Research

JAMA Internal Medicine | Original Investigation

Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites

in the United States, March 23-May 12, 2020
Fiona P. Havers, MD, MHS; Carrie Reed, PhD; Travis Lim, DrPH; Joel M. Montgomery, PhD; John D. Klena, PhD;
Aron J. Hall, DVM; Alicia M. Fry, MD; Deborah L. Cannon, BS; Cheng-Feng Chiang, PhD; Aridth Gibbons, BS;
Inna Krapiunaya, MS; Maria Morales-Betoulle, PhD; Katherine Roguski, MPH; Mohammad Ata Ur Rasheed, PhD;
Brandi Freeman, PhD; Sandra Lester, PhD; Lisa Mills, PhD; Darin S. Carroll, PhD; S. Michele Owen, PhD;
Jeffrey A. Johnson, PhD; Vera Semenova, PhD; Carina Blackmore, DVM; Debra Blog, MD; Shua J. Chai, MD;
Angela Dunn, MD; Julie Hand, MSPH; Seema Jain, MD; Scott Lindquist, MD; Ruth Lynfield, MD;
Scott Pritchard, MPH; Theresa Sokol, MPH; Lynn Sosa, MD; George Turabelidze, MD; Sharon M. Watkins, PhD;
John Wiesman, DrPH; Randall W. Williams, MD; Stephanie Yendell, DVM;
Jarad Schiffer, MS; Natalie J. Thornburg, PhD

Supplemental content
IMPORTANCE Reported cases of severe acute respiratory syndrome coronavirus 2 Related article at
(SARS-CoV-2) infection likely underestimate the prevalence of infection in affected jama.com
communities. Large-scale seroprevalence studies provide better estimates of the proportion
of the population previously infected.

OBJECTIVE To estimate prevalence of SARS-CoV-2 antibodies in convenience samples from

several geographic sites in the US.

DESIGN, SETTING, AND PARTICIPANTS This cross-sectional study performed serologic testing
on a convenience sample of residual sera obtained from persons of all ages. The serum was
collected from March 23 through May 12, 2020, for routine clinical testing by 2 commercial
laboratory companies. Sites of collection were San Francisco Bay area, California;
Connecticut; south Florida; Louisiana; Minneapolis-St Paul-St Cloud metro area, Minnesota;
Missouri; New York City metro area, New York; Philadelphia metro area, Pennsylvania; Utah;
and western Washington State.

EXPOSURES Infection with SARS-CoV-2.

MAIN OUTCOMES AND MEASURES The presence of antibodies to SARS-CoV-2 spike protein
was estimated using an enzyme-linked immunosorbent assay, and estimates were
standardized to the site populations by age and sex. Estimates were adjusted for test
performance characteristics (96.0% sensitivity and 99.3% specificity). The number of
infections in each site was estimated by extrapolating seroprevalence to site populations;
estimated infections were compared with the number of reported coronavirus disease 2019
(COVID-19) cases as of last specimen collection date.

RESULTS Serum samples were tested from 16 025 persons, 8853 (55.2%) of whom were
women; 1205 (7.5%) were 18 years or younger and 5845 (36.2%) were 65 years or older.
Most specimens from each site had no evidence of antibodies to SARS-CoV-2. Adjusted
estimates of the proportion of persons seroreactive to the SARS-CoV-2 spike protein
antibodies ranged from 1.0% in the San Francisco Bay area (collected April 23-27) to 6.9%
of persons in New York City (collected March 23-April 1). The estimated number of infections
ranged from 6 to 24 times the number of reported cases; for 7 sites (Connecticut, Florida,
Louisiana, Missouri, New York City metro area, Utah, and western Washington State),
an estimated greater than 10 times more SARS-CoV-2 infections occurred than the number
of reported cases.

CONCLUSIONS AND RELEVANCE During March to early May 2020, most persons in 10 diverse
geographic sites in the US had not been infected with SARS-CoV-2 virus. The estimated
Author Affiliations: Author
number of infections, however, was much greater than the number of reported cases in all affiliations are listed at the end of this
sites. The findings may reflect the number of persons who had mild or no illness or who did article.
not seek medical care or undergo testing but who still may have contributed to ongoing virus Corresponding Author: Fiona P.
transmission in the population. Havers, MD, MHS, CDC COVID-19
Response Team, Centers for Disease
Control and Prevention, 1600 Clifton
JAMA Intern Med. doi:10.1001/jamainternmed.2020.4130 Rd, MS H24-6, Atlanta, GA 30329
Published online July 21, 2020. ([email protected]).

(Reprinted) E1
© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 08/02/2020

 Research Original Investigation Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, 2020

T
he first case of severe acute respiratory syndrome coro-
navirus 2 (SARS-CoV-2) infection in the US was re- Key Points
ported in Washington State on January 20, 2020. The
Question What proportion of persons in 10 US sites had
first US case linked to community transmission was reported detectable antibodies to severe acute respiratory syndrome
in California on February 26, 2020, followed by subsequent coronavirus 2 (SARS-CoV-2) from March 23 to May 12, 2020?
cases resulting from community transmission reported in
Findings In this cross-sectional study of 16 025 residual clinical
Washington on February 28 and New York on March 3.1-5 Since
specimens, estimates of the proportion of persons with detectable
January 2020, states have been recommended to report all SARS-CoV-2 antibodies ranged from 1.0% in the San Francisco Bay
laboratory-confirmed cases to the Centers for Disease Con- area (collected April 23-27) to 6.9% of persons in New York City
trol and Prevention (CDC).6 Reported cases, however, likely rep- (collected March 23-April 1). Six to 24 times more infections were
resent only a fraction of SARS-CoV-2 infections, as an un- estimated per site with seroprevalence than with coronavirus
known proportion of cases are mild or asymptomatic, or they disease 2019 (COVID-19) case report data.
are otherwise not diagnosed or ascertained through passive Meaning For most sites, it is likely that greater than 10 times more
public health reporting.7-9 Furthermore, viral testing has been SARS-CoV-2 infections occurred than the number of reported
limited in many sites and was often reserved for severely ill COVID-19 cases; most persons in each site, however, likely had no
patients early in the US outbreak, and testing availability has detectable SARS-CoV-2 antibodies.
changed rapidly. Each of these issues could confound esti-
mates of incident cases and epidemic dynamics that use only
case-based reporting data.
Detection of antibodies to SARS-CoV-2 in a person’s blood at least 300 specimens per age group. Specimens from all
likely indicates that they were infected at some point since the sites were deduplicated through laboratory records, except
start of the pandemic. Thus, serologic assays can be used to for specimens provided by Lab B from WA and NY. Both
provide population-based estimates of infection that include laboratories provided specimens from NY and WA; Lab A
people who had mild or asymptomatic infection or who were also provided specimens from CA, FL, and LA, and Lab B
never tested despite having symptoms. provided specimens from CT, MO, PA, MN, and UT. The zip
We used convenience samples of residual clinical speci- code of patient residence was known for all Lab A specimens
mens obtained from 2 commercial diagnostic laboratories to and for Lab B specimens from CT, MO, PA, MN, and UT, but
conduct a serologic survey. Our goal was to estimate the se- not for Lab B specimens from NY and WA. Based on informa-
roprevalence in the population—that is, the proportion of the tion from Lab B, which indicated that the majority of speci-
population with evidence of previous infection with SARS- mens from its facilities in WA and NY were drawn from the
CoV-2, by age group, in 10 geographically diverse US sites with areas of greatest population density in western WA and NY
known community transmission. metro areas, respectively, we assumed that Lab B specimens
from WA and NY were from a similar geographic distribution
to those received from Lab A for those sites. For individual
specimens, no information on the reason for specimen col-
Methods lection was available.
We obtained convenience samples of deidentified residual pa- After reviewing the protocol, CDC human subjects re-
tient sera collected for routine screening (eg, cholesterol screen- search officials determined that the testing represented non-
ing) or clinical management by 2 commercial clinical labora- research activity in the setting of a public health response to
tories (Lab A and Lab B) from 10 sites. For Lab A, data on the the coronavirus disease 2019 (COVID-19) pandemic and ex-
breakdown between inpatients and outpatients were not avail- empted it from further review. Informed consent was waived
able. For Lab B, almost all the samples were from outpa- as deidentified data were used. On June 26, 2020, data from
tients. The samples were collected during discrete periods from 6 sites (CT, FL, MO, NY, UT, and WA) were released on CDC’s
March 23 through May 12 (Table 1; Figure 1). Sites and dates of website,18 and an early version of the non–peer-reviewed
collection included western Washington State (WA) (defined manuscript was posted on a preprint server.19
broadly; March 23-April 1); the New York City metro area (NY)
(defined broadly; predominantly Manhattan, Bronx, Brook- Laboratory Methods
lyn, Queens, and Nassau counties; March 23-April 1); south Sera were tested at CDC in a 2-step process—a screening assay
Florida (FL) (restricted to Miami-Dade, Broward, Palm Beach, followed by a confirmatory assay for presumptive reactive
and Martin counties; April 6-10); Philadelphia metro statisti- specimens identified through screening. The CDC developed
cal area counties and Lancaster and Cumberland counties (PA) and validated an enzyme-linked immunosorbent assay (ELISA)
(April 13-25); San Francisco Bay area, including San Jose (CA) that was used as the confirmatory assay, as has been previ-
(April 23-27); Minneapolis-St Paul-St Cloud combined statis- ously described.20 A specimen was considered reactive if, on
tical areas (MN) (April 30-May 12); and all of Missouri (MO) confirmatory testing, at a background corrected optical den-
(April 20-26), Utah (UT) (April 20-May 3), Connecticut (CT) sity of 0.4 and at a serum dilution of 1:100, it had a signal to
(April 26-May 3), and Louisiana (LA) (April 1-8) (eFigure 1 in threshold ratio greater than 1. The screening assay was simi-
the Supplement). Age or age group, patient sex, and collec- lar to the confirmatory assay. Sera were screened at a 1:100 di-
tion date were available for all specimens; we aimed to have lution using a qualitative pan immunoglobulin (Ig) ELISA

E2 JAMA Internal Medicine Published online July 21, 2020 (Reprinted) jamainternalmedicine.com

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 08/02/2020

 Table 1. Number of Residual Clinical Specimens From Commercial Laboratories Tested for Antibodies to Severe Acute Respiratory Syndrome Coronavirus 2 in 10 Geographic Sites,
by Sex, Age Group, and Location, With Dates of Specimen Collection for Each Site

No. (%)
Minneapolis-
New York City Philadelphia San Francisco St Paul-St Cloud
Western metro area metro area Bay area metro area
Characteristic All sites Washington State (New York) Louisiana South Florida (Pennsylvania) Missouri Utah (California) Connecticut (Minnesota)

jamainternalmedicine.com
Dates of specimen March 23-May March 23-April 1, March 23-April 1, April 1-8, 2020 April 6-10, 2020 April 13-25, 2020 April 20-26, April April 23-27, April April 30-May 12,
collection 12, 2020 2020 2020 2020 20-May 3, 2020 26-May 3, 2020
2020 2020
Date of first case NA February 28, 2020 February 27, 2020 March 9, 2020 March 1, 2020 March 9, 2020 March 13, March 14, February 26, March 11, March 9, 2020
of community 2020 2020 2020 2020
transmissiona
Sex

Downloaded From: https://jamanetwork.com/ on 08/02/2020

Female 8853 (55.2) 1930 (59.1) 1333 (53.7) 677 (57.2) 964 (55.3) 422 (51.2) 1018 (54.1) 673 (59.5) 653 (53.4) 729 (50.9) 454 (52.8)
Male 7178 (44.8) 1334 (40.9) 1149 (46.3) 507 (42.8) 778 (44.7) 402 (48.8) 864 (45.9) 465 (41.1) 571 (46.7) 702 (49.1) 406 (47.2)
Age group, y
0-18 1203 (7.5) 219 (6.7) 311 (12.5) 33 (2.8) 69 (4.0) 75 (9.1) 158 (8.4) 25 (2.2)b 45 (3.7) 219 (15.3) 49 (5.7)
19-49 5327 (33.2) 1213 (37.2) 909 (36.6) 619 (52.3) 491 (28.2) 193 (23.4) 394 (20.9) 470 (41.5) 371 (30.3) 297 (20.8) 370 (43.0)
50-64 3701 (23.1) 782 (24.0) 455 (18.3) 322 (27.2) 326 (18.7) 221 (26.8) 405 (21.5) 328 (29.0) 323 (26.4) 300 (21.0) 239 (27.8)
≥65 5802 (36.2) 1050 (32.2) 807 (32.5) 212 (17.9) 856 (49.1) 335 (40.7) 925 (49.2) 315 (27.8) 485 (39.6) 615 (43.0) 202 (23.5)
Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, 2020

Totalc 16 025 3264 2482 1184 1742 824 1882 1132 1224 1431 860
b
Abbreviation: NA, not applicable. Excluded from analysis owing to small sample size.
a c
Sources for dates of community transmission were state and city health departments and Centers for Disease The total number of specimens that had complete data for both age and sex.
Control and Prevention data.5,10-17

© 2020 American Medical Association. All rights reserved.

(Reprinted) JAMA Internal Medicine Published online July 21, 2020
Original Investigation Research

E3
 Research Original Investigation Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, 2020

Figure 1. Estimates of Seroprevalence to Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2)

Antibodies and Timeline of Specimen Collection

A Estimates of seroprevalence

8
Population reactive for SARS-CoV-2 antibodies, %

New York City metro

7
area, New York

6
Louisiana
Connecticut
5

4
Philadelphia Missouri
metro area,
3 Minneapolis-St
Pennsylvania
Paul-St Cloud
Utah metro area,
2
South Florida Minnesota
Western Washington
1
San Francisco Bay area,
California
0
March March March April April April April May May May
15 22 29 5 12 19 26 3 10 17
Dates of sample collection

B Timeline of specimen collection

New York City metro March 23-

area, New York April 1

March 23-
Western Washington
April 1

April
Louisiana
1-8

April
South Florida
5-10

Philadelphia metro area, April

Pennsylvania 13-25

April
Missouri
20-26

San Francisco Bay area, April

California 23-27

April 20- A, Estimates are shown with 95% CIs

Utah for 10 geographic sites from which
May 3
residual clinical specimens were
April 26-
Connecticut
May 3
collected. Seroprevalence estimate is
shown at the midpoint of the
Minneapolis-St Paul-St Cloud April 30- specimen collection date range.
metro area, Minnesota May 12
B, Timeline with specimen collection
dates for each site.

against the prefusion stabilized ectodomain of the SARS- Analysis

CoV-2 spike protein.21 However, a greater coating concentra- We calculated seroprevalence as the proportion of specimens
tion of spike protein was used, only 1 dilution was tested for that were confirmed reactive, stratified by sex and age group
each serum sample (1:100), and different optical density cut- (0 to ≤18 years, 19 to ≤49 years, 50 to ≤64 years, and ≥65 years).
offs were first used to identify presumptive reactive speci- We calculated age-standardized and sex-standardized sero-
mens, which were then referred for confirmatory testing.20 prevalence estimates using weights derived from US census
Using the above definition of reactivity, specificity was 99.3% county-level population projections for the most sampled
(95% CI, 98.3%-99.9%) and sensitivity was 96.0% (95% CI, counties for CA, FL, MN, NY, PA, and WA, and from US census
90.0%-98.9%).20 Results of testing against sera from poly- state-level data for CT, LA, MO, and UT. We estimated 95% CIs
merase chain reaction–confirmed infections with other coro- by generating 10 000 bootstrapped samples with replace-
naviruses indicate that antibodies to commonly circulating hu- ment (eMethods in the Supplement). We conducted addi-
man coronaviruses exhibited some cross-reactivity, but the tional analyses with bootstrapping to account for assay test
level of cross-reactivity was below the limits of detection for performance, using sensitivity and specificity parameters
this assay.20 described above. We defined estimates that were age-

E4 JAMA Internal Medicine Published online July 21, 2020 (Reprinted) jamainternalmedicine.com

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 08/02/2020

 Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, 2020 Original Investigation Research

standardized and sex-standardized and adjusted for test char- Table 3 shows estimates of the number of SARS-CoV-2 in-
acteristics as fully adjusted estimates. To assess for potential fections suggested by seroprevalence estimates in each site and
differences in populations using different laboratories, we com- compares these with the number of reported cases as of the
pared seroprevalence in specimens from Lab A with those from last date of specimen collection (eFigure 2 in the Supple-
Lab B for NY and WA, the 2 sites where both laboratories ment). Our estimate for underascertainment was lowest in CT,
collected specimens. where the estimation of 176 012 infections was 6.0 (range, 4.3-
To estimate the degree of underascertainment of re- 7.8) times greater than the 29 287 reported cases as of May 3,
ported cases for all sites, we assumed that the presence of SARS- 2020, and highest for MO, where the estimation of 161 936 in-
CoV-2 antibodies represented infections that occurred prior to fections was 23.8 (range, 14.8-34.7) times greater than the 6794
the last date of specimen collection. We applied the esti- reported cases as of April 25, 2020. Estimated numbers of
mated age-adjusted and sex-adjusted seroprevalence esti- infections for 7 sites—CT, FL, LA, MO, NY, UT, and WA—were
mates to the respective populations to estimate total infec- at least 10 times greater than the number of reported cases.
tions. We then divided these numbers by the cumulative case Estimates of underascertainment using the date 7 days
counts reported to health departments22 as of the last date of prior to start of specimen collection are shown in eTable 2 in
specimen collection for each site. Because antibodies may take the Supplement. Using these earlier dates, our point estimate
an average of 10 to 14 days to be detectable after infection,23-25 for underascertainment was lowest in CT, where the number
and collection periods were 6 to 14 days in length, we ac- of estimated infections was 8.9 times greater than the num-
counted for a lag in the development of antibodies in a sce- ber of cases reported as of April 19, 2020, and highest in NY,
nario analysis using the cumulative number of reported cases where the estimation of 641 778 infections was more than 1000
as of 7 days prior to the start of specimen collection. The speci- times greater than the 545 cases reported as of March 16, 2020.
men collection period in relationship to reported cases is shown These estimates do not account for delays in reporting re-
for each site in eFigure 2 in the Supplement, and the number sults, which may have been longer earlier in the pandemic.
of cases reported daily in the US is shown in eFigure 3 in the
Supplement.
R (version 3.6.1) and Rstudio (version 1.2.1335) (R Foun-
dation for Statistical Computing) were used to perform statis-
Discussion
tical analyses. Two-sided P values less than .05 were consid- Our study estimated seroprevalence of antibodies to SARS-
ered statistically significant. CoV-2 in 10 diverse geographic sites in the US, with discrete
collection periods from late March through mid-May 2020.
Seroprevalence estimates varied from 1.0% in the San Fran-
cisco Bay area in late April to 6.9% in the New York City metro
Results area in late March. Our results for each site suggest that the
We tested 16 025 residual sera specimens from 10 sites col- number of infections was much greater than the number of re-
lected from March 23 through May 12, with discrete collec- ported cases throughout the study period; these infections
tion periods for each site (Table 1). A total of 6320 (39.4%) speci- likely include asymptomatic and mild infections for which
mens were from Lab A, and 9705 (60.6%) specimens were from health care was not sought, as well as symptomatic infec-
Lab B. Of all specimens, 8853 (55.2%) were from women. The tions in persons who either did not seek care or in whom SARS-
age group of 0-18 years comprised the smallest number of CoV-2 viral testing was not performed. It is possible that
specimens (n = 1205, 7.5%), with the age group of 65 years and false-positive ELISA results could lead us to overestimate
older comprising the largest number (n = 5845, 36.5%). Labo- seroprevalence and infections. The estimates are the first
ratory catchment areas as determined by the number of speci- reported from these 10 sites, from which specimens are to be
mens were predominantly major cities and their metro areas, collected at a variety of time points.26
including some suburban or exurban counties for CA, FL, MN, The results of several US seroprevalence studies have been
NY, PA, and WA. Laboratories receiving specimens from the released, including those conducted in Santa Clara County
entire state (CT, LA, MO, and UT) received specimens from areas (California), Idaho, Los Angeles (California), and New York.27-31
in numbers approximately proportionate to state population As of early July 2020, 3 of these 4 studies had only been posted
density (eFigure 1 in the Supplement). as preprints without peer review.27-30 Studies have used dif-
Table 2 shows the seroprevalence estimates by sex and age ferent assays and participant selection methods. The Santa
as well as fully adjusted estimates. Seroprevalence ranged from Clara County study, conducted April 3 and 4, 2020, approxi-
1.0% (95% CI, 0.3%-2.4%) in CA to 6.9% (95% CI, 5.0%-8.9%) mately 5 weeks after the first case of community transmis-
in NY. Seroprevalence estimates fell within this range for the re- sion of COVID-19 was detected in the San Francisco Bay area,
maining 8 sites. There was no clear association between sero- estimated a seroprevalence rate of 2.5% to 4.2%.27 The au-
prevalence by age and sex across sites (Figure 2). In NY, there thors noted that seroprevalence estimates were largely driven
was a significant difference in fully adjusted seroprevalence be- by estimates of test performance characteristics, which is to
tween specimens obtained from Lab A (11.5%) and Lab B (5.7%) be expected, particularly in a low-prevalence setting. A study
(P < .01). In WA, there was no difference in fully adjusted sero- in New York City conducted between late February and mid-
prevalence between specimens obtained from Lab A or Lab B April 2020 showed seroprevalence estimates of 2.2% and 10.1%
(1.9% vs 1.5%; P = .47) (eTable 1 in the Supplement). for the weeks ending March 29 and April 5, respectively30; the

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online July 21, 2020 E5

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 08/02/2020

 E6
Table 2. Sex-Specific, Age-Specific, and Age- and Sex-Standardized Seroprevalence Estimates, Adjusted for Assay Performance Characteristics, for Specimens Collected in 10 Geographic Sitesa,b

Persons reactive, % (95% CI)a

Western New York City Philadelphia Minneapolis-St Paul-
Washington metro area metro area San Francisco Bay St Cloud metro area
Characteristic State (New York) Louisiana South Florida (Pennsylvania) Missouri Utah area (California) Connecticut (Minnesota)
Sex
Male 1.4 (0.8-2.4) 5.9 (4.5-7.6) 6.8 (4.2-9.3) 2.2 (1.1-3.6) 3.0 (1.3-5.2) 3.1 (1.8-4.6) 2.2 (0.9-3.6) 1.2 (0.4-2.7) 5.7 (3.8-7.6) 0.7 (0-2.3)
Female 1.7 (0.7-1.9) 5.7 (4.2-7.0) 7.0 (4.7-9.4) 2.2 (1.2-3.4) 1.9 (0.7-3.7) 2.6 (1.5-3.7) 2.5 (1.2-4.1) 0.7 (0.2-1.9) 4.1 (2.6-5.9) 2.7 (1.2-4.8)
Research Original Investigation

Age group, y
0-18 0.7 (0-2.5)c 2.7 (0.9-5.0)c 2.8 (0-11.5) 2.4 (0-7.8) 2.2 (0-6.9) 1.4 (0-4.1) NAd 1.7 (0-7.7) 0.8 (0-2.9) 5.8 (0-14.3)
c c
19-49 1.3 (0.7-2.3) 8.3 (6.2-10.2) 7.4 (4.7-10.0) 0.9 (0.2-2.2) 5.9 (2.4-9.8) 3.4 (1.4-5.5) 1.8 (0.6-3.5) 1.1 (0-2.6) 6.1 (3.1-9.3) 2.3 (0.8-4.2)
50-64 0.9 (0.3-1.9) 6.5 (4.3-9.6) 8.3 (4.5-11.9) 2.0 (0.3-4.0) 0.8 (0-2.8) 2.0 (0.5-3.8) 2.9 (0.9-5.2) 0.7 (0-2.4) 8.1 (4.8-11.6) 0.7 (0-2.8)

Downloaded From: https://jamanetwork.com/ on 08/02/2020

≥65 1.7 (0.9-2.7) 3.7 (2.2-5.2) 4.4 (1.5-8.0) 3.0 (1.7-4.5) 1.6 (0.3-3.5) 3.2 (1.9-4.6) 2.7 (0.9-5.0) 0.9 (0.2-2.5) 4.2 (2.3-6.0) 1.0 (0-3.2)
Overall estimate, age 1.1 (0.7-1.9)e 6.9 (5.0-8.9)e 5.8 (3.9-8.2)f 1.9 (1.0-3.2)e 3.2 (1.7-5.2)e 2.7 (1.7-3.9)f 2.2 (1.2-3.4)f 1.0 (0.3-2.4)e 4.9 (3.6-6.5)f 2.4 (1.0-4.5)e
and sex standardizedb
d
Abbreviation: NA, not available. The number of specimens for persons aged 0-18 y was inadequate, and those aged ⱕ18 y were excluded from
a 20 the analysis.
A specimen was considered reactive if, on confirmatory testing with the assay described in Freeman et al, at
e

JAMA Internal Medicine Published online July 21, 2020 (Reprinted)

a background corrected optical density of 0.4, at a serum dilution of 1:100, it had a signal to threshold ratio of >1. Standardized to the age and sex distribution of the counties in each area from which most specimens originated
b and adjusted for test performance characteristics as described above.
All estimates are adjusted for test performance characteristics (specificity, 99.3%; 95% CI, 98.3%-99.9%;
f
sensitivity, 96.0%; 95% CI, 90.0%-98.9%). Standardized to the age and sex distribution of the state and adjusted for test performance characteristics as
c described above.
A subset of samples did not have age by year available and were instead classified as age 5-17, 18-49, 50-64, and
ⱖ65 y. The 5-17 and 19-49 y age groups were combined with those aged 5-18 and 19-49 y, respectively.

© 2020 American Medical Association. All rights reserved.

jamainternalmedicine.com
Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, 2020
 Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, 2020 Original Investigation Research

Figure 2. Strata-Specific Estimates of Seroprevalence to Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Antibodies
in 10 Geographic Sites

A Estimates of seroprevalence by sex B Estimates of seroprevalence by age group

State Date State Date
New York City metro March 30-April 1 New York City metro area, New York March 30-April 1
area, New York ≥65 y
50-64 y
Men
19-49 y
Women 0-18 y
Louisiana April 1-8
Louisiana April 1-8
≥65 y
Men 50-64 y
19-49 y
Women 0-18 y
Connecticut April 26-May 3 Connecticut April 26-May 3
≥65 y
Men 50-64 y
19-49 y
Women
0-18 y
Philadelphia metro April 13-25 Philadelphia metro area, Pennsylvania April 13-25
area, Pennsylvania ≥65 y
50-64 y
Men
19-49 y
Women 0-18 y
Missouri April 20-26
Missouri April 20-26
≥65 y
Men 50-64 y
19-49 y
Women 0-18 y
Minneapolis-St Paul- April 30-May 12 Minneapolis-St Paul-St Cloud April 30-May 12
metro area, Minnesota
St Cloud metro area,
Minnesota ≥65 y
50-64 y
Men 19-49 y
0-18 y
Women
South Florida April 6-10
South Florida April 6-10 ≥65 y
50-64 y
Men
19-49 y
Women 0-18 y
Utah April 20-May 3
Utah April 20-May 3 ≥65 y
Men 50-64 y
19-49 y
Women 0-18 y
Western Washington March 23-April 1
Western Washington March 23-April 1
≥65 y
Men 50-64 y
19-49 y
Women
0-18 y
San Francisco Bay April 23-27 San Francisco Bay area, California April 23-27
area, California ≥65 y
50-64 y
Men
19-49 y
Women 0-18 y

0 2 4 6 8 0 2 4 6 8
Prevalence, % Prevalence, %

A, Estimates of seroprevalence to SARS-CoV-2 antibodies by sex, from highest to lowest overall seroprevalence. B, Strata-specific estimates of seroprevalence to
SARS-CoV-2 antibodies by age group, from highest to lowest overall seroprevalence.

NY specimen collection period in our study was March 23 to occurring several weeks later, during a period when SARS-
April 1. Another study in New York state used sera collected CoV-2 was circulating widely in New York City.
from April 19 to April 28, approximately 8 weeks after com- At present, the relationship between detectable antibod-
munity transmission was first identified in New York City. A ies to SARS-CoV-2 and protective immunity against future in-
seroprevalence of 22.7% was estimated.31 The higher sero- fection is not known.25 Extrapolating these estimates to make
prevalence found in this study compared with our results may assumptions about population immunity should not be done
reflect a different study population and specimen collection until more is known about the correlations between the

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online July 21, 2020 E7

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 08/02/2020

 E8
Table 3. Estimated Number of Infections Based on Seroprevalence Estimates and Comparison With the Number of Reported Cases as of the Last Date of Specimen Collection for 10 Sites
Estimated Cases reported by Estimated
Catchment seroprevalence, % date of last specimen Estimated cumulative infections/reported
Site Catchment description population, No. (95% CI)a collection, No.b,c infections, No. (95% CI) cases, No. (range)d
Western Washington King, Snohomish, Pierce, Kitsap, 4 273 548 1.1 (0.7-1.9) 4308 48 291 (29 915-82 907) 11.2 (6.9-19.2)
Grays Harbor counties
New York City metro area Manhattan, Bronx, Queens, Kings, 9 260 870 6.9 (5.0-8.9) 53 803 641 778 (464 896-826 070) 11.9 (8.6-15.4)
(New York) Nassau counties
Louisiana Statewide 4 644 049 5.8 (3.9-8.2) 17 030 267 033 (179 725-382 205) 15.7 (10.6-22.4)
Research Original Investigation

South Florida Miami-Dade, Broward, Palm Beach, 6 345 345 1.9 (1.0-3.2) 10 525 117 389 (63 453-204 955) 11.2 (6.0-19.5)
Martin counties
Philadelphia metro area Bucks, Chester, Cumberland, Delaware, 4 910 139 3.2 (1.7-5.2) 22 987 156 633 (82 981-254 836) 6.8 (3.6-11.1)
(Pennsylvania) Lancaster, Montgomery, Philadelphia
counties

Downloaded From: https://jamanetwork.com/ on 08/02/2020

Missouri Statewide 6 110 800 2.7 (1.7-3.9) 6794 161 936 (100 828-235 877) 23.8 (14.8-34.7)
Utah Adults aged ≥19 y (statewide) 2 173 082 2.2 (1.2-3.4) 4493e 47 373 (26 294-74 537) 10.5 (5.5-15.5)
San Francisco Bay area Alameda, Contra Costa, San Francisco, 6 662 454 1.0 (0.3-2.4) 7151 64 626 (22 652-162 564) 9.0 (3.2-22.7)
(California) San Mateo, Marin, Santa Clara counties
Connecticut Statewide 3 562 989 4.9 (3.6-6.5) 29 287 176 012 (128 624-232 307) 6.0 (4.3-7.8)
Minneapolis-St Paul-St Cloud Anoka, Benton, Carver, Chisago, Dakota, 3 857 479 2.4 (1.0-4.5) 8880 90 651 (37 803-173 587) 10.2 (4.3-19.5)

JAMA Internal Medicine Published online July 21, 2020 (Reprinted)

metro area (Minnesota) Goodhue, Hennepin, Isanti, Le Sueur,
McLeod, Mille Lacs, Ramsey, Rice, Scott,
Sherburne, Stearns, Steele, Washington,
Wright counties
a
Standardized to the age and sex distribution of the counties in each region from which most specimens Minneapolis-St. Paul-St. Cloud metro area: May 12, 2020.
originated. All estimates are adjusted for test performance characteristics (specificity, 99.3%; 95% CI, d
Estimated number of times greater for infections suggested by seroprevalence estimates compared with
98.3%-99.9%; sensitivity, 96.0%; 95% CI, CI 90.0%-98.9%). reported cases. Range derived from 95% CIs of seroprevalence estimates.
b
Information from USA Facts.22 e
The number of specimens for persons aged 0-18 y was inadequate, and those aged <18 y were excluded from the
c
Dates of last specimen collection, by site, were western Washington: April 1, 2020; New York City metro area: analysis. Estimates are for adults aged ⱖ19 y. Case report data were not available by age in all cases and were
April 1, 2020; Louisiana: April 8, 2020; south Florida: April 10, 2020; Philadelphia metro area: April 25, 2020; estimated by publicly reported age grouping.
Missouri: April 25, 2020; Utah: May 3, 2020; San Francisco Bay area: April 27, 2020; Connecticut: May 3, 2020;

© 2020 American Medical Association. All rights reserved.

jamainternalmedicine.com
Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, 2020
 Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, 2020 Original Investigation Research

presence, titer, and duration of antibodies and protection minimal. In addition, although the overall sample size was
against this novel, emerging disease. large, in some sites, there were few specimens from persons
The timing of the development of SARS-CoV-2–specific 18 years or younger, which limited our ability to estimate se-
antibodies is variable; it is unknown when infection roprevalence among children. Furthermore, at this stage in the
occurred for individuals in this study. Although humoral pandemic, infections may not be evenly distributed even
response kinetics to SARS-CoV-2 infection are not well within these geographic sites. Thus, seroprevalence esti-
understood, reactive IgA, IgM, and IgG antibodies have been mates for large geographic sites may not be accurate if the ma-
detected as soon as 1 day after symptom onset.32 In other jority of samples come from specific areas with higher infec-
studies, neutralizing antibodies were detected 10 to 15 days tion rates. We also had limited geographic data on a subset of
after symptom onset; the median time to development of specimens from Lab B for NY and WA, which may have been
total antibody, IgM, and IgG has been estimated as 11, 12, and drawn from a larger geographic site than those from Lab A with
14 days, respectively.23,25 zip code–level data. The inclusion of some specimens from
We compared the number of estimated cases in the popu- other sites in WA and NY states, especially sites of lower se-
lation based on our seroprevalence estimates with the re- roprevalence around NY, may lead to inaccurate seropreva-
ported cases as of the last day of specimen collection. From lence estimates for these areas and may explain the differ-
this comparison, we estimated that there were from 6 times ences in the seroprevalence estimates between Lab A and Lab
as many SARS-CoV-2 infections as reported cases in CT to B for NY. Finally, the representation of specific geographic pock-
24 times the number of infections as reported cases in MO. ets may not be the same between the 2 commercial laborato-
Specimen collection in CT started later than other sites, and ries, and underlying patient populations may differ between
lower underascertainment estimates for CT may reflect in- the laboratories; therefore, combining results from Lab A and
creasing availability of testing as the pandemic progressed. Lab B is problematic. Follow-up serosurveys will include zip
These estimates of underascertainment are conservative; they code data for all specimens.
would be higher if an earlier date had been used to take into It is possible that the ELISA may exhibit cross-reactivity
account infected persons who had not yet developed detect- with antibodies to other common human coronaviruses; there-
able antibodies at the time of specimen collection. Our sero- fore, some results may represent a false-positive result for
prevalence estimates are more likely to reflect infections that SARS-CoV-2, potentially leading to overestimation of the ac-
occurred a minimum of 1 to 2 weeks prior to the specimen tual seroprevalence. The assay used has high specificity for
collection. SARS-CoV-2, and cross-reactivity with common coronavi-
ruses generated results below the cutoff used for this assay.20
Limitations However, even with a highly specific test, the effect of false-
Our study has limitations that are associated with both the positive test results may be more marked in lower prevalence
samples and with the tests used. The specimens were col- settings, including CA, FL, and WA. We did consider the per-
lected for clinical purposes from persons seeking health care formance characteristics of the ELISA when making seroprev-
and were shared with the CDC with minimal accompanying alence estimates. Although the assay has high sensitivity (96%),
data. No data on recent symptomatic illness, underlying con- it is not 100% sensitive and thus will not detect all persons with
ditions, or possible COVID-19 exposures were available. It is antibodies. Finally, several early reports indicate that not all
possible that specimens were drawn from patients seeking care persons with SARS-CoV-2 infection mount an antibody re-
for suspected COVID-19 symptoms, potentially biasing re- sponse, and antibody titers may be lower in those with milder
sults, particularly in settings such as NY where disease inci- disease; furthermore, levels of IgG and neutralizing antibod-
dence was higher. Lab B sampled sera from metabolic panels ies decrease in some persons within 2 to 3 months after
taken at routine outpatient visits; Lab A sampled randomly with infection.25,33-35 For these reasons, seroprevalence estimates
respect to clinical test type and admission status. Residual clini- may underestimate the proportion of persons with prior in-
cal specimens from screening or routine care are more likely fection in any population.
to come from persons who require monitoring for chronic Tracking population seroprevalence for SARS-CoV-2 in-
medical conditions despite the ongoing pandemic. These per- fection serially, in a variety of specific geographic sites, should
sons may not be representative of the general population, in- inform models of transmission dynamics and policy deci-
cluding in their health care seeking and social distancing be- sions regarding the effects of social distancing and other pre-
havior, immune response to infection, and disease exposure ventive measures. To inform understanding of the epidemi-
risk. Representativeness may vary by age group as well. There- ology of COVID-19, the CDC plans to conduct repeated sampling
fore, our seroprevalence estimates should be confirmed and in these and other geographic sites around the US on an
extended by other studies, including serosurveys that use ongoing basis.26
targeted sampling frames to enroll more representative
populations.33 For Lab B samples from NY and WA, it is pos-
sible that more than 1 specimen was from the same indi-
vidual, as samples were not deduplicated. Given the large num-
Conclusions
bers of specimens from each site and that the potential for In conclusion, the seroprevalence estimates we report sug-
duplication should be unbiased with respect to SARS-CoV-2 in- gest that at the time of specimen collection from March to early
fection, the influence on seroprevalence estimates is likely May 2020, a large majority of persons in 10 diverse geo-

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online July 21, 2020 E9

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 08/02/2020

 Research Original Investigation Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, 2020

graphic sites in the US had not been infected with SARS- population. Because persons often do not know if they are in-
CoV-2. The estimated number of infections, however, was fected with SARS-CoV-2, the public should continue to take
much greater than the number of reported cases in all sites. steps to help prevent the spread of COVID-19, such as wear-
This finding may reflect persons who had mild or no illness or ing cloth face coverings when outside the home, remaining
who did not seek medical care or undergo testing but who still 6 feet apart from other people, washing hands frequently, and
may have contributed to ongoing virus transmission in the staying home when sick.

ARTICLE INFORMATION Conflict of Interest Disclosures: Dr Wiesman www.cdph.ca.gov/Programs/OPA/Pages/NR20-

Accepted for Publication: July 7, 2020. reported receiving grants from US Department of 006.aspx
Health and Human Services during the conduct of 3. Additional cases of COVID-19 in Washington
Published Online: July 21, 2020. the study. No other disclosures were reported.
doi:10.1001/jamainternmed.2020.4130 State. News release. Washington State Department
Funding/Support: This work was supported by of Health; February 28, 2020. Accessed July 10,
Author Affiliations: CDC COVID-19 Response the Centers for Disease Control and Prevention, 2020. https://www.doh.wa.gov/Newsroom/
Team, Centers for Disease Control and Prevention, Atlanta, Georgia. Articles/ID/1103/Additional-Cases-of-COVID-19-in-
Atlanta, Georgia (Havers, Reed, Lim, Montgomery, Washington-State
Klena, Hall, Fry, Cannon, Chiang, Gibbons, Role of the Funder/Sponsor: The Centers for
Krapiunaya, Morales-Betoulle, Roguski, Rasheed, Disease Control and Prevention was involved in 4. During coronavirus briefing, Governor Cuomo
Freeman, Lester, Mills, Carroll, Owen, Johnson, the design and conduct of the study; collection, signs $40 million emergency management
Semenova, Schiffer, Thornburg); Florida management, analysis, and interpretation of the authorization for coronavirus response. News
Department of Health, Tallahassee (Blackmore); data; preparation, review, and approval of the release; March 3, 2020. Accessed July 10, 2020.
New York State Department of Health, Albany manuscript; and decision to submit the manuscript https://www.governor.ny.gov/news/during-
(Blog); Division of State and Local Readiness, for publication. coronavirus-briefing-governor-cuomo-signs-40-
Centers for Disease Control and Prevention, Group Information: The author members of the million-emergency-management-authorization
Atlanta, Georgia (Chai); Utah Department of Health, CDC COVID-19 Response Team are Drs Havers, 5. Jorden MA, Rudman SL, Villarino E, et al; CDC
Salt Lake City (Dunn); Louisiana Department of Reed, Lim, Montgomery, Klena, Hall, Fry, Chiang, COVID-19 Response Team. Evidence for limited
Health, New Orleans (Hand); California Department Morales-Betoulle, Rasheed, Freeman, Lester, Mills, early spread of COVID-19 within the United States,
of Health, Richmond (Jain); Washington State Carroll, Owen, Johnson, Semenova, and Thornburg; January-February 2020. MMWR Morb Mortal Wkly
Department of Health, Tumwater (Lindquist); Mses Cannon, Gibbons, Krapiunaya, and Roguski; Rep. 2020;69(22):680-684. doi:10.15585/mmwr.
Minnesota Department of Health, St Paul (Lynfield); and Mr Schiffer. The author members of the State mm6922e1
Florida Department of Health, Tallahassee Collaborator Group are Drs Blackmore, Blog, Chai, 6. Nationally Notifiable Diseases Surveillance
(Pritchard); Louisiana Department of Health, New Dunn, Jain, Lindquist, Lynfield, Sosa, Turabelidze, System. Coronavirus disease 2019 (COVID-19).
Orleans (Sokol); Connecticut Department of Public Watkins, Wiesman, Williams, and Yendell; Mses Accessed June 10, 2020. https://wwwn.cdc.gov/
Health, Hartford (Sosa); Missouri Department of Hand and Sokol; and Mr Pritchard. nndss/conditions/coronavirus-disease-2019-
Health and Senior Services, Jefferson City Disclaimer: The findings and conclusions in the covid-19/
(Turabelidze); Pennsylvania Department of Health, article are those of the authors and do not
Harrisburg (Watkins); Washington State 7. Sutton D, Fuchs K, D’Alton M, Goffman D.
necessarily represent the views of the US Centers Universal screening for SARS-CoV-2 in women
Department of Health, Tumwater (Wiesman); for Disease Control and Prevention.
Missouri Department of Health and Senior Services, admitted for delivery. N Engl J Med. 2020;382(22):
Jefferson City (Williams); Minnesota Department of Additional Contributions: We thank LabCorp and 2163-2164. doi:10.1056/NEJMc2009316
Health, St Paul (Yendell). Quest for supplying specimens. From Quest: 8. Mizumoto K, Kagaya K, Zarebski A, Chowell G.
William A. Meyer III, PhD, Larry A. Hirsch, BS, Taylor Estimating the asymptomatic proportion of
Author Contributions: Drs Havers and Lim had full Hwang, BS, and Janet M. Rochat, MS. From the
access to all of the data in the study and take coronavirus disease 2019 (COVID-19) cases on
New York City Department of Mental Health and board the Diamond Princess cruise ship, Yokohama,
responsibility for the integrity of the data and the Hygiene: Marcelle Layton, MD. From the Centers
accuracy of the data analysis. Japan, 2020. Euro Surveill. 2020;25(10). doi:10.
for Disease Control and Prevention Molecular 2807/1560-7917.ES.2020.25.10.2000180
Study concept and design: Havers, Reed, Lim, Hall, Pathogenesis and Immunology Research
Fry, Owen, Blackmore, Blog, Lindquist, Turabelidze, Laboratory team: Bailey Alston, MS, Muyiwa 9. Kimball A, Hatfield KM, Arons M, et al; Public
Wiesman, Schiffer, Thornburg. Ategbole, MPH, Shanna Bolcen, MSPH, Darbi Health—Seattle & King County; CDC COVID-19
Acquisition, analysis, or interpretation of data: Boulay, BS, Peter Browning, BS, Li Cronin, MS, Investigation Team. Asymptomatic and
Havers, Reed, Lim, Montgomery, Klena, Cannon, Ebenezer David, PhD, Rita Desai, BS, Monica presymptomatic SARS-CoV-2 infections in residents
Chiang, Gibbons, Krapiunaya, Morales-Betoulle, Epperson, PhD, Yamini Gorantla, PhD, Lily Jia, MS, of a long-term care skilled nursing facility—King
Roguski, Rasheed, Freeman, Lester, Mills, Carroll, Han Li, PhD, Pete Maniatis, MS, Jeff Martin, BA, County, Washington, March 2020. MMWR Morb
Owen, Johnson, Semenova, Blog, Chai, Dunn, Kimberly Moss, BS, Kristina Ortiz, MS, Palak Patel, Mortal Wkly Rep. 2020;69(13):377-381. doi:10.
Hand, Jain, Lynfield, Pritchard, Sokol, Sosa, MS, So Hee Park, BS, Yunlong Qin, PhD, Evelene 15585/mmwr.mm6913e1
Watkins, Williams, Yendell, Schiffer, Thornburg. Steward-Clark, MS, Heather Tatum, BS, Andrew 10. New York City confirms a second COVID-19
Drafting of the manuscript: Havers, Lim, Klena, Fry, Vogan, MS, and Briana Zellner, PhD. We also thank case. News release; March 3, 2020. Accessed July
Thornburg. Ian E. Fellows, PhD, of Fellows Statistics for input on 14, 2020. https://www1.nyc.gov/site/doh/about/
Critical revision of the manuscript for important statistical methodology. These individuals were not press/pr2020/city-confirms-second-covid-19-case.
intellectual content: All authors. compensated directly by CDC for their participation page
Statistical analysis: Havers, Lim, Gibbons, Roguski, in this specific study.
Schiffer. 11. Health officials confirms two new presumptive
Obtained funding: Havers. positive COVID-19 cases in Louisiana as state
REFERENCES increases testing for the virus. News release; March
Administrative, technical, or material support:
Havers, Montgomery, Klena, Cannon, Chiang, 1. Harcourt J, Tamin A, Lu X, et al. Severe acute 10, 2020. Accessed July 14, 2020. http://ldh.la.gov/
Morales-Betoulle, Mills, Carroll, Owen, Johnson, respiratory syndrome coronavirus 2 from patient index.cfm/newsroom/detail/5463
Semenova, Blackmore, Blog, Chai, Dunn, Jain, with coronavirus disease, United States. Emerg 12. Department of Health announces two
Lindquist, Pritchard, Sokol, Sosa, Turabelidze, Infect Dis. 2020;26(6):1266-1273. doi:10.3201/ presumptive positive COVID-19 cases in Florida.
Watkins, Wiesman, Williams, Schiffer, Thornburg. eid2606.200516 News release; March 1, 2020. Accessed July 14,
Study supervision: Havers, Reed, Montgomery, Hall, 2. CDC confirms possible first instance of COVID-19 2020. http://www.floridahealth.gov/newsroom/
Fry, Johnson, Blog, Pritchard, Schiffer, Thornburg. community transmission in California 2020. News 2020/03/030120-two-presumptive-positive-
release. California Department of Public Health; covid19-cases.pr.html
February 26, 2020. Accessed July 10, 2020. https://

E10 JAMA Internal Medicine Published online July 21, 2020 (Reprinted) jamainternalmedicine.com

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 08/02/2020

 Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, 2020 Original Investigation Research

13. Philadelphia announces first case of COVID 19 investigations and serosurveillance. bioRxiv. 29. Sood N, Simon P, Ebner P, et al. Seroprevalence
coronavirus 2. News release; March 10, 2020. Preprint posted online April 25, 2020. doi:10.1101/ of SARS-CoV-2–specific antibodies among adults in
Accessed July 14, 2020. https://www.phila.gov/ 2020.04.24.057323 Los Angeles County, California, on April 10-11, 2020.
2020-03-10-philadelphia-announces-first-case-of- 21. Wrapp D, Wang N, Corbett KS, et al. Cryo-EM JAMA. 2020;323(23):2425-2427. doi:10.1001/jama.
covid-19-coronavirus-2/ structure of the 2019-nCoV spike in the prefusion 2020.8279
14. Governor Parson signs executive order 20-02 conformation. Science. 2020;367(6483):1260-1263. 30. Stadlbauer D, Tan J, Jiang K, et al.
declaring a state of emergency in Missouri. News doi:10.1126/science.abb2507 Seroconversion of a city: longitudinal monitoring
release; March 13, 2020. Accessed July 14, 2020. 22. USA Facts. Coronavirus locations: COVID-19 of SARS-CoV-2 seroprevalence in New York City.
https://governor.mo.gov/press-releases/archive/ map by county and state. Accessed April 22, 2020. medRxiv. Preprint posted online June 29, 2020.
governor-parson-signs-executive-order-20-02- https://usafacts.org/visualizations/coronavirus- doi:10.1101/2020.06.28.20142190
declaring-state-emergency covid-19-spread-map/ 31. Rosenberg ES, Tesoriero JM, Rosenthal EM, et al
15. New case of COVID-19 in Summit County signals 23. Zhao J, Yuan Q, Wang H, et al. Antibody Cumulative incidence and diagnosis of SARS-CoV-2
community spread. News release; March 14, 2020. responses to SARS-CoV-2 in patients of novel infection in New York. Ann Epidemiol. Published
Accessed July 14, 2020. https://health.utah.gov/ coronavirus disease 2019. Clin Infect Dis. Published online June 17, 2020. doi:10.1016/j.annepidem.2020.
featured-news/new-case-of-covid-19-in-summit- online March 28, 2020. doi:10.1093/cid/ciaa344 06.004
county-signals-community-spread 32. Guo L, Ren L, Yang S, et al. Profiling early
24. Sethuraman N, Jeremiah SS, Ryo A.
16. Governor Lamont provides update on Interpreting diagnostic tests for SARS-CoV-2. JAMA. humoral response to diagnose novel coronavirus
Connecticut’s Coronavirus preparedness and 2020;323(22):2249-2251. doi:10.1001/jama.2020. disease (COVID-19). Clin Infect Dis. Published online
response efforts. News release; March 11, 2020. 8259 March 21, 2020. doi:10.1093/cid/ciaa310
Accessed July 14, 2020. https://portal.ct.gov/ 33. Centers for Disease Control and Prevention.
Office-of-the-Governor/News/Press-Releases/ 25. Wu F, Wang A, Liu M, et al. Neutralizing
antibody responses to SARS-CoV-2 in a COVID-19 Community Assessment for Public Health
2020/03-2020/Governor-Lamont-Coronavirus- Emergency Response (CASPER): sampling
Update-March-11-2020-3PM recovered patient cohort and their implications.
medRxiv. Preprint posted online April 20, 2020. methodology. Accessed May 27, 2020. https://
17. Health officials confirm third case of COVID-19 doi:10.1101/2020.03.30.20047365 www.cdc.gov/nceh/casper/sampling-methodology.
in Minnesota. News release; March 10, 2020. htm
https://www.health.state.mn.us/news/pressrel/ 26. Centers for Disease Control and Prevention.
Large-scale geographic seroprevalence surveys. 34. Long QX, Tang XJ, Shi QL, et al. Clinical and
2020/covid031020.html immunological assessment of asymptomatic
Accessed June 8, 2020. https://www.cdc.gov/
18. Centers for Disease Control and Prevention. coronavirus/2019-ncov/cases-updates/geographic- SARS-CoV-2 infections. Nat Med. Published online
Commercial laboratory seroprevalence survey data. seroprevalence-surveys.html June 18, 2020. doi:10.1038/s41591-020-0965-6
Accessed July 5, 2020. https://www.cdc.gov/ 35. To KK, Tsang OT, Leung WS, et al. Temporal
coronavirus/2019-ncov/cases-updates/commercial- 27. Bendavid E, Mulaney B, Sood N, et al. COVID-19
antibody seroprevalence in Santa Clara County, profiles of viral load in posterior oropharyngeal
lab-surveys.html saliva samples and serum antibody responses
California. medRxiv. Preprint posted online April 30,
19. Havers FP, Reed C, Lim TW, et al. 2020. doi:10.1101/2020.04.14.20062463 during infection by SARS-CoV-2: an observational
Seroprevalence of antibodies to SARS-CoV-2 in six cohort study. Lancet Infect Dis. 2020;20(5):565-574.
sites in the United States, March 23-May 3, 2020. 28. Bryan A, Pepper G, Wener MH, et al. doi:10.1016/S1473-3099(20)30196-1
medRxiv. Preprint posted online June 26, 2020. Performance characteristics of the Abbott Architect
doi:10.1101/2020.06.25.20140384 SARS-CoV-2 IgG assay and seroprevalence testing in
Idaho. medRxiv. Preprint posted online May 2,
20. Freeman B, Lester S, Mills L, et al. Validation of 2020. doi:10.1101/2020.04.27.20082362
a SARS-CoV-2 spike ELISA for use in contact

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online July 21, 2020 E11

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 08/02/2020