文献速递关节外科

髋膝关节置换随机对照试验质量评估:一项系统综述

An Assessment of Quality in Hip and Knee Arthroplasty Randomized Controlled Trials: A Systematic Review.
PubMed 文献 · J Arthroplasty · 2026-02-03

该系统综述使用改良Jadad量表评估2014至2023年四本骨科期刊中566项髋膝关节置换RCT的质量,47.5%为高质量,39.8%为中等质量,12.7%为低质量,主要缺陷为缺乏双盲、试验注册报告不足和效能分析缺失。

要点

  • 47.5%的髋膝关节置换RCT为高质量,39.8%中等,12.7%低质量
  • 缺乏双盲是质量分级偏低的主要原因
  • 行业资助和手术技术类RCT更常见低或中等质量

核心结论

  • 超过半数近期髋膝关节置换RCT为低至中等质量,在盲法、试验注册和效能分析方面存在明显缺陷。
  • 需要持续努力提高方法学严谨性和透明度以提升关节置换研究的质量与可信度。

研究设计与样本

  • 系统综述,检索PubMed(2024年5月)纳入2014至2023年四本领先骨科期刊发表的髋膝关节置换RCT,共566项。
  • 使用改良Jadad量表评估随机化、盲法和参与者退出,将研究分为低、中、高质量,并用卡方分析评估文章特征间的关联。

主要发现

  • 566项RCT中,47.5%为高质量,39.8%为中等质量,12.7%为低质量,质量偏低主要因缺乏双盲。
  • 报告行业资助的RCT低或中等质量更常见(相对风险1.27;95%置信区间1.08至1.49;P=0.006)。
  • 聚焦手术技术的RCT低或中等质量更常见(相对风险1.65;95%置信区间1.42至1.90;P<0.001)。
  • 41.2%的文章未报告试验注册,22.3%的RCT未识别资助声明,16.3%的RCT未进行预先效能分析。
  • 进行效能分析的474项RCT中,130项(27.4%)未保留足够患者以满足其效能分析所需的足够统计效能;这130项中77项(59.2%)未发现组间任何统计学显著差异。

关键数值与推荐

  • 改良Jadad量表评估随机化、盲法和参与者退出,将研究分为低、中、高质量。
  • 566项RCT中47.5%高质量、39.8%中等质量、12.7%低质量。
  • 行业资助与低或中等质量的相对风险1.27(95%置信区间1.08至1.49;P=0.006)。
  • 手术技术类与低或中等质量的相对风险1.65(95%置信区间1.42至1.90;P<0.001)。
  • 41.2%未报告试验注册,22.3%未识别资助声明,16.3%未进行预先效能分析;474项进行效能分析的RCT中130项(27.4%)未保留足够患者,其中77项(59.2%)未发现统计学显著差异。

临床怎么用

  • 解读髋膝关节置换RCT时需关注盲法、试验注册和效能分析等质量指标。
  • 对行业资助和手术技术类研究应谨慎评估其方法学质量。
  • 设计RCT时应确保预先效能分析并保留足够患者以达到足够统计效能。

局限性

  • 仅检索PubMed并限定四本领先骨科期刊2014至2023年的RCT,可能未涵盖其他数据库或期刊的研究。
  • 使用改良Jadad量表评估质量,该工具可能无法全面反映研究的方法学质量。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Randomized controlled trials (RCTs) remain the gold standard for evaluating the efficacy and safety of novel interventions. This systematic review assessed the current quality of hip and knee arthroplasty RCTs using a modified Jadad scale, along with other key metrics, including trial registration and conduct of power analyses. METHODS: PubMed was queried in May 2024 to identify hip and knee arthroplasty RCTs published from 2014 through 2023 in four leading orthopaedic journals. Each RCT was scored using a modified Jadad scale, which assesses randomization, blinding, and participant withdrawal, categorizing studies into low, moderate, or high quality. Chi-square analyses were used to assess associations between article characteristics. RESULTS: Of 566 RCTs, 47.5% were graded as high quality, 39.8% as moderate quality, and 12.7% as low quality according to the modified Jadad scale, largely due to a lack of double blinding. Low or moderate quality was more common in RCTs that reported industry funding (relative risk, 1.27; 95% confidence interval, 1.08 to 1.49; P = 0.006) or those focused on surgical technology (relative risk, 1.65; 95% confidence interval, 1.42 to 1.90; P < 0.001). Trial registration was not reported in 41.2% of articles, funding statements were not identified in 22.3% of RCTs, and a priori power analyses were not conducted in 16.3% of RCTs. Of the 474 RCTs that did conduct a power analysis, 130 (27.4%) did not retain enough patients to meet sufficient statistical power per their power analyses. Among these 130 studies, 77 (59.2%) did not identify any statistically significant difference between groups. CONCLUSIONS: More than half of recent hip and knee arthroplasty RCTs were of low to moderate quality, with notable deficiencies in blinding, trial registration, and power analyses. Continued efforts to improve methodological rigor and transparency are essential to advance the quality and credibility of arthroplasty research.

来源:J Arthroplasty,查看原文。版权归原作者所有。

引用

  1. Carpenter LS, Barakat N, Temple JR, et al. An Assessment of Quality in Hip and Knee Arthroplasty Randomized Controlled Trials: A Systematic Review. J Arthroplasty. 2026. PMID: 41643870. doi:10.1016/j.arth.2026.01.083

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。