Reporting quality of papers published in Chilean dental journals . Evaluation period : 2002-2012

Structured summary of trial design, methods, results, and conclusions 67 6 28 0 Introduction Scientific background and explanation of rationale 100 0 0 0 Specific objectives or hypotheses 100 0 0 0 Materials and Description of trial design (such as parallel, factorial) including 67 28 60 methods allocation ratio Important changes to methods after trial commencement (such as 6 0 50 44 eligibility criteria), with reasons Eligibility criteria for participants 78 17 6 0 Settings and locations where the data were collected 78 0 22 0 The interventions for each group with sufficient details to allow 50 28 22 0 replication, including how and when they were actually administered Completely defined pre-specified primary and secondary outcome 39 28 33 0 measures, including how and when they were assessed Any changes to trial outcomes after the trial commenced, with reasons 0 0 50 50 How sample size was determined 22 11 67 0 Table 3. Percentage of compliance with reporting quality of 18 clinical trials. In bold print items with less than 50% of compliance. (c) Cohort studies: Summarizes the follow-up time (e.g., average and 0 0 5 95


INTRODUCTION.
The practice of evidence-based dentistry involves the integration of patient's preferences and expectations, ability and clinical judgment of the dentist and the use of the best evidence available 1 . Best evidence is the evidence produced with an appropriate methodological design that allows to achieve the objective of the research. However, conducting a research is not enough, it must be reported in such a way that other researchers and clinicians may evaluate it in detail and replicate it to verify the reported results.
Recent studies have detected the lack of reporting quality of important methodological aspects in many publications in areas such as orthodontics 2 or implantology 3 . This lack of reporting quality of key methodological elements may be found even in high-impact journals. Pandis et al. 4 found that the 6 major clinical dental journals had significant differences in the quality of their papers. This lack of reporting quality of methodological elements prevents that other researchers evaluate the internal validity of the studies and affects the reproducibility of results.
What are the elements that must be included in a research report? This question has begun to be answered with various systematic reviews that assess the content and quality of available papers. Most reviews have detected the omission of relevant methodological aspects in publications. For this reason, there have been agreements about what are the minimum elements that a scientific publication in the clinical and epidemiological fields should contain. The current agreements are collected in an international, multidisciplinary initiative called Equator-Network, which gathers and organizes the recommendations for the elements and aspects that must be reported in several types of studies 5,6 , clinical trials 7 , observational studies 8 and case reports 9 , among others.
As there are numerous studies available in dental journals, we ask what is the reporting quality of papers published in indexed Chilean dental journals over the period 2002-2012. Therefore, the objective of this study is to evaluate by international guidelines the reporting quality of scientific papers published in Chilean dental journals during 2002-2012. This information would allow to estimate the reproducibility and relative applicability of the papers published in that period.

MATERIALS AND METHODS.
A bibliometric study of publications available in Chilean dental journals over the period 2002-2012 was performed. This research uses the database of a previous study 10 , which establishes the criteria for the selection of journals and papers.
Study sample.
The analyzed journals belong to dental journals published over the period 2002-2012. They were Revista Dental de Chile, Revista Chilena de Ortodoncia, Revista de la Facultad de Odontología de la Universidad de Valparaíso, Revista de la Sociedad Chilena de Odontopediatría, Revista Canal Abierto, Revista Clínica Periodoncia Implantología y Rehabilitación Oral y Oseointegración and International Journal of Odontoestomatology.
The unit of study was the scientific paper. Only case reports, epidemiological observational studies and clinical trials reported in IMRD format (Introduction, Methods, Results and Discussion) were included. All papers available were used. Papers having a different methodological design to the one declared were excluded; for example, a report that was presented as a clinical research but that was actually an in-vitro study was excluded.
Quality assessment The quality of reporting was evaluated according to guidelines recommended by the editors of biomedical journals available in Equator-Network 5 . We define the methodological quality of the report as the characteristic of including sufficient information as to replicate the methodology used by the researchers. Operationally, we expect a report to include as many required items declared in the materials and methods section as possible. Thus, a higher quality report must include most of the items, particularly those described in the Materials and Methods section.
The guidelines used were Case Report Guidelines (CARE) for case reports 9 STrengthening the Reporting of OBservational Studies in Epidemiology (STROBE) for observational studies 8 and Consolidated Standards of Reporting Trials (CONSORT) for clinical trials 7 . A researcher with 10 years of experience in the methodological evaluation of papers conducted the calibration for the implementation of the guidelines designed to assess the methodological quality of a report. Calibration was performed until an intra-examiner agreement of linear weighted Kappa of 0.83 was obtained.
Data extraction Each paper was printed and evaluated by each of the two reviewers (NH, PQ ). They extracted from each paper the title, declared design and manually filled the guideline for each methodological design. Discrepancies in the evaluation were resolved by consultation with a third reviewer (SU). The agreement of the final evaluation for each paper was registered.
Statistical analysis Data were tabulated and analyzed using descriptive statistics. The results for each study were described and compliance percentages were obtained for each item in each of the guidelines to identify areas where the report meets the required aspects.

RESULTS.
Of the 827 items available, we found and analyzed 150 case reports by CARE guidelines, 37 observational studies by STROBE and 18 clinical trials by CONSORT guidelines.
The percentage of compliance of reports by section and type of paper are shown in Figure 1. It shows that sections with greater compliance were Introduction for all designs, while sections with lower compliance were Discussion and Material and Methods for case reports; and Results for observational studies and clinical trials.

Case reports
The analysis of case reports showed that the item with the highest percentage of compliance was presenting the facts of the case (84%). Moreover, the majority of case reports did not report about the perspective of the patient (1%) or if there were problems for diagnosis (3%).
Also, only 4% of case reports reported if the patient had given consent to publish his/her clinical information. Details are shown in Table 1.

Observational studies
The reporting in observational studies shows the highest level of achievement in the Abstract, which describes adequately the essential aspects of the study (95%). However, it fails at reporting the criteria used to select the cohort, the study diagram, and the way how researchers dealt with missing or incomplete data, and how they dealt with losses in follow-up cohort studies, as seen in Table 2.

Clinical trials
Clinical trials showed a good reporting of scientific background and rationale of the tests. On the other hand,
In bold print items with less than 50% of compliance.
they do not explicit whether or not there were changes in the study protocol, who was in charge of the sequence of randomization of patients, which were the periods of re-cruitment and follow-up, if the study was stopped prematurely or not, or if there was a record of the clinical trial protocol. Details are shown in Table 3.

DISCUSSION.
Significant deficiencies in the quality of reporting of the papers analyzed in Chilean dental journals over the period 2002-2012 were found when they were assessed according to international guidelines. In should be no-ted that low reporting quality does not necessarily imply a flawed research, as the name implies, it refers only to the verification of the presence of explicit methodological elements in the scientific publication. The lack of key aspects in the papers analyzed allows us to affirm that Because CARE guidelines are available only since 2013 9 , this is the first report of their use in dental publications. The guidelines used in this study were CARE (2014) 9 .
Observational studies are another category of research design. They allow to evaluate epidemiological characteristics of populations. The three most common designs are prevalence studies, cohort studies, and case-control; these accounted for 18% of the analyzed papers. The aspect that from a methodological point of view negatively affects the quality of these reports is the fact that none of the studies indicated how researchers dealt with the missing data. The importance of explaining the handling of missing data is particularly relevant as such studies frequently use clinical records or forms, which may be incomplete, and it is not clear how researchers handle these situations.
We analyzed 18 clinical trials. From the clinical standpoint, these studies show the importance of evidence on the experimental effectiveness of interventions or treatments. As with the previously discussed designs, they are lacking important elements. Most (73%) do not report the sample size or do it only partially. This aspect is critical to evaluate the results, so that while the majority (72%) of the studies report the statistical methods used, the statistical power of the studies is unclear. Moreover, none of the evaluated clinical trials reported dates of recruitment and follow-up of patients, and the registration of clinical intervention protocol. Currently, most indexed journals require clinical trials to have a prior registration of their protocol. This registration is usually done in the database of the US National Institute of Medicine at https:// clinicaltrials.gov/ or in the International Clinical Trials Registry Platform (ICTRP) of the World Health Organization at http://www.who.int/ictrp/es/. Respect to the quality of reporting of clinical trials our results are similar to those found in pediatric dentistry 15 implantology 16 , community dental health 17 , and even in high-impact journals 18 . However, Froud et al. found that the report of experimental designs with a greater degree of complexity, such as clinical trials grouped by cluster, showed a good reporting quality 19 , only a very small number of Chilean studies published in dental journals over the period 2002-2012 could be considered or included in systematic reviews. Moreover, the absence of these key elements makes replicability of studies difficult and affects negatively its applicability in clinical practice.
We acknowledge that the guidelines used to evaluate the papers were issued after the publication of most of the papers discussed here. While this could be a limitation, these guidelines focus on the essential aspects of the report, those that allow an evaluation of the validation and applicability of the designs used to obtain the results of a particular research. Hence, they provide an objective guide to analyze the reporting quality. Besides, they have been used in other fields of medicine, such as oncology 11 , psychiatry 12 or oncological surgery for head and neck cancer 13 . Researchers in these fields have confirmed that there are deficiencies in the quality of papers in different areas of medicine.
Case reports, which are studies based on anecdotal evidence, are the most common type of papers in the sample. The best reported sections (items 4, 7 and 9A of Table 1) show a good summary of the facts of the case and the main clinical findings of a clinical report, as well as a description of relevant events in the history of the case and details of the intervention. On the other hand, we found a potential problem because only 2% of case reports reported full or in part if the patient had given consent for the publication of his/her medical record. The present research includes studies until 2012. That year law number 20584, which regulates the rights and duties of patients, was passed in Chile. This law states that patients must give researchers explicit consent for the use of their clinical data for any type of scientific research 14 . This is an important aspect to consider for papers published from 2012 in order to avoid any potential legal conflict. Furthermore, only 4% of case reports included some information about the perspective or experience of the patient. These two aspects must be urgently improved to prevent legal, ethical and scientific misconduct. with some deficiencies in specific aspects, such as the calculation of sample size.
Our results show that little evidence published in Chilean dental journals over the period studied would have direct clinical applicability or could be used to generate lines of research due to the lack or omission of essential methodological aspects.
Following the guidelines by the editors does not improve by itself the quality of papers 20 , it requires active implementation. In this regard, Pandis et al. report the results of active interaction between the publisher or editor and researchers as a strategy to improve the quality of clinical trials submitted to the American Journal of Orthodontics and Dentofacial Orthopedics.
They found a significant improvement in quality, with a few items that remained with little report in some specific aspects such as indicating the premature end of the trial, if there were interim analyzes or changes to the protocol 21 .
We also agree with the suggestion of Stevens et al. that publishers or editors should have a more active role in order to provide better quality evidence for medical professionals 22 . Today in Chile, the Journal of Oral Research is the only journal that explicitly requires researchers to report their studies according to established guidelines. Our results strongly suggest that only a small number of papers published in Chilean journals would provide what is considered as good quality evidence.
We believe that a long-term solution can be achieved by strengthening the curriculum research in dentistry schools. In this regard, although the majority (60%) of Chilean schools of dentistry require a research thesis, most of the curricula show that there is lack of organization in students' training, with a poorly organized curriculum. For example, in the case of dental students, statistics is taught at their first years in college and research at the last. There is also a real lack of integration between clinical and research courses 23 .

CONCLUSION.
Case reports, observational and clinical studies published in Chilean dental journals over the period 2002-2012 show a lack of basic design elements and results.
This hampers both the replicability of research and its clinical applicability. Authors, reviewers and editors should make efforts to ensure that future papers published in Chilean journals include the necessary methodological elements to assess their internal and external validity. A useful guide with recommendations is available on the Equator-Network website 5,6 .

ACKNOWLEDGEMENTS.
This research is part of the requirements of NFH and PAQ to obtain the degree of Bachelor of Dentistry, School of Dentistry, School of Medicine, Universidad Austral de Chile.
SU proposed the research, performed the calibration and analysis and wrote the manuscript. DU and WS manually searched and sorted the papers, and reviewed the final draft.
NFH and QPA performed tabulation, data entry, data validation and reviewed the final draft. All authors approved the final version.