Event

Abstract

It has long been assumed that most peer-reviewed publications in health and medical research, even those of demonstrably poor quality, low integrity and high risk of bias, were at least based on real data. Little attention has been paid to the trustworthiness of reports from randomised clinical trials (RCTs) submitted for peer review. That is, journal editors and peer reviewers have not, habitually, considered the possibility that the study never took place and that the results of the supposed data analysis were fabricated. It is estimated that as much as 30% of the data presented in published RCTs are not genuine. This has led to a distorted evidence base containing fabricated and otherwise corrupted data that end up in meta-analyses and, ultimately, the guidelines that inform clinical practice in Australia and around the world. In this seminar, Ben will tell of his personal odyssey in research integrity, summarize the problem in 2025 and discuss solutions. Lyle will present the statistical toolbox that the team has developed to assess post-publication concerns with RCTs, and talk about emerging methods for assessing the often remarkable similarly between tables of summary data published in unrelated papers.

Ben Mol is Professor of Obstetrics and Gynaecology at Monash University. During his time in Australia, Ben has worked on clinical trial capacity in women’s health at Monash. He developed extensive relationships with Asian universities, resulting in large RCTs. Ben is also involved in many Individual Participant Data Meta-Analysis (IPD-MA), in which data from RCT’s conducted worldwide in Obstetrics and Reproductive Medicine data are brought together. Recently, Ben has worked also on systems to detect data-fabrication in RCTs. He has published more than 1600 papers and supervised 130 PhD students. His professional adage is: A day without randomisation is a day without progress.

Lyle Gurrin is Professor of Biostatistics at the University of Melbourne. He has run his own epidemiological studies looking at why some people with a genetic or immunological predisposition to disease never experience symptoms, such as HFE-associated hereditary haemochromatosis (iron overload) and childhood food allergy. Before moving to Melbourne in 2001, Lyle worked at medical research institutes in Perth in the 1990’s researching the developmental origins of health and disease. Recently, Lyle has been working with colleagues at the Monash Medical Centre and a network of data sleuths and integrity specialists around Australia and in the global north to develop statistical methods that can be used to detect data fabrication and fraud in peer-reviewed publications of RCT’s.

This session will be held in person, and streamed via Zoom. No RSVP or registration is required. The presentation will be recorded.

Location

Melbourne School of Population and Global Health

207 Bouverie St Parkville 3053

Level 5, Seminar room 515

Zoom

Click here to join the meeting https://unimelb.zoom.us/j/81807878632?pwd=rt8d9TF4eaBOLkMca0a6Up7hBIsOxa...

Meeting ID: 81807878632

Password: 870819