TY - JOUR T1 - Machine learning models aimed at identifying risk factors for reducing morbidity and mortality still need to consider confounding related to calendar time variations JF - medRxiv DO - 10.1101/2022.05.24.22275482 SP - 2022.05.24.22275482 AU - Andreas Rieckmann AU - Tri-Long Nguyen AU - Piotr Dworzynski AU - Ane Bærent Fisker AU - Naja Hulvej Rod AU - Claus Thorn Ekstrøm Y1 - 2022/01/01 UR - http://medrxiv.org/content/early/2022/05/25/2022.05.24.22275482.abstract N2 - Machine learning models applied to health data may help health professionals to prioritize resources by identifying risk factors that may reduce morbidity and mortality. However, many novel machine learning papers on this topic neither account for nor discuss biases due to calendar time variations. Often, efforts to account for calendar time (among other confounders) are necessary since patterns in health data – especially in low- and middle-income countries – may be influenced by calendar time variations such as temporal changes in risk factors and changes in the disease and mortality distributions over time (epidemiological transitions), seasonal changes in risk factors and disease and mortality distributions, as well as co-occurring artefacts in data due to changes in surveillance and diagnostics. Based on simulations, real-life data from Guinea-Bissau, and examples drawn from recent studies, we discuss how including calendar time variations in machine learning models is beneficial for generating more relevant and actionable results. In this brief report, we stress that explicitly handling temporal structures in machine learning models still remains to be considered (like in general epidemiological studies) to prevent resources from being misdirected to ineffective interventions.Competing Interest StatementThe authors have declared no competing interest.Funding StatementAR was supported by an international postdoc grant by the Independent Research Fund Denmark (9034-00006B). PD was supported by a research grant from the Danish Diabetes Academy funded by the Novo Nordisk Foundation.Author DeclarationsI confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.YesThe details of the IRB/oversight body that provided approval or exemption for the research described are given below:The BHP HDSS surveillance was initiated in 1990 at the request of the Ministry of Health. Surveyed women provided oral consent at the time of registration. Protocols for concurrent trials nested in the HDSS and describing the data collection have been approved by the Ministry of Health (Nucleo de Coordencao das Pesquisas do Ministerio da Saude: NPC no. 12/2007, NPC no. 02/2008), National Ethics Committee in Guinea-Bissau (Comite Nacional de Etica na Saude: no 34/CNES/2010, 08/CNES/2011) and received consultative approval from the Central Ethical Committee in Denmark (2006-7041-99; 1103988)I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.YesI understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).YesI have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.YesThe simulated dataset can be generated using the R script in the supplementary material. Request for data access is referred to Bandim Health Project, bandim@bandim.org. ER -