AI, machine learning, and data science are three of the biggest weapons we have in our arsenal to cope with the inevitable pandemics we are going to continue to experience. This means the multidisciplinary study of the COVID-19 pandemic and its wide-ranging impact has become an urgent endeavor worldwide. In an attempt to further and deepen global understanding of the crisis, the Harvard Data Science Review (an open access platform of the Harvard Data Science Initiative) is publishing a special issue examining the novel coronavirus and its impact through the lens of data science.
The issue covers a wide range of topics addressing the societal, epidemiological, political, and educational issues that have rapidly emerged from the SARS-CoV2 pandemic. Articles include:
Five leading biostatisticians and epidemiologists debate the probable scope and duration of the pandemic, the kinds of medical responses that we need, and some of the impacts they foresee on the U.S. and on the world. They also discuss the pandemic’s likely effect on higher education.
The reasons why we are currently “socially distancing” are based on an understanding of exponential growth and the idea of “flattening the curve.” The authors present and discuss a pair of survey experiments that explore the public’s statistical literacy by examining its ability to calculate and understand exponential growth. These findings may be used to help better ground effective communication strategies aimed at the general public.
In the midst of the COVID-19 pandemic, how should regulatory agencies adapt their normally lengthy clinical trial and approval process to address the urgency of finding treatments and saving lives? The authors propose a Bayesian adaptive patient-centered framework to optimize the clinical trial development path for anti-infective therapies and vaccines. Their research provides a rational, systematic, transparent, repeatable, and practical framework for regulators, policymakers, and clinical researchers to evaluate the efficacy of anti-infective therapeutics during the course of any epidemic outbreak when the cost of false negatives far outweighs the cost of false positives.
The economic value of a drug or medical device development program is typically computed by assessing the program’s cumulative revenues if successful, and companies rely on this data to make business decisions about which programs to pursue and how to fund them.
In this article, the authors provide estimates of clinical trial outcomes for vaccines and other anti-infective therapeutics using 43,414 unique triplets of clinical trial, drug, and disease between January 1, 2000, and January 7, 2020, yielding 2,544 vaccine programs and 6,829 non-vaccine programs targeting infectious diseases–the largest dataset of its kind. As governments around the world begin to formulate a more systematic strategy for dealing with pandemics beyond COVID-19, these estimates can be used by policymakers to identify areas most likely to be undeserved by private-sector engagement and in need of public-sector support.
Innovations in data science and artificial intelligence (AI) have a central role to play in supporting global efforts to combat COVID-19 and address a broad range of biomedical, epidemiological, and socio-economic challenges. However, this wide-reaching scientific capacity also raises ethical challenges.
The authors present a practice-based path to responsible AI design and discovery centered on open, accountable, equitable, and democratically governed processes and products. When taken from the start, these steps will not only enhance the capacity of innovators to tackle Covid-19 responsibly, they will help to set the data science and AI community down a path that is both better prepared to cope with future pandemics and better equipped to support a more humane, rational, and just society of tomorrow.
The special issue will be published on a rolling/continuous basis with new articles appearing weekly through the beginning of July 2020.
Source: The MIT Press