Manuscripts can change from their pre-printed or submitted version to their (peer-revied) published version of record. I identified 21 studies that analysed manuscript changes till December 2020. In this talk I will present a summary of those studies as well results of a study in which I analysed changes between 121 epidemiology preprints and their subsequent published versions. These 121 were published in 73 different journals (of median IF 4), with a median time of 204 days following the posting of the preprint. Half of preprints were posted before or at the same time as being submitted to journals. Out of the 121 pairs, 26% had differences in their titles, 5% in the number or order of authors, 44% in their section subtitles (e.g., IMRaD), 84% in the number of references, 26% in the number of tables, 23% in the number of figures, 44% in acknowledgment sections, 61% in conflicts of interest sections, 59% in limitations sections of discussion, and 18% in the first sentence they used to describe their (main) results in the discussion section. I hope to incite a discussion on which of these changes do we as scholarly community find important and wish we knew about, including on whether it is time for journals to declare changes manuscripts undergo due to the peer review.
KEYWORDS: peer review, preprints, versions of record