We recently questioned the utility of testing for proportional hazards in survival analysis. Here, we expand on why the proportional hazards assumption is both implausible and unnecessary in most medical studies, particularly in randomized trials. We conclude that using survival analysis methods that do not rely on proportional hazards is typically the preferred course of action.
Keywords: causal inference; hazard ratios; proportional hazards; survival analysis.
© The Author(s) 2025. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.