Software engineering standards for epidemiological models
- PDF / 562,490 Bytes
- 24 Pages / 439.37 x 666.142 pts Page_size
- 35 Downloads / 203 Views
Software engineering standards for epidemiological models Jack K. Horner1 · John F. Symons1
Received: 16 June 2020 / Accepted: 18 October 2020 © Springer Nature Switzerland AG 2020
Abstract There are many tangled normative and technical questions involved in evaluating the quality of software used in epidemiological simulations. In this paper we answer some of these questions and offer practical guidance to practitioners, funders, scientific journals, and consumers of epidemiological research. The heart of our paper is a case study of the Imperial College London (ICL) covid-19 simulator, set in the context of recent work in epistemology of simulation and philosophy of epidemiology. Keywords COVID-19 · Public-health policy · Simulation · Software engineering
1 Introduction There are many tangled normative and technical questions involved in evaluating the quality of software used in epidemiological simulations. In this paper we answer some of these questions and offer practical guidance to practitioners, funders, scientific journals, and consumers of epidemiological research. The heart of our paper is a case study in which we provide an analysis of the Imperial College London (ICL) covid-19 simulator. This simulator has been extensively used by the United Kingdom to help formulate public-health policy; it has been used to a somewhat lesser extent in public-health policy decision-making in the United States. Developed primarily to predict the effects of public-health interventions such as shutdowns,
* Jack K. Horner [email protected] John F. Symons [email protected] 1
Department of Philosophy, University of Kansas, Lawrence, KS, USA
13
Vol.:(0123456789)
54
Page 2 of 24
J. K. Horner, J. F. Symons
quarantines, social-distancing, and the administration of vaccines, it can be viewed as a complicated data-driven variant of the “susceptible-infected-recovering” (SIR) family of epidemiological models, first described in Kermack and McKendrick (1927). Our case study, combined with reflection on the state of the art in the philosophy of epidemiology and the ethics of engineering, serves as the basis for our recommendations for future epidemiological modeling projects. We contend that epidemiological simulators should be engineered and evaluated according to a set of public norms. We take as our model for the kinds of norms that we regard as appropriate, the framework of safety–critical standards developed by consensus of the software engineering community for applications such as automotive and aircraft control systems. To achieve that goal, the development and use of epidemiological simulators must have high levels of transparency, explainability, and reproducibility for stakeholders. Furthermore, we recommend that such standards be mandated by funding agencies for epidemiological contexts that have direct and significant public policy implications. The structure of our argument is straightforward. In Sect. 2 we explain some recent work on the role computation in the philosophy of epidemiology.
Data Loading...
