The rate of academic publishing by authors after their death has increased significantly in biomedical fields since 2000, a new study has found. In the absence of current guidelines, the findings highlight the need for an agreed framework to recognise deceased authors.
Journal publication has become one of the most recognised ways to acknowledge a researcher’s contribution to a particular body of work.
From biochemistry and genetics to pharmacology, biomedical studies can take years to conduct and publish. The review process itself can take months, even years, so it is an unfortunate reality that some authors may pass away between the stage of manuscript acceptance to publication.
However, new research shows the rate of publication for deceased authors has grown significantly since 2000, occurring at a much faster rate than can be accounted for by an increased pool of authors and their corresponding total publications.
The work was led by Professor Chris Hovens from the University of Melbourne and Royal Melbourne Hospital with colleagues from the University of California, Los Angeles (UCLA), Melbourne Bioinformatics, Victorian Comprehensive Cancer Centre and Western Health.
Their results highlight the problem that journal publishing criteria generally do not mention deceased authors. The existing guidelines are, in fact, an obstacle to doing so.
For example, the International Committee of Medical Journal Editors (ICMJE) stipulates that “authors make substantial contributions, participate in manuscript drafting or revision, approve the final manuscript and agree to be accountable for it” – conditions a deceased author would be unable to fulfil.
So, in the absence of appropriate guidelines, how can researchers acknowledge a colleague who dies before a journal paper is published but has significantly contributed to the work?
As a starting point, and to guide future authorship frameworks, Professor Hovens and co-authors conducted the first study to systematically quantify how many authors are still publishing after death in the field of biomedicine.
“I first became interested in this phenomenon after noticing the names of colleagues in authorship lists after they had died,” Professor Hovens says.
“But how they were acknowledged as deceased was very inconsistent. Some journals do include a mention in the Acknowledgements section, but in other journals, there is no mention at all.”
Professor Hovens and colleagues screened 2,601,457 peer-reviewed papers in the biomedical literature from 1990 to 2020 using the full-text Europe PubMed Central database.
Next, natural language processing, filtering and manual curation from obituaries were used to identify a final set of 1,439 deceased authors. The study found that these authors published a total of 38,907 papers over their careers with 5,477 published after death.
Since 1998, the number of papers in this dataset with deceased authors rose from a single paper in 1998 to 952 in 2018. This pattern represents an 18.4 per cent increase in the number of such publications since 2000, significantly outpacing the 5 per cent annual growth rate in total PubMed publications for the same period.
“The top 20 of 1200 deceased authors we identified produced an average of 100 papers after they died, which is more than many living scientists achieve in their whole careers,” Professor Hovens says.
“We also saw a clear trend where authors with the most post-death publications were strongly correlated to the number prior to death, a phenomenon that was significantly higher in the United States.”
Some of the most-published deceased researchers had collected invaluable specimen samples or spearheaded new clinical areas of treatment.
“Our analysis indicates that the attribution of deceased authorship in the literature is not an occasional occurrence but a growing trend."
Professor Hovens explains it's likely that journals don't have guidelines regarding deceased authors because publications of this type in the biomedical literature would have been viewed as an occasional occurrence and of only anecdotal interest, which would likely have been the case in 1978 when the ICMJE criteria were first drafted.
The study authors agree that authorship guidelines are essential to reduce instances of fraudulent and purely honorary authorship, but that improper motives may be less likely to be at play in the case of deceased authors, because they cannot exert undue pressure or coerce co-authors for honorary authorship after death.
Rather, it is more likely that living co-authors recognise the important contribution of their deceased colleagues and genuinely wish to attribute recognition for this.
“One way to address these issues may be a system similar to the CRediT system. This allows authors to note the specific contribution made by each researcher, from conceptualising the study, to providing resources and acquiring funding,” Professor Hovens says.
“A consistent, considered consensus framework to address authorship by deceased biomedical scientists is now needed to create a system that reflects our current research environment.”
Perish and Publish: Dynamics of Biomedical Publications by Deceased Authors