头颈癌随机对照试验中的失访:系统综述
该研究系统检索了Pubmed/MEDLINE、Cochrane和Scopus数据库,纳入美国基于头颈癌诊断、治疗或预防的随机对照试验,共分析128项研究、22,016例随机化患者。结果显示仅35项研究(27.3%)报告了失访,平均失访率为4.37%,除2项统计学离群值外,研究特征不能预测报告失访的比值比。
要点
- 美国头颈癌随机对照试验中仅27.3%报告失访,平均失访率4.37%。
- 除2项统计学离群值外,发表年份、试验中心数、期刊学科、资金来源和干预类型不能预测报告失访的比值比。
- 与95%报告合格性和100%报告随机化的试验相比,仅47%和57%分别报告退出和分析细节。
核心结论
- 美国大多数头颈癌临床试验未报告失访,阻碍了对可能影响显著结果解读的 attrition bias 的评估。
- 需要标准化报告以评估试验结果对临床实践的普适性。
研究设计与样本
- 系统综述,检索Pubmed/MEDLINE、Cochrane和Scopus数据库中的标题。
- 纳入标准为美国基于头颈癌诊断、治疗或预防的随机对照试验;排除回顾性分析和试点研究。
- 共审查3255条标题,128项研究符合纳入标准,共22,016例患者被随机化,平均年龄58.6岁。
- 记录平均年龄、随机化患者数、发表细节、试验中心、资金和失访数据,并进行二元逻辑回归评估研究特征与报告失访的关联。
主要发现
- 共128项研究纳入分析,22,016例患者被随机化,平均年龄58.6岁。
- 35项研究(27.3%)报告了失访,平均失访率为4.37%。
- 除2项统计学离群值外,发表年份、试验中心数、期刊学科、资金来源和干预类型不能预测报告失访的比值比。
- 95%的试验报告了合格性阶段的参与者,100%报告了随机化,但仅47%报告了退出,57%报告了分析细节。
临床怎么用
- 原文未据此给出推荐。
局限性
- 原文未报告局限性。
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
OBJECTIVE: To evaluate the reporting and rates of loss to follow-up (LTFU) in head and neck cancer (HNC) randomized controlled trials based in the United States. DATA SOURCES: Pubmed/MEDLINE, Cochrane, Scopus databases. REVIEW METHODS: A systematic review of titles in Pubmed/MEDLINE, Scopus, and Cochrane Library was performed. Inclusion criteria were US-based randomized controlled trials focused on the diagnosis, treatment, or prevention of HNC. Retrospective analyses and pilot studies were excluded. The mean age, patients randomized, publication details, trial sites, funding, and LTFU data were recorded. Reporting of participants through each stage of the trial was documented. Binary logistic regression was performed to evaluate associations between study characteristics and reporting LTFU. RESULTS: A total of 3255 titles were reviewed. Of these, 128 studies met the inclusion criteria for analysis. A total of 22,016 patients were randomized. The mean age of participants was 58.6 years. Overall, 35 studies (27.3%) reported LTFU, and the mean LTFU rate was 4.37%. With the exception of 2 statistical outliers, study characteristics including publication year, number of trial sites, journal discipline, funding source, and intervention type did not predict the odds of reporting LTFU. Compared to 95% of trials reporting participants at eligibility and 100% reporting randomization, only 47% and 57% reported on withdrawal and details of the analysis, respectively. CONCLUSION: The majority of clinical trials in HNC in the United States do not report LTFU, which inhibits the evaluation of attrition bias that may impact the interpretation of significant findings. Standardized reporting is needed to evaluate the generalizability of trial results to clinical practice.
引用
- Sawaf T, Quinton BA, Moss G, et al. Loss to Follow-up in Head and Neck Cancer: A Systematic Review of Randomized Controlled Trials. Otolaryngol Head Neck Surg. 2023. PMID: 36861844. doi:10.1002/ohn.312
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。