Nothing to Hide: The Merits of Open Data in Dental Research

Scientific research is a lengthy and costly endeavor. It can take years and the collaborative efforts of a team of researchers to gather data for a high-quality manuscript. In the traditional model of scientific publication, these data are summarized in graphs, tables, and carefully selected exemplar images that lead the reader through the story. While this approach facilitates understanding of the work, the failure to share original research data that have not been transformed or processed is an opportunity lost. Making raw data openly available in an accessible format and appropriately annotated allows validation of the work through replication (Macleod 2022), enables integration of findings into broader meta-analyses, and supports the reuse of data for additional analyses (Donaldson and Koepke 2022).

In recognition of these issues, funders of scientific research have moved to promote data sharing. For example, the US National Institutes of Health (NIH) Policy for Data Management and Sharing, introduced in 2023, requires the submission of Data Management and Sharing Plans for certain award types (National Institutes of Health 2023). Many journals also either encourage or require data sharing, often with policies aligned to some extent with 1 of 3 tiers recommended by the Centre for Open Science: (1) disclosure of whether or not data are shared, (2) sharing and citation of data required, or (3) sharing, citation, and validation of data sharing through peer review (Grant et al 2025). Open data sets are gradually becoming recognized as important scholarly output, incentivizing researchers to practice good management of research data (Hahnel et al 2024). Despite these efforts, data sharing has been very limited in the field of dental research. For example, in recent years, the Journal of Dental Research (JDR) and JDR Clinical and Translational Research (JDR CTR) have encouraged data sharing, yet only 47% of original research articles in the JDR published in 12 months up to October 2025 and 20% of those in JDR CTR contained a clear data-sharing statement. Of these, only 60% of JDR papers and 43% of JDR CTR papers stated that data and/or code were already available, with the others indicating that data or code were available from authors upon reasonable request. Studies have shown that only around 7% of researchers who say they will share data actually do so when asked, so statements indicating that data are available upon request are generally not helpful (Gabelica et al 2022). In our brief analysis, we did not validate the data-sharing statements. It is likely that some of those stating they shared data will not have provided raw data in a reusable format. A more detailed analysis of open data in dental research journals found that only 1.5% of manuscripts shared original data openly (Uribe et al 2022).

While the open sharing of data is beneficial in many circumstances, there are many challenges and obstacles. The need to protect patient confidentiality may restrict the sharing of data, particularly when identifiable characteristics cannot be removed. However, even when data are personalized or potentially identifiable, openness can still be promoted through controlled and well-governed access. Instead of full public release, researchers can deposit de-identified or pseudonymized datasets in repositories that operate under data access committees or data-use agreements. This approach allows validation and secondary analysis while maintaining compliance with privacy regulations. Alternatively, abbreviated or synthetic data can be provided in some cases as sharable derivates. In fields such as genomics, sharing of data on specialized repositories has been common for many years. However, these data are useful for others only if they are adequately annotated with metadata that fully describe the samples, processing methods, and analysis tools. Metadata should hence remain openly available to ensure discoverability and citation in line with the FAIR (findable, accessible, interoperable, and reusable) principles (Towse et al 2021). Notably, there have been persistent problems with standardizing metadata that have not yet been resolved (Sheffield et al 2023). On the other hand, there are concerns that making data too easily available has led to a proliferation of low-quality formulaic papers that add little to the scientific record (Spick et al 2025). Controlled data access may be required to help prevent the exploitation and misuse of data. Another concern is that researchers from high-income countries may benefit unfairly from data generated in low- and middle-income countries. Adopting frameworks such as the CARE Principles (Collective benefit, Authority to control, Responsibility, Ethics) or institutional review can help balance openness with respect for participant rights and collective interests. By combining technical measures (robust de-identification, secure repositories) with ethical and governance safeguards, researchers can make their work as open as possible while protecting individual privacy.

Qualitative research presents additional concerns since the epistemological approach in social science often does not aim to identify a single, fixed truth, meaning that replicability cannot be a desired goal (Lamb et al 2024). In addition, sharing full verbatim interview transcripts (even if they can be anonymized) could potentially risk participant disclosure, breaching research ethics committee approval requirements (Prosser et al 2023). However, qualitative research reporting guidance, such as the COREQ checklist (Tong et al 2007), recommends sharing coding trees rather than interview transcripts to demonstrate credibility and dependability. Care must also be taken to understand the study design for clinical trials since not all designs allow direct head-to-head comparisons between treatments (see Adams and Clauw 2025, for an illustrative example).

Difficulties in making data fully open should not prevent us from trying. Therefore, at the JDR and JDR CTR, we have updated our policies to require the sharing of data in articles. We recognize the difficulties with sharing certain data types, so exceptions will be permitted. However, the open sharing of data will be the default and the expectation in most cases. There are many different repositories for sharing data, and guidelines have been published to help authors comply with open data principles (Towse et al 2021). Adherence to these principles will ensure transparency and will help to maximize the benefits from investment in research. We cannot afford to stand still!

Author Contributions

N.S. Jakubovics, contributed to conception and design, data manuscript, critically revised manuscript; F. Schwendicke, V. Muirhead, J. Feine, contributed to design, drafted and critically revised manuscript. All authors gave final approval and agree to be accountable for the work.