The session “Modelling Futures, Modelling Pasts” (Chair: Thomas Macho, Humboldt University, Berlin) on Friday, Feb 8, 4:45-6:15 pm, at this week’s Science Futures Conference in Zurich offers talks on epidemiology, modelling, numerical experiments and simulation studies:
Susanne Bauer, University of Copenhagen: Epidemiological Studies as ‘Future Machines’: Modelling Population Health and Predicting Individual Risk
Mikaela Sundberg, University of Stockholm: Exploring Simulation Models: Numerical Experiments and Virtual Worlds
Erika Mattila, London School of Economics: Predicting the Future by Modelling the Past
For the abstract of my talk, see here:
“Epidemiological studies as ‘future machines’: Modelling population health and predicting individual risk”
Susanne Bauer, Medical Museion, University of Copenhagen
Much of contemporary epidemiology –‘the science of the distribution of health and disease in populations’ – is about prognostic statistics, about calculating the past and predicting the future. In epidemiological studies, facts come as probabilistic risks – derived from the past and projected into the future. From evidence-based medicine to evidence-based policy, probabilistic reasoning increasingly constitutes its own culture of decision-making. In epidemiological studies, probabilistic measures of effect are calculated from the collective experience of defined populations, i.e. health outcomes for follow-up periods over decades. Risk estimates are derived and then spelled out for subgroups and diversified according to profiles and applied to individuals. Establishing and maintaining large scale databases on health and disease as future opportunities for research is key to the production of epidemiological knowledge. This paper looks at the complex ‘assemblages’ of information networks within an epidemiological study in Denmark and related data gathering and mining practices. Drawing on STS approaches, I will analyse study designs as methodological ‘machines’ and databases as ‘emergent structures’. Following biostatistical modelling further towards predictive knowledge and evidence-based prevention, I will explore how futures are created via mining the past. Optimising health, reducing risk and adjusting profiles to projected futures have become ubiquitous practices that capitalise on uncertainty and enact biomedical progress as an economy of promises.