St-Louis Etienne, Oosenbrug Marcus, Landry Tara, Baird Robert
Division of General Surgery, McGill University, Montréal, Quebec, H3G 2M1, Canada; Division of Pediatric General and Thoracic Surgery, McGill University Health Centre, Montreal, Quebec, H4A 3J1, Canada.
Faculty of Medicine, McGill University, Montréal, Quebec, H3G 2M1, Canada.
J Pediatr Surg. 2018 May;53(5):879-884. doi: 10.1016/j.jpedsurg.2018.02.009. Epub 2018 Feb 8.
Pediatric surgical randomized clinical trials (RCTs) are labor-intensive and costly. This systematic review investigated patient accrual and estimates of study duration in RCTs by interrogating enrollment and registration practices.
We performed a peer-review search of multiple databases from 2000 to 2016 evaluating RCTs salient to the field with inclusion mandated that a self-identified pediatric surgeon be listed as an author. Trial registries were also searched. RCTs were appraised, and predictors of success were evaluated using multivariate logistic regression, with success defined as achievement of recruitment objectives.
After screening, 137 RCTs were analyzed. Mean Jadad score was 1.80 (median=2). CONSORT scores ranged between 17% and 97% (median=58%). Sixty-seven studies described sample-size determination, 49 reported projected enrollment, and 26 were successful. Among 26 registered RCTs, 15 disclosed their expected completion date, which was achieved by 8. On average, protocols underwent 3.42 iterations. 9% of trials were terminated before completion, most commonly owing to poor recruitment. Trial registration and urgent cases significantly predicted success on multivariable analysis (p<0.05).
Overall quality of reporting in pediatric surgical trials is poor. Sample-size calculation and patient accrual are frequently poorly performed or underestimated, resulting in trial overrun and/or premature termination. These data may help inform subsequent study design and facilitate successful completion.
Level III-Systematic Review and Observational (Case-Control) Analysis.
儿科外科随机临床试验（RCT）工作量大且成本高昂。本系统评价通过审视入组和注册实践，调查了RCT中的患者入组情况及研究持续时间的估计。
我们对2000年至2016年多个数据库进行同行评审搜索，评估该领域相关的RCT，纳入标准要求列出一名自认为是儿科外科医生的作者。还搜索了试验注册库。对RCT进行评估，并使用多变量逻辑回归评估成功的预测因素，成功定义为实现招募目标。
筛选后，分析了137项RCT。Jadad评分均值为1.80（中位数=2）。CONSORT评分在17%至97%之间（中位数=58%）。67项研究描述了样本量的确定，49项报告了预计入组人数，26项取得成功。在26项注册的RCT中，15项披露了预期完成日期，其中8项实现了该日期。方案平均进行了3.42次迭代。9%的试验在完成前终止，最常见原因是招募不佳。多变量分析显示，试验注册和紧急病例显著预测成功（p<0.05）。
儿科外科试验的总体报告质量较差。样本量计算和患者入组常常执行不佳或估计不足，导致试验超期和/或提前终止。这些数据可能有助于为后续研究设计提供参考并促进成功完成。
III级——系统评价和观察性（病例对照）分析。
