Sense and sensitivity: can an inaccurate test be better than no test at all?

Journal of Medical Ethics (forthcoming)
  Copy   BIBTEX

Abstract

The UK government has put lateral flow antigen tests at the forefront of its strategy to scale-up testing in the coronavirus pandemic. However, evidence from a pilot trial using an LFAT to identify asymptomatic infections in the community suggested that the test missed over half of the positive cases in the tested population. This raises the question of whether it can be ethical to use an inaccurate test to guide public health measures. We begin by explicating different dimensions of test accuracy, and why they matter morally, before highlighting key data from the Liverpool pilot. We argue that the poor sensitivity of the LFAT in this pilot trial suggests that there are important limitations to what we can expect these tests to achieve. A test with low sensitivity will provide false-negative results, and in doing so generate the risk of false assurance and its attendant moral costs. However, we also suggest that the deployment of an insensitive but specific test could identify many asymptomatic carriers of the virus who are currently being missed under existing arrangements. Having outlined ways in which the costs of false reassurance could potentially be mitigated, we conclude that the use of an insensitive LFAT in mass testing may be ethical if it is used predominantly to identify positive cases, it is a cost-effective method of achieving that goal and if other public health tools can effectively prevent widespread false reassurance.

Links

PhilArchive



    Upload a copy of this work     Papers currently archived: 91,386

External links

Setup an account with your affiliations in order to access resources via your University's proxy server

Through your library

Similar books and articles

To test or not to test: A clinical dilemma.David B. Resnik - 1995 - Theoretical Medicine and Bioethics 16 (2).
The cartesian test for automatism.Gerald J. Erion - 2001 - Minds and Machines 11 (1):29-39.
Should we genetically test everyone for haemochromatosis?K. Allen & R. Williamson - 1999 - Journal of Medical Ethics 25 (2):209-214.
Turing's two tests for intelligence.Susan G. Sterrett - 1999 - Minds and Machines 10 (4):541-559.

Analytics

Added to PP
2021-04-06

Downloads
19 (#778,470)

6 months
3 (#992,474)

Historical graph of downloads
How can I increase my downloads?

Author Profiles

Jonathan Pugh
University of Oxford
Dominic Wilkinson
Oxford University

Citations of this work

Meta-surrogate decision making and artificial intelligence.Brian D. Earp - 2022 - Journal of Medical Ethics 48 (5):287-289.

Add more citations