On the ethics of algorithmic decision-making in healthcare

Thomas Grote, Philipp Berens

Journal of Medical Ethics · 2019 · 513 citations · 25 references

DOIFull text

Open access

Concepts

TL;DR

Recent high‑profile studies report that machine‑learning algorithms outperform clinicians in diagnosis and treatment, sparking interest in deploying them to improve healthcare decision‑making. The paper argues that deploying machine‑learning in healthcare involves epistemic and normative trade‑offs, highlighting pitfalls such as paternalism, moral responsibility, fairness, and shifts in evidentiary norms, and calls for further ethical reflection. The authors draw on social epistemology and moral responsibility literature to contend that uncertainty undermines clinicians’ epistemic authority and that machine‑learning use raises paternalism, moral responsibility, and fairness concerns. They find that while machine‑learning can improve diagnostic accuracy, it reduces opacity, making it harder to assess diagnosis reliability.

Abstract

In recent years, a plethora of high-profile scientific publications has been reporting about machine learning algorithms outperforming clinicians in medical diagnosis or treatment recommendations. This has spiked interest in deploying relevant algorithms with the aim of enhancing decision-making in healthcare. In this paper, we argue that instead of straightforwardly enhancing the decision-making capabilities of clinicians and healthcare institutions, deploying machines learning algorithms entails trade-offs at the epistemic and the normative level. Whereas involving machine learning might improve the accuracy of medical diagnosis, it comes at the expense of opacity when trying to assess the reliability of given diagnosis. Drawing on literature in social epistemology and moral responsibility, we argue that the uncertainty in question potentially undermines the epistemic authority of clinicians. Furthermore, we elucidate potential pitfalls of involving machine learning in healthcare with respect to paternalism, moral responsibility and fairness. At last, we discuss how the deployment of machine learning algorithms might shift the evidentiary norms of medical diagnosis. In this regard, we hope to lay the grounds for further ethical reflection of the opportunities and pitfalls of machine learning for enhancing decision-making in healthcare.

References

25