Concerns about data integrity across 263 papers by one author

Abstract

Comprehensive investigation of published work by authors suspected of academic misconduct can reveal further concerns. We aimed to test for data integrity concerns in papers published by an author with eight retracted articles. We investigated the integrity of all papers reporting on prospective clinical studies by this author. We assessed the feasibility of study methods, baseline characteristics, and outcomes. We plotted the author's clinical research activity over time. We conducted pairwise comparisons of text, tables, and figures to identify duplicate publications, and checked for consistency between conference abstracts, interim analyses, trial registrations, and final papers. Where indicated, we recalculated p-values from the reported summary statistics. We identified 263 papers claiming to have enrolled 74,667 participants between January 2009 and July 2022, 190 (72%) of which reported on studies that recruited from the Assiut Women's Health Hospital in Assiut, Egypt. The number of active studies per month was greatest between 2016 and 2019, with 88 ongoing studies in May 2017. We found evidence of data integrity concerns in 130 (49%) papers, 43 (33%) of which contained concerns sufficient to suggest that they could not be based on data reliably collected from human participants. Our investigation finds evidence of widespread integrity concerns in the collected work of one author. We recommend that the involved journals collaborate in a formal investigation.