Department of Ophthalmology and ORL, Faculty of Medicine.
Faculty of Medicine, Laval University.
J Glaucoma. 2024 Jan 1;33(1):8-14. doi: 10.1097/IJG.0000000000002262. Epub 2023 Jul 20.
Our results demonstrate that, among randomized clinical trials examining the use of surgical treatments for glaucoma, the majority were not registered. Publication bias (PB) was less likely. More than a third of registered trials presented outcome reporting bias (ORB).
Despite the optimum of quality evidence provided by randomized controlled trials (RCTs), biases may be introduced and hinder their application. The primary objective of this study was to investigate ORB and PB in RCTs assessing surgical treatments of glaucoma, as well as their registration status.
A literature review was conducted in MEDLINE, EMBASE, and CENTRAL databases. Inclusion criteria were RCTs published in English between 2007 and 2021 that focused on surgical treatments of patients of all ages with glaucoma or elevated intraocular pressure. Exclusion criteria included cadaveric and animal studies. Registration status was correlated with entries from clinical trial registries. PB was determined by the proportion of trials presenting statistically significant results. ORB was evaluated by comparing the study's primary outcome with that listed in the trial registry. Trials quality was assessed using the Jadad score.
After deleting duplicates, 7561 citations were screened. One hundred sixty-one RCTs were eligible and included between 13 and 556 participants. Of the total, 91% studied an adult population and 71% included patients suffering from primary open angle glaucoma. Among included studies, 63% were not registered and 47% had statistically significant results. An upward trend in registration was observed with time. However, 37% of the studies showed discrepancies between objectives in cited clinical trial registries and the published results.
PB in surgical glaucoma trials was not obvious. Among the minority of trials that were registered, more than a third presented ORB. Unregistered trials had lower quality. RCT registration is crucial for the transparent interpretation of studies, improved patient care in surgery, and informed decision-making.
我们的研究结果表明，在随机临床试验中，大多数针对青光眼手术治疗的试验并未注册。发表偏倚（PB）的可能性较小。超过三分之一的已注册试验存在结局报告偏倚（ORB）。
尽管随机对照试验（RCT）提供了最佳的高质量证据，但可能会引入偏倚并阻碍其应用。本研究的主要目的是调查评估青光眼手术治疗的 RCT 中的 ORB 和 PB，以及它们的注册状态。
对 MEDLINE、EMBASE 和 CENTRAL 数据库进行文献回顾。纳入标准为 2007 年至 2021 年期间发表的英文 RCT，重点关注各种年龄段的青光眼或眼内压升高患者的手术治疗。排除标准包括尸体和动物研究。注册状态与临床试验注册库中的条目相关联。通过呈现统计学显著结果的试验比例来确定 PB。通过比较研究的主要结局与试验注册库中的列出的结局来评估 ORB。使用 Jadad 评分评估试验质量。
删除重复项后，筛选了 7561 条引文。有 161 项 RCT 符合纳入标准，纳入了 13 至 556 名参与者。其中，91%的研究对象为成年人，71%的研究对象患有原发性开角型青光眼。在所纳入的研究中，63%未注册，47%的研究具有统计学意义。随着时间的推移，注册率呈上升趋势。然而，37%的研究在引用的临床试验注册库中的目标与发表的结果之间存在差异。
在针对青光眼手术的 RCT 中，PB 并不明显。在少数已注册的试验中，超过三分之一的试验存在 ORB。未注册的试验质量较低。RCT 注册对于研究的透明解释、手术中患者护理的改善和决策的制定至关重要。
