Dodwell Emily, Dua Shiv, Dulai Sukhdeep K, Astone Kristina, Mulpuri Kishore
*Hospital for Special Surgery, Weill Cornell Medical College, New York, NY †George Washington University Medical School, Washington, DC ‡Stollery Children's Hospital, Edmonton, AB §British Columbia Children's Hospital, University of British Columbia, Vancouver, BC, Canada.
J Pediatr Orthop. 2015 Jul-Aug;35(5):536-45. doi: 10.1097/BPO.0000000000000324.
The quality of randomized controlled trials (RCTs) in orthopaedics is a topic of considerable importance, as RCTs play a major role in guiding clinical practice. The quality of RCTs published between 1995 and 2005 has previously been documented. The purpose of the current study was to assess and describe the quality of pediatric orthopaedic RCTs published from 2005 to 2012, by identifying study characteristics associated with higher quality and outlining areas for improvement.
A standardized literature search was used to identify pediatric orthopaedic RCTs published in 7 well-recognized journals between September 2005 and July 2012 inclusive. The Detsky Quality Assessment Scale and the CONSORT checklist for Non-Pharmacologic Trials were used to assess the quality of the RCTs. Scores for the Detsky and CONSORT were calculated by 2 independent blinded orthopaedic surgeon reviewers with epidemiologic training.
Forty RCTs were included in this analysis. The mean percentage score on the Detsky quality scale was 67%. Sixteen (40%) of the articles satisfied the threshold for a satisfactory level of methodological quality (Detsky >75%). Twenty-five (63%) of these studies were negative studies, concluding no difference between treatment arms. In 52% of the negative studies, an a priori sample size analysis was absent, and 28% were self-described as underpowered. In multiple variable regression analysis, only working with a statistician was significantly associated with higher Detsky percentage scores (P=0.01).
There is a trend for improving quality in pediatric orthopaedic RCTs. Compared with past reports, the mean Detsky score improved from 53% to 67%, and the proportion meeting an acceptable level of quality improved from 19% to 40%. One of the most concerning findings of this study was the lack of attention to sample size and power analysis, and the potential for underpowered studies. Ongoing efforts are necessary to improve the conduct and reporting of clinical trials in pediatric orthopaedics.
Pediatric orthopaedic surgeons, JPO, and POSNA are working toward improving levels of quality in pediatric orthopaedic research. This paper highlights progress that has been made, and addresses some high-yield areas for future improvement.
骨科随机对照试验（RCT）的质量是一个相当重要的话题，因为随机对照试验在指导临床实践中起着主要作用。此前已有关于1995年至2005年间发表的随机对照试验质量的记录。本研究的目的是通过确定与更高质量相关的研究特征并概述改进领域，来评估和描述2005年至2012年间发表的儿科骨科随机对照试验的质量。
采用标准化文献检索，以确定在2005年9月至2012年7月期间（含）在7种知名期刊上发表的儿科骨科随机对照试验。使用德茨基质量评估量表和非药物试验的CONSORT清单来评估随机对照试验的质量。德茨基量表和CONSORT的评分由2名经过流行病学培训的独立双盲骨科外科医生评审员计算得出。
本分析纳入了40项随机对照试验。德茨基质量量表的平均百分比评分为67%。16篇（40%）文章达到了方法学质量令人满意水平的阈值（德茨基评分>75%）。其中25项（63%）研究为阴性研究，结论是各治疗组之间无差异。在52%的阴性研究中，未进行预先的样本量分析，28%的研究自称功效不足。在多变量回归分析中，只有与统计学家合作与更高的德茨基百分比评分显著相关（P = 0.01）。
儿科骨科随机对照试验的质量有提高的趋势。与过去的报告相比，德茨基平均评分从53%提高到了67%，达到可接受质量水平的比例从19%提高到了40%。本研究最令人担忧的发现之一是对样本量和功效分析缺乏关注，以及存在功效不足研究的可能性。有必要持续努力改进儿科骨科临床试验的实施和报告。
儿科骨科外科医生、《儿科矫形外科学杂志》（JPO）和儿科骨科医师协会（POSNA）正在努力提高儿科骨科研究的质量水平。本文强调了已取得的进展，并探讨了一些未来改进的高产领域。
