Research that matters – clinical studies

High-quality clinical studies are required to form the endodontic evidence-base for which dentists and patient treatment relies. Although clinical practice can be informed by in vitro, ex vivo and judiciously performed in vivo animal research, without clinical research we cannot establish and develop our knowledge base. Only recently, it was highlighted that in Endodontics, we are lacking a great deal of the vital information to support many of the treatments we routinely perform (Swedish Council on Health Technology Assessment 2012, Bergenholtz & Kvist 2013). Equally, it is evident that many of the research descriptors by which we judge outcome are either absent or insufficiently detailed; there is an urgent need for both improved reporting and consensus on this issue within Endodontics. Although this research area constitutes a priority for the International Endodontic Journal (IEJ), many of the current submissions are unpublishable due to problems with experimental planning, lack of vital methodological information or because the research findings are only preliminary. That said, it is acknowledged that carrying out high-quality clinical research is both difficult and challenging and for that reason this editorial highlights a selection of common submission problems, as well as detailing current guidelines with the aim of improving future quality in this area. Randomized controlled clinical trials (RCTs) are considered the highest standard of human clinical investigative research (Centre for Evidence-Based Medicine 2009); however, it is imperative that such trials are correctly designed, adequately powered, diligently reported and answer a question not satisfactorily answered by the existing literature. A common problem with RCT submissions to the IEJ is that the power calculation is either absent or erroneous; this leads to insignificant findings, perhaps masking an important difference which would have been discovered if a larger sample size had been used. It is critical that the power calculation relies on information retrieved from previous published studies or a relevant pilot study and also accounts for noncompliance and potential dropouts. The rationale and explanation for the power analysis should be detailed in the text; however, it is important to note that if during the study, sufficient participation numbers are not reached (i.e. the study is underpowered), a potential benefit of well-designed, well-performed and well-described RCTs is that the results may still be relevant and used in subsequent meta-analyses. Equally important is the measure itself, it should aim to report not just a significant difference, but a clinically significant difference. To explain this we will give an example; visual analogue scales (VAS) are frequently used in endodontics to describe pre- and postoperative pain experience; however, the chosen significant difference in VAS is often selected as 5-mm on a 100-mm scale. Yet in the general medical literature, it has been regularly indicated that at a difference of 11-15 mm (Ostelo et al. 2008, Kelly 2001) is necessary for a minimal clinical significant difference (MCAD). Often the reported findings reported may be significant, but not clinically relevant, thus negating the purpose of clinical research. Additionally, RCTs should also have outcome measures, one primary and secondary, and the power calculation should be carried out separately for both measures; unfortunately, this is rarely carried out in current endodontic submissions. It is important to acknowledge that not all clinical endodontic questions can be addressed by designing superiority RCTs, some for ethical reasons (e.g. perforations, instrument fracture) and others because randomization is not possible (e.g. systemic health, smoking) or due to patient preference (e.g. surgery versus nonsurgical approaches). As a result, the IEJ encourages the submission of other well-designed prospective clinical trials in these areas using comparative controls; longitudinal observation in cohort studies, community trials and intervention cases series without controls. Prospective cohort studies in particular have an important role to play as endodontic treatment is generally well-standardized and significant prognostic factors are well-recognized [e.g. pain, apical periodontitis (AP)]. Indeed, nonsurgical and surgical endodontic success rates are high (>80% for root canal treatment), which necessitates large numbers of patients for detection of significant differences. This is difficult, expensive and likely prohibitive for RCTs. Practice-based cohorts and multicentred institutional studies may be more able to readily accumulate relevant numbers and therefore present an attractive alternative to collect comparative data (Bergenholtz & Kvist 2013). Retrospective (or case–control studies) by definition cannot be carefully planned, but can be particularly useful for unusual events or exposures (e.g. root fracture, perforation, systemic disease). Although, well-executed original case–control studies will be given due consideration by the editorial board, others retrospectively analysing issues amenable to a clinical trial will be given a lower priority for publication in the journal. Equally cross-sectional epidemiological studies can be sources of valuable clinical data if the study design is adequate (e.g. prevalence of AP, markers of systemic disease/AP); however, if the research is confirmatory of previous research, except from a different population or geographical location, it will also be give a lower priority for publication in the journal. A source of increasing submissions to the IEJ is qualitative clinical studies, which examine quality of life, or patient's attitudes and beliefs. Traditionally, there have been less defined appraisal frameworks for this type of study; however, the consolidated criteria for reporting qualitative research (COREQ) checklist has been recommended (but not introduced yet by leading dental journals) to improve reporting (Tong et al. 2007). Logically qualitative submissions should have a clearly defined research question, a validated methodology and a robust analysis that is not purely descriptive. Systematic reviews of RCTs are considered the gold standard of clinical evidence (Moher et al. 2010) and their inclusion is clearly welcome in the IEJ; however, by definition, the review conclusions are dependent on the quality of the clinical research included (Spångberg 2007). Due to the paucity of RCTs and high-quality cohort studies in endodontics, most systematic reviews are not of RCTs and the findings are limited to a call for further high-quality studies to be carried out. This appears to add little to existing knowledge; however, a rigorous assessment of current knowledge is essential to generate further questions and identify priority areas for future research. Indeed, in the medical journal the Lancet it is stipulated that prior to manuscript acceptance, the relationship between existing and new evidence should be highlighted with direct reference to an existing systematic review (Young & Horton 2005, Clarke et al. 2007). Within endodontics, this goal is aspirational; however, systematic reviews should still clearly address a clinically important question, underline the need for further trials and follow the PICO principle (population, intervention/prognostic factor, comparison and outcome) in their description as well as identifying a list for study exclusion/inclusion, prior to grading the level of evidence of included studies. If for increased statistical power, meta-analysis is included with the review, it is only meaningful as an analytical tool if the data between studies is well-described and homogenous in nature (e.g. if the systematic review included retrospective and prospective studies, it is not appropriate to combine this data statistically). It is common in submissions to this journal that inappropriate research questions are asked as the basis of a systematic review or completely heterogeneous data is combined in a statistical meta-analysis. For several years, RCT submissions to the IEJ have required to conform to the ‘consolidated standards of reporting trials (CONSORT; www.consort-statement.org)’ guidelines, which applies for superiority, equivalence and noninferiority RCT designs (Piaggio et al. 2012), whilst systematic review should conform to the ‘preferred reporting items for systematic reviews and meta-analyses’ (PRISMA; http://prisma-statement.org/) guidelines. Both guidelines involve preparation and inclusion of a checklist and a flow diagram at submission with noncompliant manuscripts returned to the authors before peer review. It is acknowledged that both statements have improved the quality of submissions to medical journals (Turner et al. 2012), although adherence to these guidelines should not be seen as a guarantee of quality assessment or a substitute for meticulous independent study planning (Turner et al. 2012, Savithra & Nagesh 2013). In the future, for observational and epidemiological studies, the IEJ will demand completion of the ‘strengthening the reporting of observational studies in Epidemiology (STROBE)’ checklist and statement, whilst making mandatory clinical trial registration. This registration is free with public clinical trials registries (www.clinicaltrials.gov; http://clinicaltrials.ifpma.org/clinicaltrials/; http://isrctn.org/) and generates a number to be added to the manuscript; this has until now been recommended as good practice by the IEJ. The reason for compulsory STROBE and trial registration is not only to drive improved methodological description, transparency and publication standards (Chalmers & Glasziou 2009, Vandenbroucke et al. 2014), but also improve subsequent clinical treatments (Pihlstrom 2012). Unfortunately at present, there is low public trial registration compliance in Dentistry (Smaïl-Faugeron et al. 2015). It is important to stress that these recommendations are not designed to make life more difficult for authors submitting clinical manuscripts to the IEJ, as we acknowledge the difficulties in carrying out this vital research, but instead by completing the indicated guidelines a resulting improvement in the homogeneity of the data will later assist analysis and systematic review. In conclusion, the IEJ welcomes meaningful high-quality clinical research that will shape the future of endodontics and believe that meticulous planning, good reporting and compliance with suggested guidelines will aid with this process. When planning prospective clinical research studies, investigators would do well to remember the three principle questions of any research project: (1) what do you want to know, (2) what do you want to measure and (3) what can you actually measure (Darvell 1995). We look forward to receiving your future submissions in this area.