Mor Vincent
Department of Community Health and Center for Gerontology and Health Care Research, Brown University School of Medicine, Providence, Rhode Island 02192, USA.
Med Care. 2004 Apr;42(4 Suppl):III50-9. doi: 10.1097/01.mlr.0000120104.01232.5e.
The Minimum Data Set (MDS) for nursing home (NH) resident assessment, designed to assess elders functional status and care needs, exemplifies how the information needs of clinical practice are congruent with those of research. Building on a review of the published literature, this article describes the development of the MDS, its reliability and validity testing, as well as the variety of different policy and research uses to which it has been applied. Interrater reliability of items and internal consistency of MDS summary scales is generally good to excellent. Validation studies reveal good correspondence to research quality instruments for cognition, activities of daily living, and diagnoses with more variable results for vision, pain, mood, and behavior scales. To date, no consistent evidence suggests that applications of MDS data for case-mix reimbursement and quality indicator monitoring systematically bias the data. Although facility variation in data quality could compromise some applications, creation of the MDS as a clinical tool for care planning provides an example of how assessment tools with clinical use can be used in administrative databases for research and policy applications.
用于疗养院（NH）居民评估的最小数据集（MDS）旨在评估老年人的功能状态和护理需求，它体现了临床实践的信息需求与研究需求的一致性。基于对已发表文献的回顾，本文描述了MDS的开发、其信效度测试，以及它所应用的各种不同政策和研究用途。项目的评分者间信度和MDS汇总量表的内部一致性总体良好至优秀。效度研究表明，对于认知、日常生活活动和诊断，MDS与研究质量工具具有良好的对应关系，而对于视力、疼痛、情绪和行为量表，结果则更具变异性。迄今为止，没有一致的证据表明将MDS数据用于病例组合报销和质量指标监测会系统性地使数据产生偏差。尽管数据质量的机构差异可能会影响某些应用，但将MDS创建为护理计划的临床工具，提供了一个范例，说明具有临床用途的评估工具如何能用于行政数据库以进行研究和政策应用。
