Objective A high prevalence of authorship problems can have a severe impact on the integrity of the research process. We evaluated the authorship practices of clinicians from the same university hospital in 2019 to compare them with our 2003 data and to find out if the practices had changed.

Methods Practitioners were randomly selected from the hospital database (Hospices Civils de Lyon, France). The telephone interviews were conducted by a single researcher (HM) using a simplified interview guide compared with the one used in 2003. The doctors were informed that their answers would be aggregated without the possibility of identifying the respondents. During the interviews, the researcher ticked the boxes with the answers on a paper file.

Results We interviewed 26 clinicians (mean age 49±8 years) from various medical specialties. They were unfamiliar with the ICMJE (International Committee of Medical Journal Editors) criteria for writing medical articles and felt that these criteria were not well met in general. With regard to ways of reducing the practice of honorary authors, the participants clearly felt that asking for a signature was hypocritical and of little use. The ghost authors were well known; this practice was considered as rather rare. The ‘publish or perish’ has always been cited as being responsible for bad practices (26/26: 100%). We compared these results with those observed in 2003 and no improvement has been observed in the past 15 years.

Conclusion For the second time in France, within a 15-year interval, we have shown that the ICMJE criteria were ignored and that honorary authorship was frequent.