Forecasting severe respiratory disease hospitalizations using machine learning algorithms.
Artificial intelligence
Flu prediction
Forecasting
Forecasting healthcare burden
Influenza-like illness
Machine learning
Probabilistic forecast
Seasonal epidemic
Severe respiratory diseases
Journal
BMC medical informatics and decision making
ISSN: 1472-6947
Titre abrégé: BMC Med Inform Decis Mak
Pays: England
ID NLM: 101088682
Informations de publication
Date de publication:
09 Oct 2024
09 Oct 2024
Historique:
received:
15
06
2024
accepted:
30
09
2024
medline:
9
10
2024
pubmed:
9
10
2024
entrez:
8
10
2024
Statut:
epublish
Résumé
Forecasting models predicting trends in hospitalization rates have the potential to inform hospital management during seasonal epidemics of respiratory diseases and the associated surges caused by acute hospital admissions. Hospital bed requirements for elective surgery could be better planned if it were possible to foresee upcoming peaks in severe respiratory illness admissions. Forecasting models can also guide the use of intervention strategies to decrease the spread of respiratory pathogens and thus prevent local health system overload. In this study, we explore the capability of forecasting models to predict the number of hospital admissions in Auckland, New Zealand, within a three-week time horizon. Furthermore, we evaluate probabilistic forecasts and the impact on model performance when integrating laboratory data describing the circulation of respiratory viruses. The dataset used for this exploration results from active hospital surveillance, in which the World Health Organization Severe Acute Respiratory Infection (SARI) case definition was consistently used. This research nurse-led surveillance has been implemented in two public hospitals in Auckland and provides a systematic laboratory testing of SARI patients for nine respiratory viruses, including influenza, respiratory syncytial virus, and rhinovirus. The forecasting strategies used comprise automatic machine learning, one of the most recent generative pre-trained transformers, and established artificial neural network algorithms capable of univariate and multivariate forecasting. We found that machine learning models compute more accurate forecasts in comparison to naïve seasonal models. Furthermore, we analyzed the impact of reducing the temporal resolution of forecasts, which decreased the model error of point forecasts and made probabilistic forecasting more reliable. An additional analysis that used the laboratory data revealed strong season-to-season variations in the incidence of respiratory viruses and how this correlates with total hospitalization cases. These variations could explain why it was not possible to improve forecasts by integrating this data. Active SARI surveillance and consistent data collection over time enable these data to be used to predict hospital bed utilization. These findings show the potential of machine learning as support for informing systems for proactive hospital management.
Sections du résumé
BACKGROUND
BACKGROUND
Forecasting models predicting trends in hospitalization rates have the potential to inform hospital management during seasonal epidemics of respiratory diseases and the associated surges caused by acute hospital admissions. Hospital bed requirements for elective surgery could be better planned if it were possible to foresee upcoming peaks in severe respiratory illness admissions. Forecasting models can also guide the use of intervention strategies to decrease the spread of respiratory pathogens and thus prevent local health system overload. In this study, we explore the capability of forecasting models to predict the number of hospital admissions in Auckland, New Zealand, within a three-week time horizon. Furthermore, we evaluate probabilistic forecasts and the impact on model performance when integrating laboratory data describing the circulation of respiratory viruses.
METHODS
METHODS
The dataset used for this exploration results from active hospital surveillance, in which the World Health Organization Severe Acute Respiratory Infection (SARI) case definition was consistently used. This research nurse-led surveillance has been implemented in two public hospitals in Auckland and provides a systematic laboratory testing of SARI patients for nine respiratory viruses, including influenza, respiratory syncytial virus, and rhinovirus. The forecasting strategies used comprise automatic machine learning, one of the most recent generative pre-trained transformers, and established artificial neural network algorithms capable of univariate and multivariate forecasting.
RESULTS
RESULTS
We found that machine learning models compute more accurate forecasts in comparison to naïve seasonal models. Furthermore, we analyzed the impact of reducing the temporal resolution of forecasts, which decreased the model error of point forecasts and made probabilistic forecasting more reliable. An additional analysis that used the laboratory data revealed strong season-to-season variations in the incidence of respiratory viruses and how this correlates with total hospitalization cases. These variations could explain why it was not possible to improve forecasts by integrating this data.
CONCLUSIONS
CONCLUSIONS
Active SARI surveillance and consistent data collection over time enable these data to be used to predict hospital bed utilization. These findings show the potential of machine learning as support for informing systems for proactive hospital management.
Identifiants
pubmed: 39379946
doi: 10.1186/s12911-024-02702-0
pii: 10.1186/s12911-024-02702-0
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
293Informations de copyright
© 2024. The Author(s).
Références
Lane CJ, Bhatnagar M, Lutrick K, Maves RC, Weiner D, Olvera DR, et al. ICU resource limitations during peak seasonal influenza: Results of a 2018 national feasibility study. Critical Care Explorations. 2022;4(1): e0606.
doi: 10.1097/CCE.0000000000000606
pmcid: 8735785
Abubakar A, Malik M, Pebody RG, Elkholy AA, Khan W, Bellos A, et al. Burden of acute respiratory disease of epidemic and pandemic potential in the WHO Eastern Mediterranean Region: a literature review. EMHJ-Eastern Mediterranean Health J. 2016;22(7):509–22.
doi: 10.26719/2016.22.7.509
Farquharson KA, Anthony D, Menzies R, Homaira N. Burden of respiratory syncytial virus disease across the lifespan in Australia and New Zealand: a scoping review. Public Health. 2024;226:8–16.
doi: 10.1016/j.puhe.2023.10.031
Macias AE, McElhaney JE, Chaves SS, Nealon J, Nunes MC, Samson SI, et al. The disease burden of influenza beyond respiratory illness. Vaccine. 2021;39:A6-14.
doi: 10.1016/j.vaccine.2020.09.048
Newall AT, Scuffham PA. Influenza-related disease: the cost to the Australian healthcare system. Vaccine. 2008;26(52):6818–23.
doi: 10.1016/j.vaccine.2008.09.086
Bontempi G, Ben Taieb S, Le Borgne YA, Machine learning strategies for time series forecasting. Business Intelligence: Second European Summer School, eBISS 2012, Brussels, Belgium, July 15–21, 2012, Tutorial Lectures 2. 2013;62–77.6. Bontempi G, Ben Taieb S, Le Borgne YA. Machine learning strategies for time series forecasting. Business Intelligence: Second European Summer School, eBISS,. Brussels, Belgium, July 15–21, 2012. Tutorial Lectures. 2012;2(2013):62–77.
Montero P, Vilar JA. TSclust: An R package for time series clustering. J Stat Softw. 2015;62:1–43.
Löning M, Bagnall A, Ganesh S, Kazakov V, Lines J, Király FJ. sktime: A unified interface for machine learning with time series. 33rd Conference on Neural Information Processing Systems. 2019.
Godahewa R, Bergmeir C, Webb GI, Hyndman RJ, Montero-Manso P. Monash time series forecasting archive. 35th Conference on Neural Information Processing Systems. 2021.
Alexandrov A, Benidis K, Bohlke-Schneider M, Flunkert V, Gasthaus J, Januschowski T, et al. Gluonts: probabilistic and neural time series modeling in python. J Mach Learn Res. 2020;21(116):1–6.
Shchur O, Turkmen AC, Erickson N, Shen H, Shirkov A, Hu T, et al. AutoGluon–TimeSeries: AutoML for probabilistic time series forecasting. International Conference on Automated Machine Learning. PMLR; 2023. p. 9/1–21.
Qi H, Xiao S, Shi R, Ward MP, Chen Y, Tu W, et al. COVID-19 transmission in Mainland China is associated with temperature and humidity: a time-series analysis. Sci Total Environ. 2020;728: 138778.
doi: 10.1016/j.scitotenv.2020.138778
pmcid: 7167225
Shrivastav LK, Jha SK. A gradient boosting machine learning approach in modeling the impact of temperature and humidity on the transmission rate of COVID-19 in India. Appl Intell. 2021;51:2727–39.
doi: 10.1007/s10489-020-01997-6
Iloanusi O, Ross A. Leveraging weather data for forecasting cases-to-mortality rates due to COVID-19. Chaos, Solitons Fractals. 2021;152: 111340.
doi: 10.1016/j.chaos.2021.111340
Da Silva RG, Ribeiro MHDM, Mariani VC, dos Santos CL. Forecasting Brazilian and American COVID-19 cases based on artificial intelligence coupled with climatic exogenous variables. Chaos Solitons Fractals. 2020;139:110027.
doi: 10.1016/j.chaos.2020.110027
pmcid: 7324930
Yang S, Santillana M, Brownstein JS, Gray J, Richardson S, Kou SC. Using electronic health records and Internet search information for accurate influenza forecasting. BMC Infect Dis. 2017;17:1–9.
doi: 10.1186/s12879-017-2424-7
Yang S, Santillana M, Kou SC. Accurate estimation of influenza epidemics using Google search data via ARGO. Proc Natl Acad Sci. 2015;112(47):14473–8.
doi: 10.1073/pnas.1515373112
pmcid: 4664296
Moss R, Zarebski A, Dawson P, McCaw JM. Forecasting influenza outbreak dynamics in Melbourne from Internet search query surveillance data. Influenza Other Respir Viruses. 2016;10(4):314–23.
doi: 10.1111/irv.12376
pmcid: 4910172
Aiken EL, Nguyen AT, Viboud C, Santillana M. Toward the use of neural networks for influenza prediction at multiple spatial resolutions. Sci Adv. 2021;7(25):eabb1237.
doi: 10.1126/sciadv.abb1237
pmcid: 8208709
Venna SR, Tavanaei A, Gottumukkala RN, Raghavan VV, Maida AS, Nichols S. A novel data-driven model for real-time influenza forecasting. Ieee Access. 2018;7:7691–701.
doi: 10.1109/ACCESS.2018.2888585
Dugas AF, Jalalpour M, Gel Y, Levin S, Torcaso F, Igusa T, et al. Influenza forecasting with Google flu trends. PLoS ONE. 2013;8(2): e56176.
doi: 10.1371/journal.pone.0056176
pmcid: 3572967
Chen TH, Chen YC, Chen JL, Chang FC. Flu trend prediction based on massive data analysis. 3rd International Conference on Cloud Computing and Big Data Analysis (ICCCBDA). IEEE; 2018. p. 304–8.
Davidson MW, Haim DA, Radin JM. Using networks to combine “big data” and traditional surveillance to improve influenza predictions. Sci Rep. 2015;5(1):8154.
doi: 10.1038/srep08154
pmcid: 5389136
Huang QS, Turner N, Baker MG, Williamson DA, Wong C, Webby R, et al. Southern hemisphere influenza and vaccine effectiveness research and surveillance. Influenza Other Respir Viruses. 2015;9(4):179–90.
doi: 10.1111/irv.12315
pmcid: 4474494
Huang QS, Baker M, McArthur C, Roberts S, Williamson D, Grant C, et al. Implementing hospital-based surveillance for severe acute respiratory infections caused by influenza and other respiratory pathogens in New Zealand. Western Pacific surveillance and response journal: WPSAR. 2014;5(2):23.
doi: 10.5365/wpsar.2014.5.1.004
pmcid: 4113656
Pei S, Shaman J. Aggregating forecasts of multiple respiratory pathogens supports more accurate forecasting of influenza-like illness. PLoS Comput Biol. 2020;16(10): e1008301.
doi: 10.1371/journal.pcbi.1008301
pmcid: 7608986
Jenkins GM. Autoregressive-Integrated Moving Average (ARIMA) Models. Encyclop Statist Sci. 2004;1:1.
Hyndman RJ, Koehler AB, Snyder RD, Grose S. A state space framework for automatic forecasting using exponential smoothing methods. Int J Forecast. 2002;18(3):439–54.
doi: 10.1016/S0169-2070(01)00110-8
Assimakopoulos V, Nikolopoulos K. The theta model: a decomposition approach to forecasting. Int J Forecast. 2000;16(4):521–30.
doi: 10.1016/S0169-2070(00)00066-2
Syntetos AA, Boylan JE. On the bias of intermittent demand estimates. Int J Prod Econ. 2001;71(1–3):457–66.
doi: 10.1016/S0925-5273(00)00143-2
Wong H, Ip WC, Zhang R, Xia J. Non-parametric time series models for hydrological forecasting. J Hydrol. 2007;332(3–4):337–47.
doi: 10.1016/j.jhydrol.2006.07.013
Nie Y, Nguyen NH, Sinthong P, Kalagnanam J. A time series is worth 64 words: Long-term forecasting with transformers. International Conference on Learning Representations. 2023.
Salinas D, Flunkert V, Gasthaus J, Januschowski T. DeepAR: Probabilistic forecasting with autoregressive recurrent networks. Int J Forecast. 2020;36(3):1181–91.
doi: 10.1016/j.ijforecast.2019.07.001
Lim B, Arık SÖ, Loeff N, Pfister T. Temporal fusion transformers for interpretable multi-horizon time series forecasting. Int J Forecast. 2021;37(4):1748–64.
doi: 10.1016/j.ijforecast.2021.03.012
Barman R, Deshpande S, Agarwal S, Inamdar U, Devare M, Patil A. Transfer learning for small dataset. In: Proceedings of the National Conference on Machine Learning. ResearchGate Berlin, Germany; 2019.
Wang Y, Nazir S, Shafiq M. An overview on analyzing deep learning and transfer learning approaches for health monitoring. Comput Math Methods Med. 2021;2021:1–10.
doi: 10.1155/2021/4530180
Kora P, Ooi CP, Faust O, Raghavendra U, Gudigar A, Chan WY, et al. Transfer learning techniques for medical image analysis: A review. Biocybernet Biomed Eng. 2022;42(1):79–107.
doi: 10.1016/j.bbe.2021.11.004
Garza A, Mergenthaler-Canseco M. TimeGPT-1. arXiv preprint arXiv:231003589. 2023.
Das A, Kong W, Sen R, Zhou Y. A decoder-only foundation model for time-series forecasting. In: Salakhutdinov R, Kolter Z, Heller K, Weller A, Oliver N, Scarlett J, Berkenkamp F, editors. Proceedings of the 41st International Conference on Machine Learning. 2023. p. 10148–10167. https://proceedings.mlr.press/v235/das24c.html .
Ansari AF, Stella L, Turkmen C, Zhang X, Mercado P, Shen H, et al. Chronos: Learning the language of time series. arXiv preprint arXiv:240307815. 2024;
Chung Y, Neiswanger W, Char I, Schneider J. Beyond pinball loss: Quantile methods for calibrated uncertainty quantification. Adv Neural Inf Process Syst. 2021;34:10971–84.
Herzen J, Lässig F, Piazzetta SG, Neuer T, Tafti L, Raille G, et al. Darts: User-friendly modern machine learning for time series. J Mach Learn Res. 2022;23(124):1–6.
Roster K, Connaughton C, Rodrigues FA. Forecasting new diseases in low-data settings using transfer learning. Chaos, Solitons Fractals. 2022;161: 112306.
doi: 10.1016/j.chaos.2022.112306
Salehi S, Kavgic M, Bonakdari H, Begnoche L. Comparative study of univariate and multivariate strategy for short-term forecasting of heat demand density: Exploring single and hybrid deep learning models. Energy and AI. 2024;16: 100343.
doi: 10.1016/j.egyai.2024.100343
Etu EE, Monplaisir L, Masoud S, Arslanturk S, Emakhu J, Tenebe I, et al. A comparison of univariate and multivariate forecasting models predicting emergency department patient arrivals during the COVID-19 pandemic. In: Healthcare. MDPI; 2022. p. 1120.
Chowell G, Viboud C. Pandemic influenza and socioeconomic disparities: Lessons from 1918 Chicago. Proc Natl Acad Sci. 2016;113(48):13557–9.
doi: 10.1073/pnas.1616537113
pmcid: 5137725
Mamelund SE, Shelley-Egan C, Rogeberg O. The association between socioeconomic status and pandemic influenza: Systematic review and meta-analysis. PLoS ONE. 2021;16(9): e0244346.
doi: 10.1371/journal.pone.0244346
pmcid: 8423272