Our study aimed to evaluate the adherence to the CONSORT checklist of RCTs addressing deep caries management, published in six specific pediatric dental journals between 2010–2022. Our results have revealed that adherence to CONSORT guidelines among RCTs focused on deep caries management in the selected pediatric dental journals, spanning the years 2010 to 2022, has been relatively low. This suggests ample room for improvement in transparency and reporting quality, and some improvement was seen in the last five years of our study. Nevertheless, the assessment of the risk of bias has uncovered methodologic flaws, with 126 studies raising concerns or displaying a high risk of bias. It raises concerns that pediatric patients may have been subjected to experimental interventions with limited benefits and possible harms. Assessing the internal validity and reporting quality of randomized controlled trials is essential to determine their applicability in clinical practice. Research waste, defined as studies not benefiting society, is a major concern that can occur throughout the research process; evidence synthesis methods help identify low priority research questions and avoid unnecessary studies.20 Reducing research waste through evidence synthesis requires collaboration among investigators, publishers, and funding agencies.
Consistent with our findings, previous studies revealed low adherence to the CONSORT statement in various medical and dental fields. For instance, studies have shown suboptimal adherence in fields such as solid organ transplantation21, pediatric dental journals22, orthodontics23, noncarious cervical lesions13, endodontic regenerative procedures for necrotic immature teeth24, posterior restorations25, and dental bleaching26. One study27, like ours, demonstrated an improvement in CONSORT adherence in recent years. The increase was relatively modest: for publications in the last five years, the improvement in CONSORT scores was only 3.46 (compared to those 6–10 years ago) and 4.04 points (compared to those more than ten years ago).
We observed a significant association between CONSORT overall scores and endorsement of the CONSORT guidelines by the journals. This proves the impact of requiring authors to adhere to CONSORT checklists, as demonstrated in previous studies.21,28 Nevertheless, there remains a need for attention from the editors and peer reviewers of all six included journals. While journals endorsing the CONSORT statement showed improvements in overall scores, a significant number of specific items still require attention. Therefore, it is suggested that journals not only endorse CONSORT but also engage in and provide improved training for editors, reviewers, and authors in the rigorous reporting of RCTs.
In contrast to an earlier study27, which reported adequate documentation of items in RCTs on caries prevention, such as estimated effect size, protocol registration, trial design, outcomes, funding, title, and flow diagram, our findings indicate that these specific items were poorly reported in deep caries management trials. This aligns with previous studies13,25 that also highlighted these items for their inadequate reporting. Of particular concern was the insufficient description of sample size calculations, consistent with previous research.13,26 Explicitly stating sample size calculations is vital as it enhances predictability when interpreting intervention effects and bolsters the reproducibility of the RCT. These calculations are based on pre-specified estimates of the expected effect size for the primary outcome. This ensures that if no statistically significant group differences are found, it reflects a true absence of clinically meaningful effects rather than simply an underpowered study unable to detect important disparities.
The majority of the included studies, like previous research findings13,24,25 did not adhere to the practice of pre-registering their trial protocols and subsequently disclosing this information within the published RCT. Trial protocol pre-registration serves to ensure transparency in research and prevents selective outcome reporting. It allows stakeholders to compare the published articles with the original planned protocol. Notably, only one study in our analysis registered its trial protocol and followed the pre-registered protocol without any deviations, aligning with the CONSORT standards and the International Committee of Medical Journal Editors (ICMJE).12,29
Most RCTs received poor scores for the “abstract” item, as they either intentionally or unintentionally omitted significant methodological details, such as type of study design, eligibility criteria, and statistical analysis methods. We underscore the importance of including the “title and abstract” items in the CONSORT tool assessment, as these sections are the most widely read and accessible parts of a research paper. Most readers form their initial judgments about a study based on these sections and quickly scan for key information to determine its relevance for further reading.30 While word limits on abstracts, often imposed by journals, may inadvertently limit authors’ ability to include all pertinent details, this limitation does not justify incomplete communication of essential information. Despite these constraints, authors have an ethical obligation to transparently communicate crucial methodological details and primary results in the abstract.
The items “allocation concealment” and “blinding” need further awareness, as they were either poorly reported or entirely omitted. In contrast, “sequence generation” received relatively good reporting, with approximately 67% of the included studies addressing the item adequately. Adequately reporting allocation concealment is needed, as it complements sequence generation by preventing knowledge of the sequence, ultimately reducing the risk of selection bias.14 The allocation concealment should not be confused with blinding, as the former prevents selection bias, while the latter mitigates performance and detection bias.14 Consistent with other studies.13,23,24, the most well-reported item was “Description of interventions”. This aspect is of particular significance, as it allows for the replication of procedures used in the treatment of deep caries in children, facilitating the testing of their validity. Our risk of bias analysis revealed notable deficiencies in specific domains, namely the randomization process, deviation from intended interventions, and selection of the reported results. This is in line with findings reported in previous studies.13,24 In our study, only 39,4% of the studies reported with low risk of bias for the domain “randomization process”, 27,6% reported with low risk of bias for the domain “deviation from intended interventions”, and domain “selection of the reported results” was reported by 0.8% low risk of bias. It is crucial to recognize that clinical trials categorized as having a significant risk of bias can substantially hinder our ability to draw reliable conclusions, potentially compromising the trustworthiness of caries management recommendations for clinical practice guideline developers. The deficiencies unveiled within RCTs highlight systemic issues in research methodology and transparency, exerting an impact on pediatric dental research. The identification of these common shortcomings paves the way for effective and cohesive quality improvement initiatives that may involve collaboration among editors, peer reviewers, and authors.
We emailed the authors of RCTs that did not report a protocol registration number in their manuscript. We requested they provide us only with the registration number for their trial protocol if it was registered and did not ask them to provide any other information. This missing data on the registry affected our assessments of both risk of bias using the RoB 2 tool and adherence to CONSORT guidelines. For any other missing information regarding study design or methodology that was not declared by the authors in their manuscript, we did not contact them to request that information and considered it a deficiency in reporting quality, in line with the methodology employed by Loguercio et al. and Ortiz et al.26,27 To ensure objectivity and avoid potential conflict of interest, particularly in articles authored by individuals associated with our study, we invited a third reviewer from outside the department with expertise in the field of quality assessment. This reviewer conducted an independent evaluation alongside the other two reviewers. We limited our search to the specific journals chosen as they are well-established, peer-reviewed journals that publish research relevant to caries management topics in pediatric dentistry and align with the methodology employed in previous empirical research to evaluate the reporting quality.31–33 Additionally, a key factor in their selection is that these six journals are indexed in the Web of Science. Being listed in the Web of Science allowed us to extract citation data for studies published in these journals, which was necessary for the analysis of this meta-research.
The adoption of the CONSORT statement by the pediatric dental journals is a very important aspect and of great relevance to improving the reporting quality of RCTs of deep caries management. Our study underscores that adherence to CONSORT remains relatively low, although improvement could be observed over the past 5 years. Furthermore, specific items highlighted in our study require further attention. This finding emphasizes the crucial role of the journal's active endorsement of CONSORT, as shown previously28. As adherence to the CONSORT statement is associated with a reduced risk of bias, it is important to note that some concerns still exist within the deep caries management RCTs in our target journals. Thus, clinical practice guideline developers and stakeholders still need to consider the results of the risk of bias analysis and evaluate the level of evidence included when formulating appropriate recommendations for clinicians.