[Skip to Navigation]
Comment & Response
April 8, 2021

Survival Analysis vs Longitudinal Modeling With Multiple Imputation—a False Dichotomy—Reply

Author Affiliations
  • 1The Royal Marsden National Health Service Foundation Trust, London, United Kingdom
  • 2National Institute for Health and Research Biomedical Centre, Moorfields Eye Hospital National Health Service Foundation Trust, London, United Kingdom
  • 3Gloucestershire Retinal Research Group, Gloucestershire Hospitals NHS Foundation Trust, Gloucestershire, United Kingdom
JAMA Ophthalmol. 2021;139(5):588-589. doi:10.1001/jamaophthalmol.2021.0518

In Reply We thank VanderBeek et al and the journal for giving us the opportunity to respond to this letter and hope that our reply offers some reassurance. VanderBeek et al point out that missing data may present issues to the unwary researcher, and we are pleased that our article1 has drawn attention to this matter. We completely agree that there is great value in the use of multiple imputation in observational studies with missing data and are somewhat perplexed by the suggestion that our article indicated otherwise.

Add or change institution