App-based COVID-19 syndromic surveillance and prediction of hospital admissions in COVID Symptom Study Sweden

Beatrice Kennedy; Hugo Fitipaldi; Ulf Hammar; Marlena Maziarz; Neli Tsereteli; Nikolay Oskolkov; Georgios Varotsis; Camilla A. Franks; Diem Nguyen; Lampros Spiliopoulos; Hans Olov Adami; Jonas Björk; Stefan Engblom; Katja Fall; Anna Grimby-Ekman; Jan Eric Litton; Mats Martinell; Anna Oudin; Torbjörn Sjöström; Toomas Timpka; Carole H. Sudre; Mark S. Graham; Julien Lavigne du Cadet; Andrew T. Chan; Richard Davies; Sajaysurya Ganesh; Anna May; Sébastien Ourselin; Joan Capdevila Pujol; Somesh Selvachandran; Jonathan Wolf; Tim D. Spector; Claire J. Steves; Maria F. Gomez; Paul W. Franks; Tove Fall

doi:10.1038/s41467-022-29608-7

App-based COVID-19 syndromic surveillance and prediction of hospital admissions in COVID Symptom Study Sweden

Beatrice Kennedy, Hugo Fitipaldi, Ulf Hammar, Marlena Maziarz, Neli Tsereteli, Nikolay Oskolkov, Georgios Varotsis, Camilla A. Franks, Diem Nguyen, Lampros Spiliopoulos, Hans Olov Adami, Jonas Björk, Stefan Engblom, Katja Fall, Anna Grimby-Ekman, Jan Eric Litton, Mats Martinell, Anna Oudin, Torbjörn Sjöström, Toomas TimpkaCarole H. Sudre, Mark S. Graham, Julien Lavigne du Cadet, Andrew T. Chan, Richard Davies, Sajaysurya Ganesh, Anna May, Sébastien Ourselin, Joan Capdevila Pujol, Somesh Selvachandran, Jonathan Wolf, Tim D. Spector, Claire J. Steves, Maria F. Gomez, Paul W. Franks, Tove Fall^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

8 Citations (Scopus)

Abstract

The app-based COVID Symptom Study was launched in Sweden in April 2020 to contribute to real-time COVID-19 surveillance. We enrolled 143,531 study participants (≥18 years) who contributed 10.6 million daily symptom reports between April 29, 2020 and February 10, 2021. Here, we include data from 19,161 self-reported PCR tests to create a symptom-based model to estimate the individual probability of symptomatic COVID-19, with an AUC of 0.78 (95% CI 0.74–0.83) in an external dataset. These individual probabilities are employed to estimate daily regional COVID-19 prevalence, which are in turn used together with current hospital data to predict next week COVID-19 hospital admissions. We show that this hospital prediction model demonstrates a lower median absolute percentage error (MdAPE: 25.9%) across the five most populated regions in Sweden during the first pandemic wave than a model based on case notifications (MdAPE: 30.3%). During the second wave, the error rates are similar. When we apply the same model to an English dataset, not including local COVID-19 test data, we observe MdAPEs of 22.3% and 19.0% during the first and second pandemic waves, respectively, highlighting the transferability of the prediction model.

Original language	English
Article number	2110
Journal	Nature Communications
Volume	13
Issue number	1
DOIs	https://doi.org/10.1038/s41467-022-29608-7
Publication status	Published - Dec 2022

Access to Document

10.1038/s41467-022-29608-7

Cite this

Kennedy, B., Fitipaldi, H., Hammar, U., Maziarz, M., Tsereteli, N., Oskolkov, N., Varotsis, G., Franks, C. A., Nguyen, D., Spiliopoulos, L., Adami, H. O., Björk, J., Engblom, S., Fall, K., Grimby-Ekman, A., Litton, J. E., Martinell, M., Oudin, A., Sjöström, T., ... Fall, T. (2022). App-based COVID-19 syndromic surveillance and prediction of hospital admissions in COVID Symptom Study Sweden. Nature Communications, 13(1), [2110]. https://doi.org/10.1038/s41467-022-29608-7

@article{cc6d213f710a4df2a615221cb4d64eef,

title = "App-based COVID-19 syndromic surveillance and prediction of hospital admissions in COVID Symptom Study Sweden",

abstract = "The app-based COVID Symptom Study was launched in Sweden in April 2020 to contribute to real-time COVID-19 surveillance. We enrolled 143,531 study participants (≥18 years) who contributed 10.6 million daily symptom reports between April 29, 2020 and February 10, 2021. Here, we include data from 19,161 self-reported PCR tests to create a symptom-based model to estimate the individual probability of symptomatic COVID-19, with an AUC of 0.78 (95% CI 0.74–0.83) in an external dataset. These individual probabilities are employed to estimate daily regional COVID-19 prevalence, which are in turn used together with current hospital data to predict next week COVID-19 hospital admissions. We show that this hospital prediction model demonstrates a lower median absolute percentage error (MdAPE: 25.9%) across the five most populated regions in Sweden during the first pandemic wave than a model based on case notifications (MdAPE: 30.3%). During the second wave, the error rates are similar. When we apply the same model to an English dataset, not including local COVID-19 test data, we observe MdAPEs of 22.3% and 19.0% during the first and second pandemic waves, respectively, highlighting the transferability of the prediction model.",

author = "Beatrice Kennedy and Hugo Fitipaldi and Ulf Hammar and Marlena Maziarz and Neli Tsereteli and Nikolay Oskolkov and Georgios Varotsis and Franks, {Camilla A.} and Diem Nguyen and Lampros Spiliopoulos and Adami, {Hans Olov} and Jonas Bj{\"o}rk and Stefan Engblom and Katja Fall and Anna Grimby-Ekman and Litton, {Jan Eric} and Mats Martinell and Anna Oudin and Torbj{\"o}rn Sj{\"o}str{\"o}m and Toomas Timpka and Sudre, {Carole H.} and Graham, {Mark S.} and {du Cadet}, {Julien Lavigne} and Chan, {Andrew T.} and Richard Davies and Sajaysurya Ganesh and Anna May and S{\'e}bastien Ourselin and Pujol, {Joan Capdevila} and Somesh Selvachandran and Jonathan Wolf and Spector, {Tim D.} and Steves, {Claire J.} and Gomez, {Maria F.} and Franks, {Paul W.} and Tove Fall",

note = "Funding Information: We thank all COVID Symptom Study Sweden participants whose participation, engagement, and feedback were essential to the study. We also thank ZOE Limited for excellent collaboration and development and maintenance of the app. We thank NOVUS for generously sharing their data with us. We thank the CSSS team for their dedication and engagement during the set up and running of the study. In particular, we thank Anna-Maria Dutius Andersson and Mattias Borell for administrative and technical support; Ulrika Blom-Nilsson, Pernilla Siming, and Riia Sustarsic for project management support; Sara Liedholm, Johanna Sandahl, Lars Uhlin, Caroline Run?us, and Katrin St?hl, along with Lund University and Uppsala University communication teams, for dissemination and outreach. Jacqueline Postma and Lund University and Uppsala University legal teams provided valuable advice regarding the legal aspects of this project; Erik Renstr?m and Stacey Ristinmaa S?rensen also provided valuable advice during the planning phase of the study. Koen Dekkers is thanked for valuable support with computational programming. The computations were enabled by resources in project sens2020559 provided by the Swedish National Infrastructure for Computing (SNIC) at UPPMAX, partially funded by the Swedish Research Council through grant agreement no. 2018-05973. Funding Information: We thank all COVID Symptom Study Sweden participants whose participation, engagement, and feedback were essential to the study. We also thank ZOE Limited for excellent collaboration and development and maintenance of the app. We thank NOVUS for generously sharing their data with us. We thank the CSSS team for their dedication and engagement during the set up and running of the study. In particular, we thank Anna-Maria Dutius Andersson and Mattias Borell for administrative and technical support; Ulrika Blom-Nilsson, Pernilla Siming, and Riia Sustarsic for project management support; Sara Liedholm, Johanna Sandahl, Lars Uhlin, Caroline Run{\'e}us, and Katrin St{\aa}hl, along with Lund University and Uppsala University communication teams, for dissemination and outreach. Jacqueline Postma and Lund University and Uppsala University legal teams provided valuable advice regarding the legal aspects of this project; Erik Renstr{\"o}m and Stacey Ristinmaa S{\"o}rensen also provided valuable advice during the planning phase of the study. Koen Dekkers is thanked for valuable support with computational programming. The computations were enabled by resources in project sens2020559 provided by the Swedish National Infrastructure for Computing (SNIC) at UPPMAX, partially funded by the Swedish Research Council through grant agreement no. 2018-05973. Funding Information: Swedish Heart-Lung Foundation (20190470, 20140776), Swedish Research Council (EXODIAB, 2009-1039; 2014-03529), European Commission (ERC-2015-CoG - 681742 NASCENT), eSSENCE@LU 8:8 (eSSENCE - The e-Science Collaboration), Swedish Foundation for Strategic Research (LUDC-IRC, 15-0067), Crafoord Foundation (20211011), and EUGLOHRIA (101017752) to M.G. and/or P.F., and European Research Council Starting Grant (ERC-2018-STG - 801965 GUTSY) to T.F. N.O. was financially supported by the Knut and Alice Wallenberg Foundation as part of the National Bioinformatics Infrastructure Sweden at SciLifeLab. A.T.C. was supported in this work through the Massachusetts Consortium on Pathogen Readiness (MassCPR). J.B. was financially supported by Swedish Research Council (2019-00198 and 2021-04665) and by Sweden{\textquoteright}s Innovation Agency (Vinnova; 2021-02648). ZOE Limited provided in-kind support for all aspects of building, running and supporting the app and service to all users worldwide. Support for this study for KCL researchers was provided by the National Institute for Health Research (NIHR)-funded Biomedical Research Centre based at Guy{\textquoteright}s and St Thomas{\textquoteright} (GSTT) NHS Foundation Trust. This work was also supported by the UK Research and Innovation London Medical Imaging & Artificial Intelligence Centre for Value-Based Healthcare. Investigators also received support from the Wellcome Trust, Medical Research Council (MRC), British Heart Foundation (BHF), Alzheimer{\textquoteright}s Society, European Union, NIHR, COVID-19 Driver Relief Fund (CDRF) and the NIHR-funded BioResource, Clinical Research Facility and Biomedical Research Centre (BRC) based at GSTT NHS Foundation Trust in partnership with KCL. ZOE Limited developed the app for data collection as a not-for-profit endeavour. None of the funding entities had any role in study design, data analysis, data interpretation, or the writing of this paper. Open access funding provided by Uppsala University. Publisher Copyright: {\textcopyright} 2022, The Author(s).",

year = "2022",

month = dec,

doi = "10.1038/s41467-022-29608-7",

language = "English",

volume = "13",

journal = "Nature Communications",

issn = "2041-1723",

publisher = "Nature Publishing Group",

number = "1",

}

Kennedy, B, Fitipaldi, H, Hammar, U, Maziarz, M, Tsereteli, N, Oskolkov, N, Varotsis, G, Franks, CA, Nguyen, D, Spiliopoulos, L, Adami, HO, Björk, J, Engblom, S, Fall, K, Grimby-Ekman, A, Litton, JE, Martinell, M, Oudin, A, Sjöström, T, Timpka, T, Sudre, CH, Graham, MS, du Cadet, JL, Chan, AT, Davies, R, Ganesh, S, May, A, Ourselin, S, Pujol, JC, Selvachandran, S, Wolf, J, Spector, TD, Steves, CJ, Gomez, MF, Franks, PW & Fall, T 2022, 'App-based COVID-19 syndromic surveillance and prediction of hospital admissions in COVID Symptom Study Sweden', Nature Communications, vol. 13, no. 1, 2110. https://doi.org/10.1038/s41467-022-29608-7

TY - JOUR

T1 - App-based COVID-19 syndromic surveillance and prediction of hospital admissions in COVID Symptom Study Sweden

AU - Kennedy, Beatrice

AU - Fitipaldi, Hugo

AU - Hammar, Ulf

AU - Maziarz, Marlena

AU - Tsereteli, Neli

AU - Oskolkov, Nikolay

AU - Varotsis, Georgios

AU - Franks, Camilla A.

AU - Nguyen, Diem

AU - Spiliopoulos, Lampros

AU - Adami, Hans Olov

AU - Björk, Jonas

AU - Engblom, Stefan

AU - Fall, Katja

AU - Grimby-Ekman, Anna

AU - Litton, Jan Eric

AU - Martinell, Mats

AU - Oudin, Anna

AU - Sjöström, Torbjörn

AU - Timpka, Toomas

AU - Sudre, Carole H.

AU - Graham, Mark S.

AU - du Cadet, Julien Lavigne

AU - Chan, Andrew T.

AU - Davies, Richard

AU - Ganesh, Sajaysurya

AU - May, Anna

AU - Ourselin, Sébastien

AU - Pujol, Joan Capdevila

AU - Selvachandran, Somesh

AU - Wolf, Jonathan

AU - Spector, Tim D.

AU - Steves, Claire J.

AU - Gomez, Maria F.

AU - Franks, Paul W.

AU - Fall, Tove

N1 - Funding Information: We thank all COVID Symptom Study Sweden participants whose participation, engagement, and feedback were essential to the study. We also thank ZOE Limited for excellent collaboration and development and maintenance of the app. We thank NOVUS for generously sharing their data with us. We thank the CSSS team for their dedication and engagement during the set up and running of the study. In particular, we thank Anna-Maria Dutius Andersson and Mattias Borell for administrative and technical support; Ulrika Blom-Nilsson, Pernilla Siming, and Riia Sustarsic for project management support; Sara Liedholm, Johanna Sandahl, Lars Uhlin, Caroline Run?us, and Katrin St?hl, along with Lund University and Uppsala University communication teams, for dissemination and outreach. Jacqueline Postma and Lund University and Uppsala University legal teams provided valuable advice regarding the legal aspects of this project; Erik Renstr?m and Stacey Ristinmaa S?rensen also provided valuable advice during the planning phase of the study. Koen Dekkers is thanked for valuable support with computational programming. The computations were enabled by resources in project sens2020559 provided by the Swedish National Infrastructure for Computing (SNIC) at UPPMAX, partially funded by the Swedish Research Council through grant agreement no. 2018-05973. Funding Information: We thank all COVID Symptom Study Sweden participants whose participation, engagement, and feedback were essential to the study. We also thank ZOE Limited for excellent collaboration and development and maintenance of the app. We thank NOVUS for generously sharing their data with us. We thank the CSSS team for their dedication and engagement during the set up and running of the study. In particular, we thank Anna-Maria Dutius Andersson and Mattias Borell for administrative and technical support; Ulrika Blom-Nilsson, Pernilla Siming, and Riia Sustarsic for project management support; Sara Liedholm, Johanna Sandahl, Lars Uhlin, Caroline Runéus, and Katrin Ståhl, along with Lund University and Uppsala University communication teams, for dissemination and outreach. Jacqueline Postma and Lund University and Uppsala University legal teams provided valuable advice regarding the legal aspects of this project; Erik Renström and Stacey Ristinmaa Sörensen also provided valuable advice during the planning phase of the study. Koen Dekkers is thanked for valuable support with computational programming. The computations were enabled by resources in project sens2020559 provided by the Swedish National Infrastructure for Computing (SNIC) at UPPMAX, partially funded by the Swedish Research Council through grant agreement no. 2018-05973. Funding Information: Swedish Heart-Lung Foundation (20190470, 20140776), Swedish Research Council (EXODIAB, 2009-1039; 2014-03529), European Commission (ERC-2015-CoG - 681742 NASCENT), eSSENCE@LU 8:8 (eSSENCE - The e-Science Collaboration), Swedish Foundation for Strategic Research (LUDC-IRC, 15-0067), Crafoord Foundation (20211011), and EUGLOHRIA (101017752) to M.G. and/or P.F., and European Research Council Starting Grant (ERC-2018-STG - 801965 GUTSY) to T.F. N.O. was financially supported by the Knut and Alice Wallenberg Foundation as part of the National Bioinformatics Infrastructure Sweden at SciLifeLab. A.T.C. was supported in this work through the Massachusetts Consortium on Pathogen Readiness (MassCPR). J.B. was financially supported by Swedish Research Council (2019-00198 and 2021-04665) and by Sweden’s Innovation Agency (Vinnova; 2021-02648). ZOE Limited provided in-kind support for all aspects of building, running and supporting the app and service to all users worldwide. Support for this study for KCL researchers was provided by the National Institute for Health Research (NIHR)-funded Biomedical Research Centre based at Guy’s and St Thomas’ (GSTT) NHS Foundation Trust. This work was also supported by the UK Research and Innovation London Medical Imaging & Artificial Intelligence Centre for Value-Based Healthcare. Investigators also received support from the Wellcome Trust, Medical Research Council (MRC), British Heart Foundation (BHF), Alzheimer’s Society, European Union, NIHR, COVID-19 Driver Relief Fund (CDRF) and the NIHR-funded BioResource, Clinical Research Facility and Biomedical Research Centre (BRC) based at GSTT NHS Foundation Trust in partnership with KCL. ZOE Limited developed the app for data collection as a not-for-profit endeavour. None of the funding entities had any role in study design, data analysis, data interpretation, or the writing of this paper. Open access funding provided by Uppsala University. Publisher Copyright: © 2022, The Author(s).

PY - 2022/12

Y1 - 2022/12

N2 - The app-based COVID Symptom Study was launched in Sweden in April 2020 to contribute to real-time COVID-19 surveillance. We enrolled 143,531 study participants (≥18 years) who contributed 10.6 million daily symptom reports between April 29, 2020 and February 10, 2021. Here, we include data from 19,161 self-reported PCR tests to create a symptom-based model to estimate the individual probability of symptomatic COVID-19, with an AUC of 0.78 (95% CI 0.74–0.83) in an external dataset. These individual probabilities are employed to estimate daily regional COVID-19 prevalence, which are in turn used together with current hospital data to predict next week COVID-19 hospital admissions. We show that this hospital prediction model demonstrates a lower median absolute percentage error (MdAPE: 25.9%) across the five most populated regions in Sweden during the first pandemic wave than a model based on case notifications (MdAPE: 30.3%). During the second wave, the error rates are similar. When we apply the same model to an English dataset, not including local COVID-19 test data, we observe MdAPEs of 22.3% and 19.0% during the first and second pandemic waves, respectively, highlighting the transferability of the prediction model.

AB - The app-based COVID Symptom Study was launched in Sweden in April 2020 to contribute to real-time COVID-19 surveillance. We enrolled 143,531 study participants (≥18 years) who contributed 10.6 million daily symptom reports between April 29, 2020 and February 10, 2021. Here, we include data from 19,161 self-reported PCR tests to create a symptom-based model to estimate the individual probability of symptomatic COVID-19, with an AUC of 0.78 (95% CI 0.74–0.83) in an external dataset. These individual probabilities are employed to estimate daily regional COVID-19 prevalence, which are in turn used together with current hospital data to predict next week COVID-19 hospital admissions. We show that this hospital prediction model demonstrates a lower median absolute percentage error (MdAPE: 25.9%) across the five most populated regions in Sweden during the first pandemic wave than a model based on case notifications (MdAPE: 30.3%). During the second wave, the error rates are similar. When we apply the same model to an English dataset, not including local COVID-19 test data, we observe MdAPEs of 22.3% and 19.0% during the first and second pandemic waves, respectively, highlighting the transferability of the prediction model.

UR - http://www.scopus.com/inward/record.url?scp=85128664211&partnerID=8YFLogxK

U2 - 10.1038/s41467-022-29608-7

DO - 10.1038/s41467-022-29608-7

M3 - Article

C2 - 35449172

AN - SCOPUS:85128664211

SN - 2041-1723

VL - 13

JO - Nature Communications

JF - Nature Communications

IS - 1

M1 - 2110

ER -

App-based COVID-19 syndromic surveillance and prediction of hospital admissions in COVID Symptom Study Sweden

Abstract

Access to Document

Other files and links

Fingerprint

Cite this