Center for Evidence Synthesis in Health, Department of Health Services, Policy, and Practice (Primary), Department of Epidemiology (Secondary), Brown University School of Public Health, Providence, Rhode Island, USA.
MRC/NIHR Trials Methodology Research Partnership, Department of Health Data Science, University of Liverpool (a member of Liverpool Health Partners), Liverpool, UK.
J Clin Epidemiol. 2021 Aug;136:168-179. doi: 10.1016/j.jclinepi.2021.04.019. Epub 2021 May 8.
Using recent systematic reviews (SRs), our objectives were to: (1) develop a framework to assess whether a given COS is relevant to the scope of a SR; (2) examine the proportion of SRs for which relevant COS exist; and (3) for SRs for which COS exist, examine the extent to which outcomes in the COS and outcomes in the SR match.
We included a sample of SRs published by the Agency for Healthcare Research and Quality Evidence-based Practice Center Program between January 1, 2018 and October 12, 2020. We searched for potentially relevant COS from the Core Outcome Measures for Effectiveness Trials (COMET) database. We assessed the matching between outcomes recommended by COS and those included in corresponding SRs. When outcomes were matched, we considered matches to be specific (i.e., exact) or general (i.e., non-specific).
Sixty-seven SRs met criteria. We found relevant COS for 36 of 67 SRs (54%). Our framework for comparing the scope of a SR and a COS describes 16 scenarios arising when the breadth of the populations and the interventions are considered. The framework guides systematic reviewers to determine whether a COS is very likely to be relevant, may be relevant, or unlikely to be relevant. Sixty-two percent of outcomes in COS (interquartile range, 40% - 80%) were either specific or general matches to outcomes in SRs.
We found a COS with relevant scope for more than half of the SRs in our sample, with almost two-thirds of the recommended core outcomes matched to outcomes chosen for the SRs. Consideration of COS appears relevant for SR planning and our framework for assessing relevance of a given COS may help with this process.
利用最近的系统评价（SR），我们的目标是：（1）制定一个框架，以评估给定的核心结局测量指标（COS）是否与 SR 的范围相关；（2）检查存在相关 COS 的 SR 的比例；以及（3）对于存在 COS 的 SR，检查 COS 中推荐的结局与 SR 中包含的结局的匹配程度。
我们纳入了 2018 年 1 月 1 日至 2020 年 10 月 12 日期间由美国医疗保健研究与质量循证实践中心计划发布的 SR 样本。我们从核心结局测量指标有效性试验（COMET）数据库中搜索潜在相关的 COS。我们评估了 COS 推荐的结局与相应 SR 中包含的结局之间的匹配程度。当结局匹配时，我们认为匹配是具体的（即精确的）或一般的（即非特定的）。
67 项 SR 符合标准。我们发现 36 项 SR（54%）有相关的 COS。我们用于比较 SR 和 COS 范围的框架描述了当考虑研究人群和干预措施的广度时会出现的 16 种情况。该框架指导系统评价者确定 COS 是否极有可能相关、可能相关或不太可能相关。在 COS 中，62%的结局（四分位间距，40%-80%）与 SR 中选择的结局是具体或一般匹配。
我们在样本中发现了与超过一半的 SR 相关的 COS，其中近三分之二的推荐核心结局与为 SR 选择的结局相匹配。考虑 COS 似乎与 SR 规划相关，我们用于评估给定 COS 的相关性的框架可能有助于这一过程。
