http://orcid.org/0000-0003-1968-3326
The ultimate goal of orthodontic research is the discovery and dissemination of novel high-quality data to the wider community. However, studies published in orthodontic journals are often compromised by incomplete reporting that can minimise their potential impact by compounding appropriate critical appraisal by the reader, replication by other research teams or inclusion in meta-research. Typical examples include studies that fail to report baseline characteristics of included subjects and standard deviations for continuous outcomes; or those that report percentages of patients with an event of interest at multiple time points, without reporting the effective sample at each time point.
From the reader point of view, not having all the required information means the magnitude, applicability and, ultimately, clinical relevance of the data can be lost.
From the researcher point of view, a lack of data needed to include a study in a meta-analysis means that this study will ultimately not contribute to the evidence base. The meta-analyst retains the option to contact a corresponding author and request missing data, but unfortunately in the vast majority of cases authors either do not respond at all or the data cannot be provided for a variety of reasons. Moreover, it is important to note that although raw measurements taken during a study and the calculated descriptive statistics are objective data, any statistical analysis used to draw inferences between two or more variables always carries a degree of subjectivity. In other words, the same data might be analysed by two persons differently, without either necessarily being wrong.
From the patient point of view, all those who agree to take part in clinical research do so on the basis that the research is carried out and reported to the highest standard and will further existing knowledge in the field, ultimately benefitting the patient community through more effective and efficient patient care. When we, even inadvertently, hold back on such potentially useful information we do our patients a great disservice, by not fulfilling our ethical obligations as researchers.
Table 1. List of general medicine and orthodontic journals and their policy regarding data sharing of research data.
Table 2. List of funders for which clinical trial archiving is required (Hahnel Citation2015).
There is a growing movement to encourage reproducibility and transparency practices within the scientific community, including public access to protocols and raw data, the conduct of replication studies, systematic integration of evidence in systematic reviews, and the documentation of funding and potential conflicts of interest (Iqbal et al. Citation2016).
What can be done
The International Committee of Medical Journal Editors (ICMJE) believes there is an ethical obligation to responsibly share data generated by clinical research, particularly with interventional clinical trials because patients put themselves at risk. The ICMJE has proposed that authors share de-identified individual patient data underlying the results presented in any article (including tables, figures, and appendices or supplementary material) no later than 6 months after publication. The data underlying the results are defined as the individual patient data required to reproduce the article findings. Therefore, ICMJE requires the following as conditions for consideration of publication of a clinical study report in its member journals. As of 1 July 2018, manuscripts submitted to ICMJE journals that report the results of clinical research must contain a data sharing statement (Taichman et al. Citation2017), and clinical trials that begin enrolling participants on or after 1 January 2019 must include a data sharing plan in the trial registration (including their ClinicalTrials.gov protocol), ethical committee approval and patient informed consent. A number of journals have now adopted data sharing policies () and whilst these are not yet strongly enforced, others do maintain more stringent requirements for data sharing (Alsheikh-Ali et al. Citation2011).
Points to consider
Just as patient confidentiality must be protected (through de-identification of the dataset) so must the reasonable rights of researchers and sponsors. The ICMJE has proposed a number of measures for this: ICMJE editors will not consider data deposition in a repository as prior publication; the authors of secondary analyses must attest that this use is in accordance with the terms (if any) agreed to upon their receipt; authors of secondary analyses must reference the source of the data using a unique DOI of a study dataset to provide appropriate credit to those who generated it (and allow searching for the studies it has supported) and authors of secondary analyses must explain completely how theirs differ from previous analyses, if it does. In addition, those who generate and then share clinical trial datasets deserve substantial credit for their work. Efforts are currently placed towards devising appropriate credit or recognition systems in the academic community (Ioannidis et al. Citation2014).
Who does this concern?
Editors of scientific journals in orthodontics can actively encourage or even demand data sharing through ‘Instructions for Authors’ on the journal's website and their online manuscript submission system. This can be done by requiring authors to submit a data statement, including what data from the study are available, who can access the data (and how) and optimally, how to replicate the study results (). This may be through uploading a dataset in an online open repository (Harvard Dataverse, Dryad, Mendeley Data, Zenodo) during manuscript submission, with the authors then providing the dataset unique Digital Object Identifier (DOI) received from the repository, which is published together with the paper. Associate editors and peer reviewers need to be aware of the advantages of data sharing, encourage this practice during peer review and acknowledge open data sharing by submitting authors. Researchers submitting to journals should also be aware that editors may take into consideration data sharing statements and data availability when making editorial decisions. Finally, funding agencies and scientific societies funding orthodontic research can include data sharing as a requirement for all grantees ().
Introducing consistent data sharing in orthodontic journals is difficult and will take time but it will ultimately increase confidence in the conclusions of clinical studies, generate new hypotheses, make the most of what may be learned from each study and enable independent confirmation of results, which is an essential tenet of the scientific process. In the end, data sharing helps fulfil our moral obligation to study participants, thereby benefiting, patients, researchers, clinicians, funding agencies and ultimately, society.
References
- Alsheikh-Ali AA, Qureshi W, Al-Mallah MH, Ioannidis JP. 2011. Public availability of published research data in high-impact journals. PLoS ONE 6:e24357. doi: 10.1371/journal.pone.0024357
- Hahnel M. 2015. Global funders who require data archiving as a condition of grants. figshare. [retrieved 2017 Oct 17]. https://doi.org/10.6084/m9.figshare.1281141.v1
- Ioannidis JP, Greenland S, Hlatky MA, Khoury MJ, Macleod MR, Moher D, Schulz KF, Tibshirani R. 2014. Increasing value and reducing waste in research design, conduct, and analysis. Lancet. 383(9912):166–175. doi: 10.1016/S0140-6736(13)62227-8
- Iqbal SA, Wallach JD, Khoury MJ, Schully SD, Ioannidis JP. 2016. Reproducible research practices and transparency across the biomedical literature. PLoS Biol. 14(1):e1002333. doi: 10.1371/journal.pbio.1002333
- Taichman DB, Sahni P, Pinborg A, Peiperl L, Laine C, James A, Hong ST, Haileamlak A, Gollogly L, Godlee F, et al. 2017. Data sharing statements for clinical trials: a requirement of the International Committee of Medical Journal Editors. JAMA. 317(24):2491–2492. doi: 10.1001/jama.2017.6514