Updates to the standardized reporting guidelines endorsed by the Journal

Donald Miller 0 1 2 3 4 Francois Donati 0 1 2 3 4 Penelope Brasher 0 1 2 3 4 Scott Beattie 0 1 2 3 4 David Mazer 0 1 2 3 4 0 P. Brasher, PhD University of British Columbia , Vancouver, BC, Canada 1 F. Donati, MD , PhD University of Montreal , Montreal, QC, Canada 2 D. Miller, MD University of Ottawa , Ottawa, ON, Canada 3 D. Miller, MD (&) The Editorial Office, Canadian Journal of Anesthesia, The Ottawa Hospital , General Campus CCW 1409, 501 Smyth Road, Ottawa , ON K1H 8L9, Canada 4 S. Beattie, MD D. Mazer, MD University of Toronto , Toronto, ON, Canada Consistent with the stance of leading journals in the specialty, the editorial board of the Canadian Journal of Anesthesia staunchly advocates clear scientific writing and transparent reporting of all research. We do so by endorsing key standardized reporting guidelines, advocating the uniform reporting requirements of the International Committee of Medical Journal Editors (ICMJE), and ensuring that the decision letters we send to authors respectfully provide sound and expert guidance to foster improved accuracy and clarity in their revised articles and subsequent reports. Our ''Instructions for Authors'' are regularly updated online (www.springer.com/12630 and www.edmgr.com/cja) to reflect evolving editorial policies, present essential information regarding the required content of submitted articles, and provide numerous hyperlinks that serve as easy navigation tools to related resources for authors. In recent editorials, we have addressed several of the most commonly identified reporting problems in manuscripts submitted to the Journal, including problems relating to sample size estimation1 and statistical issues associated with multiplicity.2 Furthermore, we are making a considerable effort to enhance the rigour of our editorial peer review process, including the statistical reviews. While there is recent evidence of a gradual improvement in the overall quality of reported trials in this and other anesthesia journals,3 there is more work to be done. Although the overall rationale for reporting guidelines was recently addressed,4 over 90 reporting guidelines currently exist. Some guidelines are regularly updated and evidence-based, while others are not. The purpose of this editorial is to provide an update regarding the three reporting guidelines that the Journal currently endorses. These reporting guidelines should serve as equally important resources for authors submitting original research and for the reviewers and editors who evaluate the submitted articles. - Randomized trials: the CONSORT statement Several years ago, the Journal adopted the CONsolidated Standards Of Reporting Trials (CONSORT) Statement (http://www.consort-statement.org), which is an evidencebased minimum set of recommendations for reporting randomized controlled trials (RCTs).5 The most recent version comprises an updated 22-item checklist and a diagram that illustrates the flow of patients throughout a trial. It is our policy to publish this type of flow diagram for larger trials, whereas the essential details of smaller clinical trials can usually be presented in several sentences in the Method section of the related article. In our Instructions for Authors, we advise authors to refer to the CONSORT statement during manuscript preparation of RCTs and to upload the completed checklist during submission. The checklist should identify the corresponding page number where each of the 22 items is reported. In following the elements of the CONSORT statement during the editorial review process, we have identified several recurrent problems. For example, we continue to observe incomplete or inconsistent reporting of the statistical methods used to compare primary outcome(s) and the methods used for analysis of secondary outcomes and subgroup analysis (CONSORT Item #12). To avoid these and related statistical problems, we encourage authors to consult with an experienced biostatistician at the time of trial design and during the subsequent data analysis. Another recurrent problem is the tendency for authors to blur the very important distinction between sequence generation in the randomization process (CONSORT Item #8) and the method of allocation concealment (also part of the randomization processCONSORT Item #9). Sequence generation is the method used to generate the random sequence to ensure comparability of the study groups. Allocation concealment is the method used to implement the random sequence (eg. numbered containers or central telephone) to ensure that this sequence was concealed from those enrolling participants into the study. Randomization is a critically important element of trial design to ensure comparability of the intervention and control groups on any known or unknown potential confounder. Concealment is equally important if the problem of treatment allocation (selection) bias is to be avoided. If the method of allocation concealment in an RCT is not reported, it is impossible to know if a concealment process was actually followed and thus the potential for bias cannot be ruled out. The process should be reported transparently. For further details on CONSORT, please refer to the supporting explanation and elaboration document.6 Observational studies: the STROBE statement The CONSORT statement applies to reports of RCTs only. Recognizing that observational studies may also provide valid scientific information, reporting guidelines have also been developed for these types of studies. The reporting guidelines for STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) Statement (http://www.strobe-statement.org) were published in 2007 and provide recommendations for the three main analytical designs used in observational research, i.e., cohort, case control, and cross-sectional studies.7 The Journal now adopts the STROBE guidelines. Consistent with CONSORT, STROBE 2007 includes a unique 22-item checklist of elements to include in reports of observational studies. As observational studies are prone to issues of selection bias and variably recorded data and outcomes, these items are carefully scrutinized during our peer review process to ensure that methods to address these and other issues have been attended to and explained. In our Instructions for Authors, we advise authors to refer to the STROBE statement during manuscript preparation of observational studies and to upload the completed checklist during submission. The checklist should identify the corresponding page number where each of the 22 items is reported. In following the elements of the STROBE statement during the editorial review process we have identified Item 7, Variables as especially problematic. When reporting observational studies authors invariably rely on multivariable models and yet seldom do they provide any information regarding why variables were considered for inclusion in the model. For further details regarding STROBE, please refer (...truncated)