App-based COVID-19 syndromic surveillance and prediction of hospital admissions in COVID Symptom Study Sweden
(2022) In Nature Communications 13(1).- 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... (More)
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.
(Less)
- author
- organization
-
- Genetic and Molecular Epidemiology (research group)
- eSSENCE: The e-Science Collaboration
- EXODIAB: Excellence of Diabetes Research in Sweden
- Department of Clinical Sciences, Lund
- Diabetic Complications (research group)
- Department of Clinical Sciences, Malmö
- Bioinformatics (research group)
- Department of Biology
- Division of Occupational and Environmental Medicine, Lund University
- publishing date
- 2022-04-21
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- COVID-19/epidemiology, Hospitals, Humans, Mobile Applications, Sentinel Surveillance, Sweden/epidemiology, Computational models, Epidemiology, Viral infection
- in
- Nature Communications
- volume
- 13
- issue
- 1
- article number
- 2110
- pages
- 12 pages
- publisher
- Nature Publishing Group
- external identifiers
-
- scopus:85128664211
- pmid:35449172
- ISSN
- 2041-1723
- DOI
- 10.1038/s41467-022-29608-7
- project
- Improved preparedness for future pandemics and other health crises through large-scale disease surveillance
- language
- English
- LU publication?
- yes
- additional info
- © 2022. The Author(s).
- id
- 8b05a33e-677d-4e91-a11e-91fb6fa60509
- date added to LUP
- 2022-05-03 09:45:56
- date last changed
- 2024-09-19 21:13:52
@article{8b05a33e-677d-4e91-a11e-91fb6fa60509, abstract = {{<p>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.</p>}}, author = {{Kennedy, Beatrice and Fitipaldi, Hugo and Hammar, Ulf and Maziarz, Marlena and Tsereteli, Neli and Oskolkov, Nikolay and Varotsis, Georgios and Franks, Camilla A and Nguyen, Diem and Spiliopoulos, Lampros and Adami, Hans-Olov and Björk, Jonas and Engblom, Stefan and Fall, Katja and Grimby-Ekman, Anna and Litton, Jan-Eric and Martinell, Mats and Oudin, Anna and Sjöström, Torbjörn and Timpka, Toomas and Sudre, Carole H and Graham, Mark S and du Cadet, Julien Lavigne and Chan, Andrew T and Davies, Richard and Ganesh, Sajaysurya and May, Anna and Ourselin, Sébastien and Pujol, Joan Capdevila and Selvachandran, Somesh and Wolf, Jonathan and Spector, Tim D and Steves, Claire J and Gomez, Maria F and Franks, Paul W and Fall, Tove}}, issn = {{2041-1723}}, keywords = {{COVID-19/epidemiology; Hospitals; Humans; Mobile Applications; Sentinel Surveillance; Sweden/epidemiology; Computational models; Epidemiology; Viral infection}}, language = {{eng}}, month = {{04}}, number = {{1}}, publisher = {{Nature Publishing Group}}, series = {{Nature Communications}}, title = {{App-based COVID-19 syndromic surveillance and prediction of hospital admissions in COVID Symptom Study Sweden}}, url = {{http://dx.doi.org/10.1038/s41467-022-29608-7}}, doi = {{10.1038/s41467-022-29608-7}}, volume = {{13}}, year = {{2022}}, }