American
1 (2.0)
0 (0.0)
Asian
1 (2.0)
1 (20.0)
Native Hawaiian/Other Pacific Islander
2 (3.9)
0 (0.0)
American Indian or Alaska Native
0 (0.0)
0 (0.0)
Multiracial
2 (3.9)
1 (20.0)
Grade in school¶
Kindergarten
5 (12.5)
0 (0.0)
1
3 (7.5)
0 (0.0)
2
2 (5.0)
0 (0.0)
3
6 (15.0)
0 (0.0)
4
6 (15.0)
2 (50.0)
5
8 (20.0)
0 (0.0)
6
10 (25.0)
2 (50.0)
Role in school
Students
40 (78.4)
4 (80.0)
Head teachers
6 (11.8)
1 (20.0)
Paraeducators**
0 (0.0)
0 (0.0)
Other teachers††
4 (7.8)
0 (0.0)
Other staff members§
§
1 (2.0)
0 (0.0)
Days in school while infectious, median (range)
2 (1–4)
0 (0–2)
Symptom status
Ever symptomatic
35 (68.6)
2 (40.0)
Asymptomatic
16 (31.4)
3 (60.0)
One or more underlying medical condition¶¶
9 (20.9)
0 (0.0)
Quarantine status after exposure to index patient***
Under quarantine
—
3 (60.0)
Notified, close contact
—
0 (0.0)
Notified, not close contact
—
2 (40.0)
Abbreviation: IQR= interquartile range.
* An index patient was defined as a student or staff member with laboratory-confirmed
SARS-CoV-2 infection who had attended in-person school while infectious for at least
1 day. Infectious period was estimated as 2 days before to 10 days after symptom onset
(if symptomatic) or first positive specimen collection date (if asymptomatic).
† School-associated transmission was excluded if 1) the school contact had an illness
onset (if symptomatic, symptom onset, if asymptomatic, first positive test date) before
the last date of school exposure, 2) a household member had an illness onset (if symptomatic,
symptom onset, if asymptomatic, first positive test date) within 14 days of the positive
school contact’s illness onset (if school contact was symptomatic) or before the last
date of school exposure (if the school contact was asymptomatic) or 3) whole genome
sequencing supported nonschool-associated transmission.
§ Dashes indicate that data are not applicable.
¶ Restricted to students. For index patients, n = 40, for secondary cases, n = 4.
** Includes teacher aides and interns.
†† Includes ethics teachers, instructional coaches, learning support teachers, special
education teachers, and substitute teachers.
§§ Includes administrators, bus drivers, and health specialists.
¶¶ Missing data: Underlying medical conditions: eight index patients, one school-associated
patient.
*** Starting January 4, 2021, the school district changed its quarantine policy based
on changes to state recommendations, and only students and staff members identified
as close contacts (i.e., within 6 ft of the index patient for a cumulative total of
≥15 minutes over a 24-hour period) of the index patient were quarantined when both
were maskless; previously, all close contacts would have been quarantined regardless
of mask use. Any close contacts identified in January who met the criteria to not
quarantine were categorized as “Notified, close contact.” Those who shared a classroom
space with the index patient but were not identified as close contacts were categorized
as “Notified, not close contact.”
TABLE 2
Characteristics of COVID-19–susceptible school contacts* in 20 elementary schools
— Salt Lake County, Utah, December 3, 2020–January 31, 2021
Characteristic
No. (%) of school contacts
Total (N = 1,041)
Tested (n = 735)
Overall participation
—
†
735 (70.6)
Median percent participation across 20 schools (range)
—
69.7 (44.4–100.0)
Median age, yrs (range)§
Students (n = 908)
9.0 (5.0–18.0)
9.0 (5.0–18.0)
Staff members (n = 112)
39.5 (19.0–83.0)
39.0 (19.0–83.0)
Sex
Male
487 (47.7)
352 (47.9)
Female
535 (52.3)
383 (52.1)
Race/Ethnicity
White, non-Hispanic
566 (55.9)
382 (52.0)
Hispanic/Latino
303 (29.9)
237 (32.2)
Black/African American
28 (2.8)
25 (3.4)
Asian
33 (3.3)
29 (3.9)
Native Hawaiian/Other Pacific Islander
28 (2.8)
15 (2.0)
American Indian or Alaska Native
8 (0.8)
7 (1.0)
Multiracial
47 (4.6)
40 (5.4)
Grade
¶
Kindergarten
110 (12.1)
61 (9.7)
1
107 (11.8)
79 (12.5)
2
139 (15.3)
108 (17.1)
3
113 (12.4)
78 (12.3)
4
134 (14.8)
95 (15.0)
5
118 (13.0)
86 (13.6)
6
182 (20.0)
121 (19.1)
≥7
5 (0.6)
4 (0.6)
Role in school
Students
908 (87.2)
632 (86.0)
Head teachers
77 (7.4)
61 (8.3)
Paraeducators**
24 (2.3)
13 (1.8)
Other teachers
††
14 (1.3)
12 (1.6)
Other staff members§§
18 (1.7)
17 (2.3)
Days between school exposure and test date, median (range)
¶¶
8 (6–15)
8 (6–15)
Quarantine status after exposure to index patient***
Quarantined
144 (13.8)
105 (14.3)
Notified, close contact
183 (17.6)
131 (17.8)
Notified, not close contact
714 (68.6)
499 (67.9)
* School contact was defined as a student or staff member who was in contact with
the index patient for a total of ≥15 minutes in a classroom, cafeteria, school bus,
or recess space during an index patient’s infectious period. This includes any contacts
who received positive SARS-CoV-2 test results but were not determined to have school-associated
cases.
† Dashes indicate that data are not applicable.
§ Missing data (also applies to Sex and Race/Ethnicity categories): Age: 21 nonparticipating
staff members; Sex: 19 nonparticipating staff members; Race/Ethnicity: 28 nonparticipants.
¶ Restricted to students (n = 908). Students in grade 7 or higher were contacts of
an elementary school student on the school bus. All five students in grade 7 or higher
were contacts of the same index patient. Bus contacts were not routinely included
on the list of school contacts for all 51 index patients.
** Includes teacher aides and interns.
†† Includes ethics teachers, instructional coaches, learning support teachers, special
education teachers, and substitute teachers.
§§ Includes administrators, bus drivers, and health specialists.
¶¶ All classroom testing occurred 6–10 days after exposure. One contact was tested
on day 8 and offered a follow-up repeat testing on day 15.
*** Starting January 4, 2021, the school district changed its quarantine policy based
on changes to state recommendations, and only students and staff members identified
as close contacts (i.e., within 6 ft of the index patient for a cumulative total of
≥15 minutes over a 24-hour period) of the index patient were quarantined when both
were maskless; previously, all close contacts would have been quarantined regardless
of mask use. Any close contacts identified in January who met the criteria to not
quarantine were categorized as “Notified, close contact.” Those who shared a classroom
space with the index patient but were not identified as close contacts were categorized
as “Notified, not close contact.”
Among all 735 tested contacts, 12 (1.6%) (11 students, one teacher) had a positive
SARS-CoV-2 test result, seven of whom were determined not to have school-associated
cases because of epidemiologic evidence (four) or because WGS suggested community
acquisition based on lineage differences (three) (Supplementary Figure, https://stacks.cdc.gov/view/cdc/104112).
WGS was only available for three pairs of index patients and their associated contacts
(Table 3). After exclusion, five cases from five separate classrooms were classified
as school-associated, for a secondary attack rate of 0.7% (five of 728). No outbreaks
were detected.
¶¶¶
Three of five persons with school-associated cases had been quarantined (the secondary
attack rate among quarantined persons who were tested was 3.0% [three of 101]); the
remaining two persons with school-associated cases had not been quarantined and were
isolated only after a positive test result (secondary attack rate among nonquarantined
contacts who were tested = 0.3% [two of 627]).**** Among the five persons with school-associated
cases, three persons were asymptomatic, and three persons were exposed to asymptomatic
index patients; four cases were attributed to student-to-student transmission, and
one was attributed to student-to-teacher transmission. Four of the five school-associated
transmission events occurred because the contact sat <6 ft from the index patient
during class (two) or during lunch (two), or the index patient or contact had poor
mask use (two) or physical distancing behavior (two) (Table 3). All five households
of persons with school-associated cases were tested. Tertiary transmission was detected
in three households; within those households, six of eight household members received
positive SARS-CoV-2 test results.
TABLE 3
Characteristics of 12 contacts who received positive SARS-CoV-2 test results and summary
of evidence for school-associated transmission in five contacts across 20 elementary
schools — Salt Lake County, Utah, December 3, 2020–January 31, 2021*
Positive contact ID
Index patient
School contact†
School-associated transmission
Factors associated with transmission
School role
Symptoms reported
School role
Symptoms reported
Basis for exclusion of school-associated transmission
School-associated transmission hypothesized
Close contact between patient and contact†
Contact sat <6 ft from index patient
Poor adherence to distancing, mask use, or neither at school
Epidemiologic data
WGS data
Index patient
Contact
I1
Student
N
Student
N
N
NA
Y
Y
Class
Distancing
Mask use, distancing
J2
Student
N
Student
Y
N
NA
Y
Y
Class
Neither
Mask use
X3
Student
Y
Student
N
N
NA
Y
N
Lunch
Neither
Distancing
AA4
Student
Y
Student
N
N
NA
Y
Y
Lunch
Neither
Neither
EE5
Student
N
Teacher
Y
N
NA
Y
N
Neither
Neither
Neither
A6
Student
Y
Student
Y
N
Y
N
Y
—§
—
—
A7
Student
Y
Student
N
N
Y
N
Y
—
—
—
L8
Student
N
Student
Y
N
Y
N
Y
—
—
—
O9
Teacher
N
Student
Y
Y
NA
N
Y
—
—
—
T10
Student
Y
Student
Y
Y
NA
N
Y
—
—
—
RR11
Teacher
Y
Student
Y
Y
NA
N
Y
—
—
—
VV12
Student
Y
Student
Y
Y
NA
N
Y
—
—
—
Abbreviations: ID = identifier; Y = yes; N = no; NA = not available; WGS = whole genome
sequencing.
* School-associated transmission was excluded by epidemiologic data if 1) the school
contact had an illness onset (if symptomatic, symptom onset; if asymptomatic, first
positive test date) before the last date of school exposure, or 2) a household member
had an illness onset (if symptomatic, symptom onset; if asymptomatic, first positive
test date) within 14 days of the positive school contact’s illness onset (if school
contact was symptomatic) or before the last date of school exposure (if the school
contact was asymptomatic). School-associated transmission was excluded by WGS data
if the index patient isolate was found to be a different lineage from the positive
school contact isolate.
† Persons were determined to be close contacts if they were <6 ft from the index patient
for a cumulative total of ≥15 minutes during a 24-hour period at school. All other
school contacts were students or staff members who were in contact with the index
patient for a cumulative total of ≥15 minutes in a classroom, cafeteria, school bus,
or recess space during an index patient’s infectious period.
§ Dashes indicate that data are not applicable.
On December 17, 2020, Utah modified its quarantine recommendations for school contacts
(students or staff members) who were identified as close contacts (persons within
6 ft of the index patient for a cumulative total of ≥15 minutes during a 24-hour period).
Previously, school contacts who were close contacts were quarantined
††††
regardless of mask use; afterwards, they were only quarantined when the index patient
or the contact did not wear a mask during the interaction. The school district implemented
this recommendation on January 4, 2021, after a holiday break, and 158 students who
were close contacts continued attending in-person school. Among these 158 students,
111 (70%) were tested; no school-associated cases were detected.
Students in 42 classrooms
§§§§
(median class size = 22 students [range = 3–33 students]) sat a median of 3 ft (range
= 1–5 ft) apart within the classroom, with a median of eight students (range = 1–16
students) sitting within a radius of 6 ft (Supplementary Table 1, https://stacks.cdc.gov/view/cdc/104112).
Among 37 teachers with available data, 23 (62%) were seated ≥6 ft from the closest
student (median = 6 ft, range = 2–10 ft), but all teachers reported daily one-on-one
or small group instruction in close proximity to students, almost always without using
plexiglass or physical barriers. Among 42 teachers, 36 (86%) reported that students
always wore masks indoors except when eating or drinking. Nineteen of 20 (95%) principals
reported using staggered mealtimes to increase spacing between students during lunch
in the cafeteria (although still <6 ft apart). All schools reported implementing multiple
measures to decrease in-school SARS-CoV-2 transmission (Supplementary Table 2, https://stacks.cdc.gov/view/cdc/104112).
Discussion
Despite high community incidence and an inability to space students’ classroom seats
≥6 ft apart, this investigation found low SARS-CoV-2 transmission and no school-related
outbreaks in 20 Salt Lake County elementary schools with high student mask use and
implementation of multiple strategies to limit transmission. Other U.S. studies have
also detected minimal school-associated transmission when implementing strict mitigation
measures, although testing was limited to symptomatic close contacts (
3
,
4
). Because children with COVID-19 are frequently asymptomatic (
5
), the expanded testing to all school contacts regardless of symptom status in this
investigation strengthens the evidence for low elementary school transmission.
In addition to implementation of multiple strategies to reduce in-school transmission,
school-related activities that increase the risk for SARS-CoV-2 transmission, such
as school-based team sports (
6
), were suspended. Although most teachers were seated ≥6 ft from students, CDC’s recommendation
at the time of the study of ≥6 ft student distancing within the classroom (
7
) was not possible because of limited space. A recent study in Massachusetts found
no difference in student and staff member case rates from school districts with ≥3
feet physical distancing requirements compared with school districts with ≥6 feet
physical distancing requirements (
8
). The study detected no teacher-driven transmission; other school investigations
have identified teachers and staff members as being central to in-school transmission
¶¶¶¶
(
9
,
10
). Although school-associated transmission was rare in this investigation, most cases
did lead to household transmission, highlighting the importance of reducing school
transmission to prevent infected children from transmitting SARS-CoV-2 to household
members.
The modified quarantine policy, allowing contacts to continue attending in-person
school if both the index patient and the contact were wearing a mask, did not lead
to additional school-associated transmission and resulted in over 1,200 student in-person
learning days saved.***** Among the five school-associated cases, the contact or index
patient often had poor mask compliance, or they sat near one another during lunch.
Findings suggest that quarantine determinations based on mask use of the index patient
and close contacts might be adequate for preventing additional school-associated transmission
in schools implementing multiple critical prevention strategies.
The findings in this report are subject to at least four limitations. First, WGS to
differentiate school-associated from community transmission in a high incidence setting
was not always available. Second, some infected contacts might have been missed because
not all contacts received testing and the winter break mid-investigation might have
interrupted additional school-associated transmission. Third, misclassification of
susceptibility might have occurred as immunity status was unknown. Finally, these
findings are specific to the current circulating SARS-CoV-2 variant distribution;
as variant distribution shifts to new variants, more transmission might occur.
In an urban county with high SARS-CoV-2 community incidence, comprehensive testing
of contacts detected low school-associated transmission in elementary schools, with
a secondary attack rate of 0.7%. These results suggest that when ≥6 ft distancing
is not feasible, schools in high-incidence communities can still limit in-school transmission
by consistently using masks and implementing other important mitigation strategies.
Summary
What is already known about this topic?
Data suggest that school-associated SARS-CoV-2 transmission is low.
What is added by this report?
SARS-CoV-2 testing was offered to 1,041 school contacts of 51 index patients across
20 elementary schools in Salt Lake County, Utah. In a high community transmission
setting, low school-associated transmission was observed with a 0.7% secondary attack
rate. Mask adherence was high, but students’ classroom seats were <6 ft apart and
a median of 3 ft apart.
What are the implications for public health practice?
These findings add to evidence that in-person elementary schools can be opened safely
with minimal in-school transmission when critical prevention strategies including
mask use are implemented, even though maintaining ≥6 ft between students’ seats might
not be possible.