MRI 亚组成像诊断准确性研究报告：对 STARD 2015 的遵循情况。
Reporting of imaging diagnostic accuracy studies with focus on MRI subgroup: Adherence to STARD 2015.
机构信息
Department of Radiology, University of Ottawa Faculty of Medicine, Ottawa, Ontario, Canada.
Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Academic Medical Center, Amsterdam, the Netherlands.
出版信息
J Magn Reson Imaging. 2018 Feb;47(2):523-544. doi: 10.1002/jmri.25797. Epub 2017 Jun 22.
PURPOSE
To evaluate adherence of diagnostic accuracy studies in imaging journals to the STAndards for Reporting of Diagnostic accuracy studies (STARD) 2015. The secondary objective was to identify differences in reporting for magnetic resonance imaging (MRI) studies.
MATERIALS AND METHODS
MEDLINE was searched for diagnostic accuracy studies published in imaging journals in 2016. Studies were evaluated for adherence to STARD 2015 (30 items, including expanded imaging specific subitems). Evaluation for differences in STARD adherence based on modality, impact factor, journal STARD adoption, country, subspecialty area, study design, and journal was performed.
RESULTS
Adherence (n = 142 studies) was 55% (16.6/30 items, SD = 2.2). Index test description (including imaging-specific subitems) and interpretation were frequently reported (>66% of studies); no important differences in reporting of individual items were identified for studies on MRI. Infrequently reported items (<33% of studies) included some critical to generalizability (study setting and location) and assessment of bias (blinding of assessor of reference standard). New STARD 2015 items: sample size calculation, protocol reporting, and registration were infrequently reported. Higher impact factor (IF) journals reported more items than lower IF journals (17.2 vs. 16 items; P = 0.001). STARD adopter journals reported more items than nonadopters (17.5 vs. 16.4 items; P = 0.01). Adherence varied between journals (P = 0.003). No variability for study design (P = 0.32), subspecialty area (P = 0.75), country (P = 0.28), or imaging modality (P = 0.80) was identified.
CONCLUSION
Imaging accuracy studies show moderate adherence to STARD 2015, with only minor differences for studies evaluating MRI. This baseline evaluation will guide targeted interventions towards identified deficiencies and help track progress in reporting.
LEVEL OF EVIDENCE
1 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;47:523-544.
目的
评估影像学杂志中诊断准确性研究对《诊断准确性研究报告的标准》（STARD）2015 年版的依从性。次要目的是确定磁共振成像（MRI）研究报告中的差异。
材料与方法
在 2016 年，我们在影像学杂志中搜索了诊断准确性研究，并评估其对 STARD 2015 版（30 项，包括扩展的影像学特定子项）的依从性。基于成像方式、影响因子、杂志对 STARD 的采用、国家、亚专科领域、研究设计和杂志，对 STARD 依从性的差异进行评估。
结果
（共纳入）142 项研究的依从性为 55%（16.6/30 项，标准差 2.2）。索引试验的描述（包括影像学特定的子项）和解释经常被报告（>66%的研究）；在 MRI 研究中，报告个别项目时，未发现明显差异。不常报告的项目（<33%的研究）包括一些对可推广性（研究设置和地点）和偏倚评估（参考标准评估者的盲法）至关重要的项目。新的 STARD 2015 项目：样本量计算、方案报告和注册很少被报告。高影响因子（IF）杂志报告的项目多于低 IF 杂志（17.2 比 16 项；P=0.001）。采用 STARD 的杂志比非采用者报告的项目多（17.5 比 16.4 项；P=0.01）。各杂志间的依从性存在差异（P=0.003）。研究设计（P=0.32）、亚专科领域（P=0.75）、国家（P=0.28）或成像方式（P=0.80）无差异。
结论
影像学准确性研究对 STARD 2015 版的依从性中等，仅对评估 MRI 的研究有较小差异。此基线评估将指导针对已确定缺陷的有针对性干预措施，并有助于跟踪报告的进展。
证据水平
1 技术功效：2 级 JMRI 2018；47：523-544
Zotero 插件