After publication of this article [1], the corresponding author notified PLOS One of an error in the p‑value calculation caused by a coding error, affecting Fig 4 and related text in the Experiment and result subsection of [1].

Following input from a member of the Editorial Board, the PLOS One Editors concluded that the conclusion drawn from applying the mathematical model to the large‑scale validation dataset, is incorrect. Specifically, after correcting the error, the revised analysis shows no causal relation between the Medicare Hospital Readmission Reduction Program (HRRP) and non-index readmissions. However, the evaluation of the model’s scalability remains sound, and the study’s methodological contribution is unaffected.

As the manuscript discusses the causal relationship extensively, and this forms a central aim of the study, the PLOS One Editors consider that a key conclusion is no longer upheld. To address this concern, the PLOS One Editors retract this article and have published an updated version of the article containing the corrected p‑value calculation, an updated Fig 4, and revised conclusions [2].

All authors agreed with the retraction.