重要修饰语对不良事件证据综合实践中有害影响的作用研究不足：一项实证研究。
Effect of important modifiers on harmful effects in evidence synthesis practice of adverse events were insufficiently investigated: an empirical investigation.
机构信息
Center of Biostatistics, Design, Measurement and Evaluation (CBDME), Department of Clinical Research Management, West China Hospital, Sichuan University, Chengdu, Sichuan, China.
Research Center of Clinical Epidemiology and Evidence-Based Medicine, West China Hospital, Sichuan University, Chengdu, Sichuan, China.
出版信息
BMC Med Res Methodol. 2023 Apr 28;23(1):106. doi: 10.1186/s12874-023-01928-2.
BACKGROUND
Safety is important in the assessment of health interventions, while the results of adverse events are often susceptive to potential effect modifiers since the event risk tends to be rare. In this study, we investigated whether the potential impact of the important effect modifiers on harmful effects was analyzed in meta-analyses of adverse events.
METHODS
Systematic reviews of healthcare interventions, had adverse events as the exclusive outcomes, had at least one meta-analysis, and published between 1 January 2015, and 1 January 2020 were collected. An adverse event was defined as any untoward medical occurrence in a patient or subject in healthcare practice. Six effect modifiers that are the most important for harmful effects were identified by a group discussion. The proportions of eligible systematic reviews that investigated the potential impact of the six effect modifiers on harmful effects were summarized.
RESULTS
We identified 279 systematic reviews eligible for this study. Except for the modifier of interventions/controls (70.61%, 197/279), most of the systematic reviews failed to investigate the potential impact of treatment duration (21.15%, 59/279), dosage (24.73%, 69/279), age (11.47%, 32/279), risk of bias (6.45%, 18/279), and source of funding (1.08%, 3/279) on harmful effects. Systematic reviews with meta-analyses containing more studies were more likely to investigate the potential impacts of these modifiers on the effects, but the proportion was still low (2.3% to 33.3%). Systematic reviews that developed a protocol were significantly more likely to investigate the potential impact of all these effect modifiers (e.g. treatment duration: odds ratio = 5.08, 95% CI: 2.76 to 9.35) on the results.
CONCLUSIONS
Current systematic reviews rarely investigated the potential impact of the important effect modifiers on harmful effects. Methodological guidelines for meta-analysis of adverse events should consider "effect modifier" as one of the domains to help systematic review authors better investigate harmful effects.
背景
安全性在健康干预措施的评估中很重要，而不良事件的结果往往容易受到潜在效应修饰因素的影响，因为事件风险往往很少见。在这项研究中，我们调查了在不良事件的荟萃分析中是否分析了重要效应修饰因素对有害影响的潜在影响。
方法
收集了 2015 年 1 月 1 日至 2020 年 1 月 1 日期间发表的以不良事件为唯一结局的医疗保健干预措施的系统评价，至少有一个荟萃分析。将不良事件定义为医疗保健实践中患者或受试者的任何不良医疗事件。通过小组讨论确定了对有害影响最重要的六个效应修饰因素。总结了有资格调查这六个效应修饰因素对有害影响潜在影响的系统评价的比例。
结果
我们确定了 279 项符合条件的系统评价。除了干预措施/对照修饰因素（70.61%，197/279）外，大多数系统评价都没有调查治疗持续时间（21.15%，59/279）、剂量（24.73%，69/279）、年龄（11.47%，32/279）、偏倚风险（6.45%，18/279）和资金来源（1.08%，3/279）对有害影响的潜在影响。包含更多研究的具有荟萃分析的系统评价更有可能调查这些修饰因素对效应的潜在影响，但比例仍然较低（2.3%至 33.3%）。制定方案的系统评价更有可能调查所有这些效应修饰因素（例如治疗持续时间：比值比=5.08，95%置信区间：2.76 至 9.35）对结果的潜在影响。
结论
目前的系统评价很少调查重要效应修饰因素对有害影响的潜在影响。不良事件荟萃分析的方法学指南应将“效应修饰因素”视为一个领域，以帮助系统评价作者更好地调查有害影响。
Zotero 插件