Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ, USA.
Department of Orthopedics, Monmouth Medical Center, Long Branch, NJ, USA.
J Shoulder Elbow Surg. 2023 Sep;32(9):1763-1769. doi: 10.1016/j.jse.2023.04.004. Epub 2023 May 22.
Prospective trial registration has become an important means of improving the transparency and reproducibility of randomized controlled trials (RCTs) and is recommended by the Journal of Shoulder and Elbow Surgery (JSES) per the Consolidated Standards of Reporting Trials (CONSORT) guidelines. Herein, we performed a cross-sectional evaluation of RCTs published in JSES from 2010 to present to determine the prevalence of trial registration and consistency of outcome reporting.
The electronic database PubMed was searched to identify all RCTs on total shoulder arthroplasty (TSA) published in JSES from 2010 to 2022 using the search terms "randomized controlled trial" AND "shoulder" AND "arthroplasty OR replacement." RCTs were considered to be registered if they provided a registration number. For articles that were registered, authors also extracted the registry name, registration date, date of first enrollment, date of last enrollment, and if the primary outcomes reported in the registry were either (1) omitted, (2) newly introduced in the publication, (3) reported as a secondary outcome or vice versa, or (4) varied in timing of assessment compared to the publication. "Early" RCTs were considered those published from 2010 to 2016, whereas "later" RCTs were from 2017 to 2022.
Fifty-eight RCTs met inclusion criteria. There were 16 early RCTs and 42 later RCTs. Twenty-three of the 58 (39.7%) studies were registered, with 9 of 22 with an available registry (40.9%) of those being enrolled prior to patient enrollment. Nineteen of the registered studies (82.6%) provided the name of the registry and a registration number. The proportion of later RCTs that were registered was not significantly different from the early RCTs (45.2% vs. 25.0%, P = .232). Seven RCTs (31.8%) had at least 1 inconsistency compared with the registry. The most common discrepancy was the timing of the assessment (ie, follow-up period) reported in the publication vs. the registry.
Although JSES recommends prospective trial registration, less than half of shoulder arthroplasty RCTs are registered and more than 30% registered trials have at least 1 inconsistency with their registry record. More rigorous review of trial registration and accuracy is necessary to limit bias in published shoulder arthroplasty RCTs.
前瞻性试验注册已成为提高随机对照试验（RCT）透明度和可重复性的重要手段，《肩肘外科杂志》（JSES）根据《临床试验报告统一标准》（CONSORT）指南推荐采用这种方法。在此，我们对 2010 年至 2022 年期间在 JSES 上发表的所有关于全肩关节置换术（TSA）的 RCT 进行了横断面评估，以确定试验注册的流行程度和结果报告的一致性。
使用“randomized controlled trial”和“shoulder”和“arthroplasty OR replacement”这两个检索词，在电子数据库 PubMed 中搜索 2010 年至 2022 年期间在 JSES 上发表的关于全肩关节置换术的所有 RCT。如果文章提供了注册编号，则认为其进行了注册。对于已经注册的文章，作者还提取了注册名称、注册日期、首次入组日期、最后一次入组日期，以及注册中报告的主要结局是否（1）遗漏、（2）在出版物中新增、（3）作为次要结局报告或反之亦然，或者（4）与出版物相比评估时间不同。“早期” RCT 是指发表于 2010 年至 2016 年的 RCT，而“晚期” RCT 是指发表于 2017 年至 2022 年的 RCT。
58 项 RCT 符合纳入标准。其中，16 项为早期 RCT，42 项为晚期 RCT。58 项研究中有 23 项（39.7%）进行了注册，22 项中有 9 项（40.9%）在患者入组前已经进行了注册。23 项已注册研究中有 19 项（82.6%）提供了注册机构名称和注册编号。晚期 RCT 的注册比例与早期 RCT 无显著差异（45.2%比 25.0%，P=0.232）。与注册机构相比，有 7 项（31.8%） RCT 至少有 1 项结果不一致。最常见的差异是出版物中报告的评估时间（即随访期）与注册机构记录的时间不同。
尽管 JSES 建议进行前瞻性试验注册，但只有不到一半的肩关节置换 RCT 进行了注册，超过 30%的注册试验与注册记录至少有 1 项不一致。需要更严格地审查试验注册和准确性，以限制已发表的肩关节置换 RCT 中的偏倚。
