Do randomised clinical trials on dental caries adopt Open Science practices? Authors: Carolina de Picoli Acosta, Jaisson Cenci, Bruna Brondani, Rokaia Ahmed Elagami, Tatiana Pereira-Cenci, Maximiliano Sergio Cenci, Marie-Charlotte D. N. J. M. Huysmans, Daniela Prócida Raggio, Mariana Minatel Braga, Fausto Medeiros Mendes Published: 2024 ABSTRACT BackgroundOpen Science Practices (OSPs) are essential when assessing research integrity and quality of Randomised Clinical Trials (RCTs). As dental caries represents a significant oral health burden, our objective was to identify and analyse the adoption of OSPs within RCTs focused on addressing this disease.MethodsWe searched PubMed to retrieve RCTs related to dental caries published from January 2000 to March 2022. Two independent researchers assessed a random sample of these articles to evaluate their eligibility until reaching the minimum sample size. Then, the same examiners reviewed the included texts regarding the OSPs adopted in the articles. The collected variables related to OSPs were reporting guidelines, protocol registration, detailed methodology available, open-source software, statistical analysis code sharing, statistical analysis plan, data sharing, open peer review, and open access. Association analyses using logistic regression were conducted considering the publication year, the continent of the first author, impact factor and open-access policy of the journals (explanatory variables), and adoption of at least one OSP or one OSP other than open access (outcomes). The recommendations for adopting OSPs were assessed by reviewing the “Instructions for Authors” section of the most frequently used journals where the included papers were published.Results64.8% of the articles (95% Confidence Interval = 59.3–70.1%) adopted at least one OSP. However, no individual OSP was adopted by more than 50% of the articles. The most adopted practices were protocol registration (37.1%), the use of reporting guidelines (33.1%) and publishing open access (37.3%). These are also the OSPs most often recommended by journals in the Instructions for Authors. A few articles adopted other practices. Older articles presented a lower frequency of adopting these practices, and articles published in higher impact factor journals were positively associated with both outcomes.ConclusionThe RCTs published on dental caries demonstrate a low frequency of adoption of most OSPs. However, a trend toward increased adoption of these practices has been notable in recent years.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12903-024-05218-1. Background Untreated dental caries is the most prevalent oral health issue, affecting approximately two billion individuals globally [1] and consistently impacting negatively their quality of life [2]. Consequently, prevention and management strategies based on the best available evidence for this public health problem are a vast need [3, 4]. Evidence-based practice is the current paradigm in healthcare, advocating the judicious use of the best available evidence to make healthcare decisions [5, 6]. Among the most relevant research designs, randomised clinical trials (RCTs) are the primary studies with the most robust evidence for evaluating health interventions’ efficacy [7]. However, to yield optimal evidence, these studies must be appropriately designed, conducted, and reported [8, 9]. To ensure evaluation of the quality of RCTs, every stage of the research process, from conception and design to the final publication and dissemination of the key findings, should remain transparent and accessible to all stakeholders, including other researchers, practitioners, health policy-makers and patients. This is the core principle of Open Science [10–12]. Open Science is an umbrella term associated with practices researchers can follow to make their work traceable and verifiable. While these practices would be more linked to science dissemination, they are, in fact, intertwined with “responsible research practices”, which focus on the rigorous conduct of research, and with “transparency” linked to the complete reporting of research at every stage of the research lifecycle [13]. Open science practices (OSP) enhance the assessment of research quality and issues related to transparency and integrity [14, 15] and also play a pivotal role in the acceleration of research on therapies for emerging and serious diseases, as observed during the COVID-19 pandemic [16, 17]. In dentistry, although most journals endorse OSPs, such as the use of reporting guidelines and recommendations for data sharing [18], very few articles have made their crude data available [19–22], and most studies have not registered the research protocol prospectively [19]. It is, therefore, crucial that the growing Open Science movement in the last decades, driven by the pressing demands accentuated by the pandemic, does not lose its momentum [16] and is increasingly incorporated into all fields of dentistry. This movement can potentially improve the management of public health issues, such as dental caries, by facilitating more efficient and expedited investigation of therapeutic strategies. This cross-sectional study aimed to identify OSPs on RCTs focusing on dental caries published since 2000 and evaluate factors associated with adopting these practices. Methods Study design This meta-research study was carried out considering published RCTs in the field of dental caries. The protocol was registered in the Open Science Framework [23]. A survey was conducted to evaluate and record OSPs in a representative sample of RCTs related to dental caries management and published since 2000. In addition, variables associated with these practices were investigated. Although it is not a conventional systematic review, we followed the Transparent Reporting of Systematic Reviews and Meta-Analyses (PRISMA) guideline to report this manuscript. The PRISMA checklist can be accessed in the supplemental material. Our research question was, “How often do RCTs related to dental caries adhere to various aspects of OSPs?” We searched articles on the PubMed database using the following search strategy: “dental caries” OR caries OR carious OR decay OR “White spot lesion,” limiting the search to studies published after 2000. The “Randomized Controlled Trial” filter was then used. Sample size and eligibility Based on this initial search and the main objective, a sample calculation was done using the online platform OpenEpi version 3 (www.openepi.com). With a finite population of 2,700 articles (articles available on PubMed using the proposed search strategy in March 2023), an anticipated frequency of 50% considering articles that followed some OSP and an absolute precision of 5%, the minimum number needed to be included was 324 published articles. To select the articles, we conducted a search using the strategy described, and all the articles screened were entered into a spreadsheet and numbered in descending order according to the day of publication. A sequence of random numbers from 1 to 1000 without repetition was then created on the random.org online platform. This order was followed to include articles until the minimum sample number of articles that met the eligibility criteria was reached. Two independent researchers (CPA and BB) evaluated the articles, first through the titles and abstracts, to evaluate their eligibility. The inclusion criteria were RCTs: (A) with subjects allocated to two or more arms, (B) related to dental caries, (C) published after 2000, and (D) fully reported in English. RCTs that investigated interventions focused on students or dentists and articles in which the full text was not found were excluded. Data collection and variables After meeting the inclusion criteria, the full articles were assessed by two trained and calibrated examiners (CPA and BB) to gather data about OSPs. They were previously trained to identify those practices by joint sessions with a senior researcher (FMM). After that, they independently evaluated 30 articles not included in the final sample to evaluate the agreement between the examiners. After they reached a kappa value higher than 0.70 for all assessed practices, they started to evaluate the included articles. The examiners assessed the full text of the included articles and extracted the following data: publication year, first author country, and journal. They also collected the impact factor of the journals according to the 2021 Journal of Citation Reports and whether the journal follows any open access policy (subscription access, hybrid, or gold access). Subscription access is when the articles are only accessible to readers with a subscription. Hybrid journals offer subscription and open-access articles, usually when the authors pay a fee to make their articles in open-access. Gold access journals are those where all content is fully and freely available to readers. The variables related to OSPs collected were: Reporting guidelines: If the authors stated to follow a reporting guideline – (0) No, (1) following the CONSORT (Consolidated Standards of Reporting Trials), or (2) following other reporting guidelines.Protocol registration: If the RCT had the protocol registered – (0) No; (1) Yes.Open notebook or detailed methodology: If the authors presented the detailed methods followed in their study in some way – (0) No, (1) Yes, in the main text, in the material and methods Sect. (2) Yes, as supplemental material, as an open notebook or in an open repository.Open source: If the authors reported the use of any open-source software – (0) No, (1) Yes, or (2) Unclear.Data analysis script sharing: If the authors shared the script code related to their statistical analyses – (0) No; or (1) Yes.Statistical analysis plan: If the RCT mentioned any publicly available statistical analysis plan – (0) No; or (1) Yes.Data sharing: If the article contained some statement about data sharing – (0) No; (1) Yes, crude data was available as supplemental material; (2) Yes, crude data was available in public repositories; (3) Yes, crude data would be shared upon reasonable request.Open peer-review: If the journal or the authors published the peer-review process – (0) No; or (1) Yes.Open access: If the paper was published in open access using any possible route – (0) No; (1) Yes, the article was openly available by the journal itself; (2) Yes, the article is available via PubMed Central; (3) Yes, the full version of the article was available in any public repository. Reporting guidelines: If the authors stated to follow a reporting guideline – (0) No, (1) following the CONSORT (Consolidated Standards of Reporting Trials), or (2) following other reporting guidelines. Reporting guidelines: If the authors stated to follow a reporting guideline – (0) No, (1) following the CONSORT (Consolidated Standards of Reporting Trials), or (2) following other reporting guidelines. Protocol registration: If the RCT had the protocol registered – (0) No; (1) Yes. Protocol registration: If the RCT had the protocol registered – (0) No; (1) Yes. Open notebook or detailed methodology: If the authors presented the detailed methods followed in their study in some way – (0) No, (1) Yes, in the main text, in the material and methods Sect. (2) Yes, as supplemental material, as an open notebook or in an open repository. Open notebook or detailed methodology: If the authors presented the detailed methods followed in their study in some way – (0) No, (1) Yes, in the main text, in the material and methods Sect. (2) Yes, as supplemental material, as an open notebook or in an open repository. Open source: If the authors reported the use of any open-source software – (0) No, (1) Yes, or (2) Unclear. Open source: If the authors reported the use of any open-source software – (0) No, (1) Yes, or (2) Unclear. Data analysis script sharing: If the authors shared the script code related to their statistical analyses – (0) No; or (1) Yes. Data analysis script sharing: If the authors shared the script code related to their statistical analyses – (0) No; or (1) Yes. Statistical analysis plan: If the RCT mentioned any publicly available statistical analysis plan – (0) No; or (1) Yes. Statistical analysis plan: If the RCT mentioned any publicly available statistical analysis plan – (0) No; or (1) Yes. Data sharing: If the article contained some statement about data sharing – (0) No; (1) Yes, crude data was available as supplemental material; (2) Yes, crude data was available in public repositories; (3) Yes, crude data would be shared upon reasonable request. Data sharing: If the article contained some statement about data sharing – (0) No; (1) Yes, crude data was available as supplemental material; (2) Yes, crude data was available in public repositories; (3) Yes, crude data would be shared upon reasonable request. Open peer-review: If the journal or the authors published the peer-review process – (0) No; or (1) Yes. Open peer-review: If the journal or the authors published the peer-review process – (0) No; or (1) Yes. Open access: If the paper was published in open access using any possible route – (0) No; (1) Yes, the article was openly available by the journal itself; (2) Yes, the article is available via PubMed Central; (3) Yes, the full version of the article was available in any public repository. Open access: If the paper was published in open access using any possible route – (0) No; (1) Yes, the article was openly available by the journal itself; (2) Yes, the article is available via PubMed Central; (3) Yes, the full version of the article was available in any public repository. Finally, an examiner (FMM) evaluated the endorsement of OSPs of the top 15 journals where most of the included papers were published by assessing the “instructions for authors” section. This assessment was done in 2023, with no attempt to evaluate older versions. Recommendations to adopt some OSPs were assessed. Besides the journals’ open access policy, the examiner evaluated if the instructions for the authors contained the recommendations related to publishing the protocol, trial registration, following the CONSORT, following other reporting guidelines, data sharing, and other practices. Statistical analyses First, a descriptive analysis was conducted to present the frequency and 95% confidence intervals (95%CI) of each OSP evaluated in the included articles. The frequency of adopting the main OSPs per year was plotted. Association analyses were further conducted. Explanatory variables were the year of publication (quantitative variable), the continent of the first author, the open access policy of the journal (closed or hybrid vs. gold) and the impact factor of the journal (quantitative variables). Two outcomes were evaluated: the adoption of at least one OSP and the adoption of one OSP other than open access. Logistic regression was conducted for both outcomes, and the odds ratio (OR) and respective 95% CI were derived. First, univariate exploratory analyses were conducted, and then a multiple regression model was built based on the variables’ significance. Only variables that had reached a p-value lower than 5% were kept in the final models. The analyses were conducted using a statistical package (Stata 15.0, Stata Corp. College Station, USA), and the significance level was set at 5%. Results The search for articles was performed on March 27, 2022, and 2,758 articles were retrieved. Following a randomly generated list of 1,000 unique numbers, we assessed 760 articles for eligibility to reach the predetermined sample size of 324 articles; hence, 436 were excluded. The reasons for the article’s exclusion are depicted in Fig. 1. Fig. 1Flow chart presenting the number of articles evaluated for eligibility included and analysed Flow chart presenting the number of articles evaluated for eligibility included and analysed The included articles were published between 2000 and 2022, with an average (standard deviation - SD) years since publication (considering 2023 for the calculation) of 9.6 (6.3) years. The mean (SD) impact factor of the journals in which the included articles were published was 3.42 (2.29). Only 62 (19.1%) were published in journals with gold open access. The geographical location of the first author was Europe for 130 (40.1%) articles, Asia for 72 (22.2%), North America for 37 (11.4%), South America for 64 (19.8%), Africa for 11 (3.4%), and Oceania for 10 (3.1%) articles. A total of 64.8% (95%CI = 59.3–70.1%) of the included articles adopted at least one OSP. However, no individual OSP was adopted by more than 50% of the articles, and often, there was 0% adherence in others. The most prevalent OSP was open access, with approximately 40% of the articles (Table 1). In around a third of the articles, the authors stated that they considered reporting guidelines for writing the manuscript (CONSORT, especially), and less than 40% of the RCTs were registered. The use of open-source software was absent or unclear in 96%, and the availability of details on methodology, statistical analysis plans or analysis scripts was (almost) completely absent. Only 13 articles indicated data sharing would be available upon request, and just two articles reported data availability in public repositories. However, we encountered difficulty accessing the data via the link provided in one of these articles. Table 1Descriptive analysis considering the open science practices adopted by the articles analysed (n = 324)Open Science practices N % Report guideline No21767.0 CONSORT10131.2 Others61.9 Protocol register No20463.0 Yes12037.0 Detailed methodology No32299.4 In the main text10.3 As supplemental material10.3 Open-source software No23472.2 Yes123.7 Unclear7824.1 Script analysis available No324100.0 Statistical analysis plan No324100.0 Data sharing No30995.4 In Public repositories20.6 Under reasonable request134.0 Open peer-review No31697.5 Yes82.5 Open access * No20362.7 Yes, in the journal9830.2 PubMed Central (PMC)4915.1 In public repositories72.2* Values exceeded the total of 324 because they may be open access in more than one format Table 1Descriptive analysis considering the open science practices adopted by the articles analysed (n = 324)Open Science practices N % Report guideline No21767.0 CONSORT10131.2 Others61.9 Protocol register No20463.0 Yes12037.0 Detailed methodology No32299.4 In the main text10.3 As supplemental material10.3 Open-source software No23472.2 Yes123.7 Unclear7824.1 Script analysis available No324100.0 Statistical analysis plan No324100.0 Data sharing No30995.4 In Public repositories20.6 Under reasonable request134.0 Open peer-review No31697.5 Yes82.5 Open access * No20362.7 Yes, in the journal9830.2 PubMed Central (PMC)4915.1 In public repositories72.2* Values exceeded the total of 324 because they may be open access in more than one format Descriptive analysis considering the open science practices adopted by the articles analysed (n = 324) * Values exceeded the total of 324 because they may be open access in more than one format In general, we observed an increase in articles adopting all OSPs over time. When we considered the frequency of adoption of the most common OSPs per year, open access emerged as the first practice adopted by the articles. Until 2008, only this practice and the use of reporting guidelines had been adopted. Since 2009, the analysed articles have also included protocol registration. Regarding other OSPs, adopting these practices has become more frequent since 2015 and is increasing (Fig. 2). Fig. 2Frequency of adoption of Open Science Practices (OSPs) per year of study publication Frequency of adoption of Open Science Practices (OSPs) per year of study publication In the univariate analysis, adopting these practices was less frequent in older articles. On the other hand, articles from Asia presented a higher frequency of OSPs than articles from Europe. Moreover, articles published in journals with higher impact factors and gold or diamond open access more frequently adopted OSPs (Table 2). In the multiple analysis, more recent articles and those published in journals with higher impact factors and gold or diamond open access policy presented more frequently adopted at least one OSP (Table 2). Table 2Logistic regression analysis to evaluate the association of explanatory variables and the adoption of at least one open science practice in the analysed articles (n = 324)Explanatory variablesAdoption of open science practicesUnadjusted OR (95%CI) p Adjusted OR(95%CI) p Years since publication 0.81 (0.77 to 0.85)< 0.0010.80 (0.76 to 0.85)< 0.001 Continent (ref.: Europe) 1.00* Asia1.92 (1.02 to 3.61)0.042 North America0.73 (0.35 to 1.51)0.393 South America1.63 (0.86 to 3.09)0.135 Africa1.20 (0.34 to 4.32)0.775 Oceania6.19 (0.76 to 50.36)0.088 Journal’s Open access policy (ref.: closed or hybrid) 1.001.00 Gold3.94 (1.86 to 8.33)< 0.0015.88 (2.33 to 14.85)< 0.001 Journal’s Impact factor (quantitative variable) 1.21 (1.08 to 1.36)0.0011.41 (1.22 to 1.64)< 0.001* Variable not included in the adjusted modelOR = Odds ratio; 95%CI = 95% Confidence interval Table 2Logistic regression analysis to evaluate the association of explanatory variables and the adoption of at least one open science practice in the analysed articles (n = 324)Explanatory variablesAdoption of open science practicesUnadjusted OR (95%CI) p Adjusted OR(95%CI) p Years since publication 0.81 (0.77 to 0.85)< 0.0010.80 (0.76 to 0.85)< 0.001 Continent (ref.: Europe) 1.00* Asia1.92 (1.02 to 3.61)0.042 North America0.73 (0.35 to 1.51)0.393 South America1.63 (0.86 to 3.09)0.135 Africa1.20 (0.34 to 4.32)0.775 Oceania6.19 (0.76 to 50.36)0.088 Journal’s Open access policy (ref.: closed or hybrid) 1.001.00 Gold3.94 (1.86 to 8.33)< 0.0015.88 (2.33 to 14.85)< 0.001 Journal’s Impact factor (quantitative variable) 1.21 (1.08 to 1.36)0.0011.41 (1.22 to 1.64)< 0.001* Variable not included in the adjusted modelOR = Odds ratio; 95%CI = 95% Confidence interval Logistic regression analysis to evaluate the association of explanatory variables and the adoption of at least one open science practice in the analysed articles (n = 324) Journal’s Impact factor (quantitative variable) * Variable not included in the adjusted model OR = Odds ratio; 95%CI = 95% Confidence interval When we excluded the open access and analysed the association with other OSPs, the prevalence of OSPs was 39.5% (95%CI = 34.2–45.1%). Again, we observed that adopting OSPs other than open-access was more frequent in more recent papers and those published in higher-impact factor journals. Moreover, articles with the first author from South America presented twice a higher chance of adopting OSPs in the unadjusted analysis (Table 3) than articles with European first authors. Nevertheless, only years since publication and the journal’s impact factor were associated with adopting OSPs other than open access in the adjusted model (Table 3). Table 3Logistic regression analysis to evaluate the association of explanatory variables and the adoption of at least one open science practice excluding open access in the analysed articles (n = 324)Explanatory variablesAdoption of Open science practices other than open accessUnadjusted OR (95%CI) p Unadjusted OR (95%CI) p Years since publication 0.74 (0.69 to 0.79)< 0.0010.70 (0.65 to 0.76)< 0.001 Continent (ref.: Europe) 1.00* Asia1.77 (0.98 to 3.20)0.058 North America0.89 (0.40 to 1.96)0.766 South America2.23 (1.21 to 4.12)0.010 Africa0.79 (0.20 to 3.11)0.731 Oceania3.14 (0.84 to 11.73)0.088 Journal’s Open access policy (ref.: closed or hybrid) 1.00* Gold1.23 (0.70 to 2.15)0.469 Journal’s Impact factor (quantitative variable) 1.39 (1.24 to 1.57)< 0.0011.68 (1.43 to 1.99)< 0.001* Variable not included in the adjusted modelOR = Odds ratio; 95%CI = 95% Confidence interval Table 3Logistic regression analysis to evaluate the association of explanatory variables and the adoption of at least one open science practice excluding open access in the analysed articles (n = 324)Explanatory variablesAdoption of Open science practices other than open accessUnadjusted OR (95%CI) p Unadjusted OR (95%CI) p Years since publication 0.74 (0.69 to 0.79)< 0.0010.70 (0.65 to 0.76)< 0.001 Continent (ref.: Europe) 1.00* Asia1.77 (0.98 to 3.20)0.058 North America0.89 (0.40 to 1.96)0.766 South America2.23 (1.21 to 4.12)0.010 Africa0.79 (0.20 to 3.11)0.731 Oceania3.14 (0.84 to 11.73)0.088 Journal’s Open access policy (ref.: closed or hybrid) 1.00* Gold1.23 (0.70 to 2.15)0.469 Journal’s Impact factor (quantitative variable) 1.39 (1.24 to 1.57)< 0.0011.68 (1.43 to 1.99)< 0.001* Variable not included in the adjusted modelOR = Odds ratio; 95%CI = 95% Confidence interval Logistic regression analysis to evaluate the association of explanatory variables and the adoption of at least one open science practice excluding open access in the analysed articles (n = 324) Journal’s Impact factor (quantitative variable) * Variable not included in the adjusted model OR = Odds ratio; 95%CI = 95% Confidence interval Finally, from the journals with more articles included, we observed that only one journal recommended all OSPs according to its “Instructions for Authors” section. The journal with more papers included has recommended most of the practices. Among the OSPs, the most commonly recommended was the adherence to CONSORT to report the paper (13 out of 15 journals), followed by the use of other reporting guidelines (12 out of 15), data sharing (10 out of 15) and RCT registration (9 out of 15) (Table 4). Most journals have a hybrid route for open access, and two journals are closed regarding open access. Considering the number of items assessed per journal, the median (range) for the hybrid route journals was 4 (3 to 6) items, and for the gold route journals was 1.5 (0 to 3) items (Table 4). Table 4Characteristics of main journals evaluated in the “Instructions to authors” section concerning the recommendations to adopt Open Science practicesJournalsOpen access policyProtocol publicationProtocol registerCONSORTOther report guidelinesData sharingOther practices N Caries ResearchHybridXXXXX47Journal of DentistryHybridXXX31Journal of Dental ResearchHybridXXX27Clinical Oral InvestigationsHybridXXXXXX25Community Dentistry and Oral EpidemiologyHybridXXXX14Pediatric DentistryClosedXX11Journal of Clinical Pediatric DentistryGoldX11JADAHybridXXXX9BMC Oral HealthGoldXXX8European Archives of Paediatric DentistryHybridXXXX7International Journal of Paediatric DentistryHybridXXXXX7Oral Health & Preventive DentistryGold6European Journal of Oral SciencesHybridXXX5Journal of Dentistry for ChildrenClosedXX5Journal of Indian Society of Pedodontics and Preventive DentistryGoldXX5N = Number of articles included in the study published in the journal Table 4Characteristics of main journals evaluated in the “Instructions to authors” section concerning the recommendations to adopt Open Science practicesJournalsOpen access policyProtocol publicationProtocol registerCONSORTOther report guidelinesData sharingOther practices N Caries ResearchHybridXXXXX47Journal of DentistryHybridXXX31Journal of Dental ResearchHybridXXX27Clinical Oral InvestigationsHybridXXXXXX25Community Dentistry and Oral EpidemiologyHybridXXXX14Pediatric DentistryClosedXX11Journal of Clinical Pediatric DentistryGoldX11JADAHybridXXXX9BMC Oral HealthGoldXXX8European Archives of Paediatric DentistryHybridXXXX7International Journal of Paediatric DentistryHybridXXXXX7Oral Health & Preventive DentistryGold6European Journal of Oral SciencesHybridXXX5Journal of Dentistry for ChildrenClosedXX5Journal of Indian Society of Pedodontics and Preventive DentistryGoldXX5N = Number of articles included in the study published in the journal Characteristics of main journals evaluated in the “Instructions to authors” section concerning the recommendations to adopt Open Science practices N = Number of articles included in the study published in the journal The raw data used in the analysis, including the articles’ titles and PMIDs, can be accessed and downloaded at http://repositorio.uspdigital.usp.br/handle/item/565 [24]. The scripts for analysis in Stata 15.0 and respective outputs were also added to the same repository (http://repositorio.uspdigital.usp.br/handle/item/6885) and are also available as supplemental material. Discussion Given the significant impact of dental caries on the population [1, 2], well-designed RCTs are crucial to investigating effective dental caries treatments [5, 9]. However, these RCTs must adhere to OSPs to allow the summarisation and incorporation of the key findings into systematic reviews and clinical guidelines. This adherence also enables a more accurate observance of the methodological quality and research integrity principles involved in these studies [10–12]. In this context, our study examined the adoption of OSPs in a representative sample of RCTs related to dental caries management published in this century. We found that approximately two-thirds of the articles adopted at least one OSP. However, considering each OSP individually, the adherence drastically decreases to less than half of the articles and is sometimes null. The most common OSP was the open access. Open access has been established since the 1990s, marked by the inception of pioneering platforms such as ArXiv.org, the Open Science Institute (later renamed to Open Science Foundations) and the Scientific Electronic Library (SciELO) in Brazil [25]. It is important to highlight that several universities have recently made it mandatory for their staff to publish open access [26], which may be reflected in our findings. Nevertheless, adopting this practice may face challenges, particularly regarding authors’ financial constraints in covering article processing charges or opting for journals with higher impact factors, which commonly impose open-access fees. In addition to open access, the other most frequently adopted OSPs were protocol registration and using a reporting guideline, mainly the CONSORT. These findings were expected since these OSPs are the most traditional ones. Similar to open access, trial registration has also become commonplace since the Food and Drug Administration Modernization Act of 1997 publication trial registration has also become prevalent since the publication of the Food and Drug Administration Modernization Act of 1997 [27] and the subsequent creation of the clinicaltrials.gov platform. This period also coincides with the conception and first version of the CONSORT guideline [28]. Both initiatives aimed to ensure transparency in the conduct and reporting of RCTs, and they gained momentum in other areas of health, including dentistry. Corroborating the findings observed for dental journals listed in the 2021 Journal Citation Reports [18], we observed that protocol registration and use of a reporting guideline were also the most frequently endorsed by the journals with the highest frequency of RCTs in our survey. Even so, for both practices, less than half of the cariology-related RCTs published after 2000 followed either of these practices and even in the past 5 years, the proportion did not exceed 75%. It is important to note that only the most updated versions of the authors’ instructions for these journals were evaluated, which is a limitation of our study. These instructions may have adopted these recommendations recently, which may partially be linked to a more frequent adoption of OSPs by more recently published articles. On the other hand, we noticed that some specific OSPs are still not covered, even in the most recent versions of the authors’ guidelines, indicating that a particular focus should be given to these practices in the coming years. The endorsement of these practices should also be explicitly included in the guidelines for reviewers, who should also be made aware of the importance of adopting these practices. Moreover, journals should encourage the publication of registered reports that incorporate the reviewers’ suggestions before the conduction of the study, increasing transparency and reducing the risk of publication bias [16]. Studies evaluating OSPs in other areas have found even lower frequencies for different practices compared to our sample, even considering that they evaluated more recently published articles. The frequency of articles with pre-registration was about 16% in articles published in surgical journals [29], about 8% for papers on medical radiology [30] and 12% for articles on sports medicine [31]. A study evaluating more than 10,000 articles on dental research observed a frequency of 7% of the papers with the registered protocol. Most papers included in this previous study were published since 2005, and the evaluation of OSPs was automatised [20]. Regarding reporting guidelines, 8% [30] and 12% [31] of the papers published in radiology and sports medicine, respectively, followed guidelines to report the manuscript. However, the higher values found in our study are probably due to our focus solely on RCTs, while previous studies considered any study design [29–31]. Indeed, guidelines for reporting RCT are pioneers in this sense [32]. In fact, a higher frequency (78%) than the one observed in our study was evidenced among interventional studies on physical activity behaviour change [33]. On the other hand, very few articles adopted other OSPs evaluated in our study. Considering data sharing, probably the OSP currently receiving the most attention, only two studies (0.6%) mentioned making data available in public repositories. However, we could not properly access the data in one of these studies [34]. In other articles, the authors only stated that the data could be made available on reasonable request, which is not counted as “current” data sharing in previous studies [20, 35]. Other articles also found a low prevalence of data sharing in articles published in dentistry, around 2% [20, 21]. While most dental journals have currently endorsed data sharing, adherence to this practice remains recommended for authors rather than being obligatory [18], which can justify the low prevalence of ‘real’ data sharing observed in the articles. This trend is also observed in the medical field, prompting some journals and institutions to require researchers to share raw data by default, except in specific cases with privacy concerns [16]. Despite the low frequency of OSPs, we could observe that the chance of adopting open science practices is higher in more recent articles, corroborating previous findings [20, 21]. While the concept of open science is not novel, its momentum has notably surged in recent years, mainly driven by advancements in digital technology [25]. Therefore, this finding was anticipated. Besides, some additional findings can be addressed considering the timeline analysed. Open access had a significant increase until 2014, but after that, the frequency of its adoption remained stable. On the other hand, protocol registration and use of reporting guidelines first occurred at the end of the first decade, but both practices have shown a steeper increase until now. The adoption of OSPs other than open access, protocol registration and use of reporting guidelines is also increasing, although modestly. Articles published in journals with a higher impact factor also showed a higher frequency of open science practices, which was also observed for journals in dentistry [22] but not in medical radiology [30]. Another variable positively associated with OSPs was the journal’s open-access policy. Although this last one is also an expected finding, adopting OSPs other than open access was only associated with the year of publication and the journal’s impact factor. This finding seems contradictory as open-access journals via the gold route would be expected to encourage other practices. One exception is the open peer review. Of the eight articles in which peer review was open, all were published in gold open-access journals. When we looked at the journal recommendations, only 4 among 15 journals evaluated were gold open-access journals. Of these four journals, one did not recommend any OSP in its instructions to authors, another recommended using reporting guidelines, and another only recommended data sharing. The fourth gold open-access journal recommended registering the protocol, using the CONSORT and sharing data. Other practices were not recommended. However, we did not evaluate changes in the journals over time regarding the recommendations to adopt OSPs, which is a limitation of the study. These findings indicate a low frequency of OSPs, particularly for less conventional ones like open-source code and data sharing, statistical analysis plans, and open notebooks with detailed methodology. Although the discussion about these practices is relatively recent, their adoption appears to be progressing slowly. If this current upward trajectory continues, it will take approximately 20 years to ensure the majority of articles incorporate at least one of these practices. However, the increased frequency observed in the last three years suggests a potential positive shift. Possible initiatives by researchers and journals to expedite the adoption of OSPs should be addressed. The high frequency of RCTs with prior registration and CONSORT adherence is likely mainly due to journal requirements regarding these steps. Therefore, journals should extend these requirements to other practices, such as sharing data and codes, prospective registering and publishing research protocols irrespective of their study designs, and making methodological and research conduct details available. This action may initially restrict some publications, mainly considering some study designs, such as qualitative research [36]. However, it must be a progressive and gradual action, and one of the FAIR principles must always be considered: as open as possible, as closed as necessary [21]. Similarly, funding agencies also have a fundamental role in this progress. Research funded by these agencies should make all steps related to the research process publicly available, a requirement that should be enforced and audited by the grant providers. Some funders have already recommended this [37–39]. Moreover, different stakeholders involved in research should value these outputs related to OSPs in the evaluations and funding distribution systems [12, 39]. Also, a change in the rewarding academic system focusing more on responsible research practices, including open science [40], is necessary to enhance the adoption of the OSPs. The most remarkable example of the importance of open science could be seen during the COVID-19 pandemic [16, 17]. Researchers working on the diagnosis and treatment of dental caries, the oral health problem that most affects the population [1, 2], should follow suit, increasingly adhering to open science practices. This step will undoubtedly bring significant advances in the search for increasingly effective therapeutic alternatives for preventing and controlling this disease. Conclusion In conclusion, although the adoption of OSPs has increased in the last few years, the published dental caries-related RCTs still present a limited and incipient adoption of these practices. Therefore, more efforts are needed to encourage the adoption of open-access policies and educate researchers about this new paradigm’s benefits. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2