Naslov
Data sharing practices among authors of biomedical publications

Autori
Gabelica, Mirko

Vrsta, podvrsta i kategorija rada
Ocjenski radovi, doktorska disertacija

Fakultet
Medicinski fakultet u Splitu

Mjesto
Split

Datum
03.12

Godina
2021

Stranica
100

Mentor
Puljak, Livia

Ključne riječi
data ; sharing

Sažetak
Background: This thesis consists of two cross- sectional studies that analyzed data sharing practices among biomedical researchers. The first study analysed the authors’ willingness to share raw data from their randomized controlled trials (RCTs) in anesthesiology. The second study analysed researchers’ compliance with their Data Availability Statement (DAS) from manuscripts published in open access journals with the mandatory DAS. Methods: The first study included RCTs of anesthesiology interventions published from January 1, 2014, to December 31, 2016, in seven journals from the Journal Citation Reports category Anesthesiology belonging to Q1, that is, the highest-ranking 25% of journals in this category, according to the Clarivate Analytics’ Journal Impact Factor distribution. We analyzed full-text articles and included only RCT’s, and there were 619 manuscripts published in those journals in the selected time period. A de- identified email with a request for raw data was sent to corresponding authors. After accessing raw data sets, we checked whether data were available in a way that enabled reanalysis, i.e., published in a file that enabled data use and whether sufficient metadata (description of data and variables that would permit re-use) were included. The second study included manuscripts published during January 2019 in all open-access journals published by BioMed Central (BMC ; part of Nature Springer). We analysed 3416 articles with DAS from 282 journals. We categorized types of DAS. We surveyed corresponding authors who wrote in their DAS that they would share the data on request or under certain circumstances. After accessing raw data sets, we checked whether data were available in a way that enabled reanalysis, i.e., published in a file that enabled data use and whether relevant metadata were included. Results: The first study showed that out of 86 (13.89%) authors from 619 pooled who responded to our query for data sharing, only 24 (3.87%) provided the requested data. Only 24 (3.87%) of manuscripts contained DAS suggesting a willingness to share trial data ; only one of those authors actually shared data. Statistically significant difference was found in the proportion of data sharing between studies with commercial and non-profit funding. Most of the contacted authors did not respond to our query at all ; among the 62 authors who rejected to provide data, reasons were rarely provided. When reasons were provided, arguments 51 included issues regarding data ownership and patient privacy. Only one of the seven analyzed journals stimulated authors toward data sharing. The second study showed that among 254 (14.17%) out of 1792 authors (52.46%) of 3419 DAS papers, who responded to our query for data sharing, only 123 (6.8%)* provided the requested data. Among 1792 manuscripts in which DAS indicated that authors are willing to share their data, 1669 (93.19%) authors either did not respond or refused to share their data with us, i.e., did not comply with their statement. Conclusion: Willingness to share data among the authors of research articles is very low. To achieve widespread availability of de-identified research data, editors should request their publication instead of only encouraging authors to do so or to provide a DAS. Even when authors state in their manuscript that they will share data on request, they have the same compliance rate as those authors that do not have a DAS in their manuscripts, indicating that DAS has little relevance for ensuring data sharing.

Izvorni jezik
Engleski

Znanstvena područja
Javno zdravstvo i zdravstvena zaštita

POVEZANOST RADA