CoChat
Sign Up
Cited by 389Data-Driven Disease SurveillanceCOVID-19 Digital Contact TracingMobile Health and mHealth Applications

The DOI matches its Crossref record; 19 of the 19 references it cites were found in a registry or an index.

Rapid implementation of mobile technology for real-time epidemiology of COVID-19

Open access
Drew, David A., Nguyen, Long H., Steves, Claire J., Menni, Cristina, Freydin, Maxim, Varsavsky, Thomas, Sudre, Carole H., Cardoso, M. Jorge, Ourselin, Sebastien, Wolf, Jonathan, Spector, Tim D., Chan, Andrew T., COPE Consortium, Chan, Andrew T., Drew, David A., Nguyen, Long H., Joshi, Amit D., Guo, Chuan-Guo, Ma, Wenjie, Lo, Chun-Han, Mehta, Raaj S., Kwon, Sohee, Sikavi, Daniel R., Magicheva-Gupta, Marina V., Fatehi, Zahra S., Flynn, Jacqueline J., Leonardo, Brianna M., Albert, Christine M., Andreotti, Gabriella, Beane-Freeman, Laura E., Balasubramanian, Bijal A., Brownstein, John S., Bruinsma, Fiona, Cowan, Annie N., Deka, Anusila, Ernst, Michael E., Figueiredo, Jane C., Franks, Paul W., Gardner, Christopher D., Ghobrial, Irene M., Haiman, Christopher A., Hall, Janet E., Deming-Halverson, Sandra L., Kirpach, Brenda, Lacey, James V., Le Marchand, Loic, Marinac, Catherine R., Martinez, Maria Elena, Milne, Roger L., Murray, Anne M., Nash, Denis, Palmer, Julie R., Patel, Alpa V., Rosenberg, Lynn, Sandler, Dale P., Sharma, Shreela V., Schurman, Shepherd H., Wilkens, Lynne R., Chavarro, Jorge E., Eliassen, A. Heather, Hart, Jamie E., Kang, Jae Hee, Koenen, Karestan C., Kubzansky, Laura D., Mucci, Lorelei A., Ourselin, Sebastien, Rich-Edwards, Janet W., Song, Mingyang, Stampfer, Meir J., Steves, Claire J., Willett, Walter C., Wolf, Jonathan, Spector, Tim·2020·Science
DOI 10.1126/science.abc0473PMC PMC7200009

Published version of a preprint

Open this paper in CoChat — free account

Abstract

Mobile symptom tracking: The rapidity with which severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spreads through a population is defying attempts at tracking it, and quantitative polymerase chain reaction testing so far has been too slow for real-time epidemiology. Taking advantage of existing longitudinal health care and research patient cohorts, Drew et al. pushed software updates to participants to encourage reporting of potential coronavirus disease 2019 (COVID-19) symptoms. The authors recruited about 2 million users (including health care workers) to the COVID Symptom Study (previously known as the COVID Symptom Tracker) from across the United Kingdom and the United States. The prevalence of combinations of symptoms (three or more), including fatigue and cough, followed by diarrhea, fever, and/or anosmia, was predictive of a positive test verification for SARS-CoV-2. As exemplified by data from Wales, United Kingdom, mathematical modeling predicted geographical hotspots of incidence 5 to 7 days in advance of official public health reports. Science , this issue p. 1362

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1126/science.abc0473

Cites

19 of 19 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Schematic of the participant workflow. After downloading the COVID Symptom Study app and providing consent, users are prompted to enter baseline demographic and clinical information and are serially queried about new or ongoing symptoms, testing results, and extent of isolation. Health care workers offer additional information about the intensity of their patient interactions, potential exposure to infected patients, and use of PPE. With informed consent, users also participating in a variety of ongoing cohorts or clinical trials (Nurses’ Health Study, TwinsUK, and others) have the option of linking COVID Symptom Study information to their extant research data.
Fig. 1 Schematic of the participant workflow. After downloading the COVID Symptom Study app and providing consent, users are prompted to enter baseline demographic and clinical information and are serially queried about new or ongoing symptoms, testing results, and extent of isolation. Health care workers offer additional information about the intensity of their patient interactions, potential exposure to infected patients, and use of PPE. With informed consent, users also participating in a variety of ongoing cohorts or clinical trials (Nurses’ Health Study, TwinsUK, and others) have the option of linking COVID Symptom Study information to their extant research data.CC BY · PubMed Central
Fig. 2 COVID Symptom Study use, reported symptoms, and testing results according to geographic location in the United Kingdom. Between 24 and 29 March 2020, more than 1.6 million unique individuals downloaded the application and shared clinical and demographic information, as well as daily symptoms and high-intensity occupational exposures (blue map). Population density of those presenting with any symptoms (left purple map) varied according to region, with widespread reports of fatigue, cough, and diarrhea, followed by anosmia and, relatively infrequently, fever (inset). Examination of individuals who reported complex symptoms (right purple map), defined as having cough or fever and at least one other symptom (diarrhea, anosmia, or fever), reveals areas that potentially need more testing. For the subset of the population that received a COVID-19 test (black map), areas with larger proportions of positive tests (orange map) appear to coincide with areas in which high proportions of the population reported complex symptoms. By contrast, some areas with low prevalence of complex symptoms have received higher rates of testing and, consequently, more negative tests (green map). This example of real-time visualization of data captured by the COVID Symptom Study may help public health and government officials reallocate resources, identify areas with unmet testing needs, and detect emerging hotspots.
Fig. 2 COVID Symptom Study use, reported symptoms, and testing results according to geographic location in the United Kingdom. Between 24 and 29 March 2020, more than 1.6 million unique individuals downloaded the application and shared clinical and demographic information, as well as daily symptoms and high-intensity occupational exposures (blue map). Population density of those presenting with any symptoms (left purple map) varied according to region, with widespread reports of fatigue, cough, and diarrhea, followed by anosmia and, relatively infrequently, fever (inset). Examination of individuals who reported complex symptoms (right purple map), defined as having cough or fever and at least one other symptom (diarrhea, anosmia, or fever), reveals areas that potentially need more testing. For the subset of the population that received a COVID-19 test (black map), areas with larger proportions of positive tests (orange map) appear to coincide with areas in which high proportions of the population reported complex symptoms. By contrast, some areas with low prevalence of complex symptoms have received higher rates of testing and, consequently, more negative tests (green map). This example of real-time visualization of data captured by the COVID Symptom Study may help public health and government officials reallocate resources, identify areas with unmet testing needs, and detect emerging hotspots.CC BY · PubMed Central
Fig. 3 Symptoms reported through the COVID Symptom Study app. By 27 March 2020, 265,851 individuals in the United Kingdom reported any symptom potentially associated with COVID-19 (top). Participants provided data on whether they were tested for COVID-19, as well as the test result. 1176 individuals reported having received a COVID-19 test (0.4% of those with symptoms). Symptom frequencies among those who tested positive (middle; n = 340) versus negative (bottom; n = 836) are shown.
Fig. 3 Symptoms reported through the COVID Symptom Study app. By 27 March 2020, 265,851 individuals in the United Kingdom reported any symptom potentially associated with COVID-19 (top). Participants provided data on whether they were tested for COVID-19, as well as the test result. 1176 individuals reported having received a COVID-19 test (0.4% of those with symptoms). Symptom frequencies among those who tested positive (middle; n = 340) versus negative (bottom; n = 836) are shown.CC BY · PubMed Central
Fig. 4 Predicting COVID-19 cases on the basis of real-time symptom reporting in Wales, United Kingdom. This time series (bar graph) displays the number of new confirmed cases (gray bars) reported by the Public Health Wales NHS Trust between 31 March 2020 and 20 April 2020. After 2 April, case numbers appear to have declined through 5 April. However, our symptom-based prediction model (18), developed from symptom reports from untested users of the COVID Symptom Study app, showed a high proportion of predicted COVID-19 cases in southern Wales on 1 April [dark red areas in (A)]. Six days later, Welsh health authorities reported a subsequent peak in cases over a 4-day period (6 to 9 April), driven primarily by these southern regions (colored bars). By 10 April, new confirmed cases across Wales declined. However, on the basis of reported symptoms (B), regions in South Wales still had high predicted levels of COVID-19, which became apparent as a second spike in confirmed cases between 15 and 16 April. As of 20 April (C), predicted COVID-19 prevalence across Wales according to symptom reporting appears to be low, which corresponds to a flattening of the cumulative incidence curve. However, several regions in southern Wales still have relatively high reports of symptoms and appear at risk for subsequent cases of COVID-19. Black dots on the maps represent the relative proportion of positive tests reported by health authorities across Wales that day by region. The prediction mapping included data from 1,339,670 users of the COVID Symptom Study on 1 April; 998,244 users on 10 April; and 1,234,918 users on 20 April. Public Health Wales NHS Trust data were current as of 21 April 2020 at 13:00 local time, taken from the “Rapid COVID-19 Virology - Public” dashboard (accessed via https://phw.nhs.wales/), and downloaded on 22 April 2020 at 12:30 p.m. Eastern Standard Time.
Fig. 4 Predicting COVID-19 cases on the basis of real-time symptom reporting in Wales, United Kingdom. This time series (bar graph) displays the number of new confirmed cases (gray bars) reported by the Public Health Wales NHS Trust between 31 March 2020 and 20 April 2020. After 2 April, case numbers appear to have declined through 5 April. However, our symptom-based prediction model (18), developed from symptom reports from untested users of the COVID Symptom Study app, showed a high proportion of predicted COVID-19 cases in southern Wales on 1 April [dark red areas in (A)]. Six days later, Welsh health authorities reported a subsequent peak in cases over a 4-day period (6 to 9 April), driven primarily by these southern regions (colored bars). By 10 April, new confirmed cases across Wales declined. However, on the basis of reported symptoms (B), regions in South Wales still had high predicted levels of COVID-19, which became apparent as a second spike in confirmed cases between 15 and 16 April. As of 20 April (C), predicted COVID-19 prevalence across Wales according to symptom reporting appears to be low, which corresponds to a flattening of the cumulative incidence curve. However, several regions in southern Wales still have relatively high reports of symptoms and appear at risk for subsequent cases of COVID-19. Black dots on the maps represent the relative proportion of positive tests reported by health authorities across Wales that day by region. The prediction mapping included data from 1,339,670 users of the COVID Symptom Study on 1 April; 998,244 users on 10 April; and 1,234,918 users on 20 April. Public Health Wales NHS Trust data were current as of 21 April 2020 at 13:00 local time, taken from the “Rapid COVID-19 Virology - Public” dashboard (accessed via https://phw.nhs.wales/), and downloaded on 22 April 2020 at 12:30 p.m. Eastern Standard Time.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 19 references

Cited by

Sorted by citations

See all 389

Related

Metadata

Authors
Drew, D. A., Nguyen, L. H., Steves, C. J., Menni, C., Freydin, M., Varsavsky, T., Sudre, C. H., Cardoso, M. J., Ourselin, S., Wolf, J., Spector, T. D., Chan, A. T., COPE Consortium, Chan, A. T., Drew, D. A., Nguyen, L. H., Joshi, A. D., Guo, C.-G., Ma, W., Lo, C.-H., Mehta, R. S., Kwon, S., Sikavi, D. R., Magicheva-Gupta, M. V., Fatehi, Z. S., Flynn, J. J., Leonardo, B. M., Albert, C. M., Andreotti, G., Beane-Freeman, L. E., Balasubramanian, B. A., Brownstein, J. S., Bruinsma, F., Cowan, A. N., Deka, A., Ernst, M. E., Figueiredo, J. C., Franks, P. W., Gardner, C. D., Ghobrial, I. M., Haiman, C. A., Hall, J. E., Deming-Halverson, S. L., Kirpach, B., Lacey, J. V., Le Marchand, L., Marinac, C. R., Martinez, M. E., Milne, R. L., Murray, A. M., Nash, D., Palmer, J. R., Patel, A. V., Rosenberg, L., Sandler, D. P., Sharma, S. V., Schurman, S. H., Wilkens, L. R., Chavarro, J. E., Eliassen, A. H., Hart, J. E., Kang, J. H., Koenen, K. C., Kubzansky, L. D., Mucci, L. A., Ourselin, S., Rich-Edwards, J. W., Song, M., Stampfer, M. J., Steves, C. J., Willett, W. C., Wolf, J., & Spector, T.
Affiliations
  • King's College London (GB)
  • Department of Twin Research and Genetic Epidemiology, King’s College London, Westminster Bridge Road, London SE1 7EH, UK.
  • Harvard University (US)
ORCID
Venue
Science
Publisher
American Association for the Advancement of Science (AAAS)
Published
2020-06-19
Volume
368
Issue
6497
Pages
1362-1367
ISSN
  • 0036-8075 (ISSN-L)
  • 1095-9203
DOI
10.1126/science.abc0473
PMC
PMC7200009
OpenAlex
W3022622144
Cited by
389
References
19
Citations by year
2026: 8 · 2025: 9 · 2024: 23 · 2023: 47 · 2022: 82 · 2021: 119 · 2020: 101
FWCI
62.76
Type
journal article
Language
en
Open access
Licence
Locations
  • Science (publishedVersion, cc-by)
  • medRxiv (acceptedVersion, cc-by-nc-nd)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Archivio Istituzionale della Ricerca (Universita Degli Studi Di Milano) (submittedVersion, other-oa)
  • UCL Discovery (University College London) (submittedVersion)
  • Research Portal (King's College London) (submittedVersion, cc-by)
Funders
  • Massachusetts General Hospital (Stuart and Suzanne Steele MGH Research Scholar Award)
  • Wellcome Trust Centre for Mitochondrial Research
  • Zoe Global Ltd.
Topics
  • Health Sciences › Medicine › Epidemiology › Data-Driven Disease Surveillance
  • Physical Sciences › Computer Science › Information Systems › COVID-19 Digital Contact Tracing
  • Health Sciences › Health Professions › General Health Professions › Mobile Health and mHealth Applications
Keywords
COVID-19 pandemic, SARS-CoV-2, mobile health apps, symptom monitoring, big data, risk factors, clinical outcomes, pandemic response
MeSH terms
Betacoronavirus, Big Data, COVID-19, SARS-CoV-2, Data Collection, United Kingdom, Humans, International Cooperation, Models, Theoretical, Pneumonia, Viral, United States, Global Health, Coronavirus Infections, Pandemics, Mobile Applications

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review