文章基本信息

标题：Impact of the Transparent Reporting of Evaluations With Nonrandomized Designs Reporting Guideline: Ten Years On
本地全文：下载
作者：Thomas Fuller ; Jaime Peters ; Mark Pearson 等
期刊名称：American journal of public health
印刷版ISSN：0090-0036
出版年度：2014
卷号：104
期号：11
页码：e110-e117
DOI：10.2105/AJPH.2014.302195
语种：English
出版社：American Public Health Association
摘要：Objectives. We assessed how the Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) reporting guideline was used by authors and journal editors in journals’ instructions to authors. We also evaluated its impact on reporting completeness and study quality. Methods. We extracted data from publications that cited TREND on how TREND was used in those reports; we also extracted information on journals’ instructions to authors. We then undertook a case–control study of relevant publications to evaluate the impact of using TREND. Results. Between 2004 and 2013, TREND was cited 412 times, but it was only evidently applied to study reports 47 times. TREND was specifically mentioned 14 times in the sample of 61 instructions to authors. Some evidence suggested that use of TREND was associated with more comprehensive reporting and higher study quality ratings. Conclusions. TREND appeared to be underutilized by authors and journal editors despite its potential application and benefits. We found evidence that suggested that using TREND could contribute to more transparent and complete study reports. Even when authors reported using TREND, reporting completeness was still suboptimal. It is now widely acknowledged that public health policymakers need to include evidence from both nonrandomized and randomized study designs. 1–6 There is evidence, however, that the quality of reporting in public health intervention studies is variable 1 and that outcome reporting bias exists. 7 In addition, poor and incomplete reporting of research adversely affects evidence synthesis and represents a major waste of scarce resources (e.g., time and research funds) and harms that could otherwise be avoided. 8 One strategy to address these concerns has been the development, publication, and promotion of reporting guidelines. Reporting guidelines are checklists, flow diagrams, or detailed texts to guide authors in reporting specific types of research that are developed using clear and transparent methods. 9 Reporting guidelines aim to ensure that fewer studies are missing information that is critical for understanding the methods and results. Simply, if key details of study methods or context are not reported, the findings of studies cannot be used by others. 10 More complete reporting should also increase the likelihood that studies will be appropriately included within evidence syntheses. Ten years ago, the American Journal of Public Health published the Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) statement. 11 The HIV/AIDS Prevention Research Synthesis team of the Centers for Disease Control and Prevention (CDC), in conjunction with editors and representatives of 18 HIV/AIDS–focused journals developed TREND over the course of a 2-day meeting. TREND differs from other reporting guidelines because it focuses on evaluations of public health and behavioral interventions with nonrandomized designs. TREND also includes items pertinent to assessing the effectiveness of interventions and external validity. TREND has 5 sections that address the title and abstract, introduction, method, results, and discussion sections of journal articles; it is available at http://www.cdc.gov/trendstatement . The authors of the TREND statement did not define what they referred to as “public health,” “behavioral,” or “nonrandomized” interventions. However, they did state that TREND was to be used for ... intervention evaluation studies using nonrandomized designs, not for all research using nonrandomized designs. Intervention studies would necessarily include 1) a defined intervention that is being studied and 2) a research design that provides for an assessment of the efficacy or effectiveness of the intervention. 11 ^(p362) Not defining key terminology has the potential benefit of being widely inclusive, but may also risk being perceived as unclear. Following the publication of TREND, several publications expressed a mix of support for it and suggestions for how it could be improved. 1,12 To date, there have been no revisions, extensions, or evaluations of TREND. This is in contrast to the Consolidated Standards of Reporting of Trials (CONSORT) 13 reporting guideline on which TREND was largely based. CONSORT focuses on aspects of study design and methodology relevant to internal validity in randomized controlled trials. It has been revised twice 13,14 and evaluated widely 15 in the time since TREND was first published. In addition, extensions to CONSORT for specific types of trials (e.g., pragmatic trials 16 ), interventions (e.g., for nonpharmacological interventions 17 ), and fields (e.g., social and psychological interventions 18,19 ) have been developed. Findings from these evaluations have indicated that the use of CONSORT is associated with the improved reporting of randomized controlled trials. 20–22 TREND has the potential to play a significant role in improving the evidence base of public health interventions, and could do so if used in conjunction with the new template for intervention description and replication checklist. 23 However, there is an outstanding need to establish whether TREND is of use and benefit. This is fundamental for either further promotion or revision of TREND or refinement of dissemination strategies. Our main objective of this study was to evaluate the impact of TREND and to address 3 research questions: (1) How is TREND used by authors of research articles? (2) How is TREND used in journals’ “instructions to authors”? (3) What is the impact of using TREND on the reporting completeness of articles? A fourth question addressing the issue of what factors affect authors’ and editors’ use of reporting guidelines was addressed in a second study (a currently unpublished article). Because of the exploratory nature of research questions 1 and 2, no hypotheses have been made. Regarding the third question, it was hypothesized that studies that reported the use of TREND would have higher levels of reporting completeness than those that could have used TREND, but did not report doing so.