The Research Unit of Evidence Synthesis (TRUES), Faculty of Pharmaceutical Sciences, Naresuan University, Phitsanulok, Thailand.
Faculty of Pharmaceutical Sciences, Burapha University, Chon Buri, Thailand.
Res Synth Methods. 2024 Nov;15(6):1161-1174. doi: 10.1002/jrsm.1764. Epub 2024 Oct 23.
Psychological interventions are complex in nature and have been shown to benefit various clinical outcomes. Gaining insight into current practices would help identify specific aspects that need improvement to enhance the quality of network meta-analysis (NMA) in this field. This scoping review aimed to explore methodological approaches in the NMA of psychological interventions. We searched PubMed, EMBASE, and Cochrane CENTRAL in September 2023. We included NMAs of psychological interventions of randomized controlled trials that reported clinical outcomes. Three independent researchers assessed the eligibility and extracted relevant data. The findings were presented using descriptive statistics. Of the 1827 articles identified, 187 studies were included. Prior protocol registration was reported in 130 studies (69.5%). Forty-six studies (24.6%) attempted to search for gray literature. Ninety-four studies (50.3%) explicitly assessed transitivity. Nearly three-quarters (143 studies, 76.5%) classified treatment nodes by the type of psychological intervention, while 13 studies (7.0%) did so by lumping different intervention types into more broader intervention classes. Seven studies (3.7%) examined active components of the intervention using component NMA. Only three studies (1.6%) classified interventions based on factors affecting intervention practices, specifically intensity, provider, and delivery platform. Meanwhile, 29 studies (15.5%) explored the influential effects of these factors using meta-regression, subgroup analysis, or sensitivity analysis. The certainty of evidence was assessed in 80 studies (42.8%). The methodological approach in NMAs of psychological interventions should be improved, specifically in classifying psychological interventions into treatment nodes, exploring the effects of intervention-related factors, and assessing the certainty of evidence.
心理干预具有复杂性，已被证明有益于各种临床结局。了解当前的实践情况有助于确定需要改进的具体方面，以提高该领域网络荟萃分析（NMA）的质量。本范围综述旨在探讨心理干预 NMA 的方法学方法。我们于 2023 年 9 月在 PubMed、EMBASE 和 Cochrane CENTRAL 进行了检索。我们纳入了报告临床结局的随机对照试验的心理干预 NMA。三位独立研究人员评估了纳入标准并提取了相关数据。研究结果采用描述性统计进行呈现。在确定的 1827 篇文章中，有 187 项研究符合纳入标准。130 项研究（69.5%）报告了事先方案注册。46 项研究（24.6%）尝试搜索灰色文献。94 项研究（50.3%）明确评估了可传递性。近四分之三（143 项，76.5%）根据心理干预的类型对治疗节点进行分类，而 13 项研究（7.0%）将不同的干预类型归入更广泛的干预类别。7 项研究（3.7%）使用成分 NMA 检查干预的活动成分。只有 3 项研究（1.6%）根据影响干预实践的因素对干预进行分类，具体为干预强度、提供者和交付平台。同时，29 项研究（15.5%）使用荟萃回归、亚组分析或敏感性分析探索这些因素的影响效应。80 项研究（42.8%）评估了证据的确定性。心理干预 NMA 的方法学方法应得到改进，特别是在将心理干预分类为治疗节点、探索干预相关因素的影响以及评估证据的确定性方面。
