Department of Surgery, St Vincent's Hospital Melbourne, The University of Melbourne, Melbourne, Australia.
School of Health Sciences and Social Work, Griffith University, Nathan Campus, Queensland, Australia.
JAMA Netw Open. 2023 Jan 3;6(1):e2250996. doi: 10.1001/jamanetworkopen.2022.50996.
Discontinuation and nonpublication are established sources of avoidable waste among surgical trials, but rates of delayed completion and recruiting shortfalls remain unclear.
To examine the rate of delayed completion, incomplete enrollment, and discontinuation among randomized clinical trials in surgical populations and the duration of delays and extent of recruiting shortfalls among these trials.
DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study examined randomized clinical trials in surgical populations registered on ClinicalTrials.gov between January 1, 2010, and December 31, 2014. Analysis was conducted between October 27, 2021, and June 30, 2022.
The main outcomes were the percentages of trials completed on time or with full enrollment. Delays and recruiting shortfalls were identified by comparing projected enrollment and study timeframes prespecified at the time of registration with the actual study duration and enrollment reported on completion or discontinuation. Absolute and relative differences between planned and actual trial conduct were presented for discontinued trials and those completed with delays or recruiting shortfalls.
In total, 2542 randomized clinical trials in surgical populations were included in the study sample, of which 370 (14.6%; 95% CI, 13.2%-15.9%) were completed both on time and with full enrollment. Approximately 1 in 5 trials (20.4%; 95% CI, 18.9%-22.0%) were completed within their planned timeframe, and 1166 trials (45.9%; 95% CI, 43.9%-47.8%) met their prespecified enrollment target. The median delay among completed trials was 12.2 months (IQR, 5.1-24.3 months) or 66.7% (IQR, 30.1%-135.8%) longer than planned. Among completed trials that did not meet their prespecified enrollment target, the median recruiting shortfall was equivalent to 31.0% (IQR, 12.7%-55.5%) of the planned study sample. A total of 546 trials (21.5%; 95% CI, 19.9%-23.1%) were discontinued. The median time to discontinuation was 26.4 months (IQR, 15.2-45.7 months), and the median recruiting shortfall among discontinued trials was equivalent to 92.7% (IQR, 65.0%-100.0%) of the trial's prespecified enrollment target.
This cross-sectional study found that delayed completion, recruiting shortfalls, and untimely discontinuation were common among surgical trials. These findings highlight the importance of ensuring that investigators and funders do not overestimate the feasibility of planned trials.
在外科试验中，试验的中断和不发表是可避免的浪费的既定来源，但完成延迟和招募不足的发生率仍不清楚。
检查在外科人群中随机临床试验的完成延迟、未完成招募和中断率，以及这些试验的延迟时间和招募不足程度。
设计、设置和参与者：本横断面研究检查了 2010 年 1 月 1 日至 2014 年 12 月 31 日期间在 ClinicalTrials.gov 上注册的外科人群中的随机临床试验。分析于 2021 年 10 月 27 日至 2022 年 6 月 30 日进行。
主要结果是按时或全额招募完成的试验百分比。通过将注册时预设的预计入组和研究时间与实际研究持续时间和完成或中断时报告的实际入组进行比较，确定了延迟和招募不足。对于已中断的试验和延迟或招募不足的试验，呈现了计划和实际试验之间的绝对和相对差异。
共纳入了 2542 项外科人群的随机临床试验，其中 370 项（14.6%；95%置信区间，13.2%-15.9%）按时且全额招募完成。大约每 5 项试验中有 1 项（20.4%；95%置信区间，18.9%-22.0%）在计划时间内完成，1166 项试验（45.9%；95%置信区间，43.9%-47.8%）达到了预设的入组目标。完成试验的中位延迟为 12.2 个月（IQR，5.1-24.3 个月）或计划时间的 66.7%（IQR，30.1%-135.8%）。对于未达到预设入组目标的完成试验，招募不足的中位数相当于计划研究样本的 31.0%（IQR，12.7%-55.5%）。共有 546 项试验（21.5%；95%置信区间，19.9%-23.1%）被中断。中断的中位时间为 26.4 个月（IQR，15.2-45.7 个月），中断试验的中位招募不足相当于试验预设入组目标的 92.7%（IQR，65.0%-100.0%）。
这项横断面研究发现，外科试验中常见的是完成延迟、招募不足和不合时宜的中断。这些发现强调了确保研究人员和资助者不过度高估计划试验可行性的重要性。
