Construction and characteristics of the RxRisk-V - A VA-adapted pharmacy-based case-mix instrument
Construction and characteristics of the RxRisk-V - A VA-adapted pharmacy-based case-mix instrument
复制标题
DOI：
10.1097/00005650-200306000-00009
复制
发表时间：
2003-06-01
期刊：
影响因子：
3
通讯作者：
Sharp, ND
中科院分区：
文献类型：
--
作者：
Sloan, KL;Sales, AE;Sharp, ND
背景。疾病负担评估是卫生保健管理许多方面的关键。患者诊断通常用于病例组合评估。然而，与诊断数据的可用性和可靠性有关的问题使基于药物的策略具有吸引力。我们的目的是为退伍军人健康管理局（VHA）人群中发现的慢性疾病提供一个可靠有效的基于药物的病例组合分类系统。详细介绍RxRisk（以前称为慢性病评分）的va改编版本的开发和类别定义；描述类别的流行度和可靠性；对ICD-9诊断进行分类标准效度检验；并在并行和预期成本模型中评估类别特定的回归系数。研究设计。临床和药理学综述，然后对诊断、药学和用药数据库进行队列分析。华盛顿州，俄勒冈州，爱达荷州和阿拉斯加州的126,075名VHA服务的资深用户。我们使用Kappa统计来评估RxRisk类别的信度和标准效度，并使用多元回归来估计并发和预期成本模型。RxRisk-V将VHA西北网络1998年70.5%的用户平均分为2.61个类别。在45个分类中，33个分类对ICD-9诊断具有良好的1年信度，25个分类对ICD-9诊断具有良好的标准效度。在并发（R-2 = 0.18）和预期成本（R-2 = 0.10）模型中，RxRisk-V占方差的显著比例。RxRisk-V为管理者更好地描述和了解其治疗人群的慢性疾病负担提供了一种可靠和有效的方法。根据VHA进行调整，可以评估针对这一人群的疾病负担。
BACKGROUND. Assessment of disease burden is the key to many aspects of health care management. Patient diagnoses are commonly used for case-mix assessment. However, issues pertaining to diagnostic data availability and reliability make pharmacy-based strategies attractive. Our goal was to provide a reliable and valid pharmacy-based case-mix classification system for chronic diseases found in the Veterans Health Administration (VHA) population.OBJECTIVE. To detail the development and category definitions of a VA-adapted version of the RxRisk (formerly the Chronic Disease Score); to describe category prevalence and reliability; to check category criterion validity against ICD-9 diagnoses; and to assess category-specific regression coefficients in concurrent and prospective cost models.RESEARCH DESIGN. Clinical and pharmacological review followed by cohort analysis of diagnostic, pharmacy, and utilization databases.SUBJECTS. 126,075 veteran users of VHA services in Washington, Oregon, Idaho, and Alaska.METHODS. We used Kappa statistics to evaluate RxRisk category reliability and criterion validity, and multivariate regression to estimate concurrent and prospective cost models.RESULTS. The RxRisk-V classified 70.5% of the VHA Northwest Network 1998 users into an average of 2.61 categories. Of the 45 classes, 33 classes had good-excellent 1-year reliability and 25 classes had good-excellent criterion validity against ICD-9 diagnoses. The RxRisk-V accounts for a distinct proportion of the variance in concurrent (R-2 = 0.18) and prospective cost (R-2 = 0.10) models.CONCLUSIONS. The RxRisk-V provides a reliable and valid method for administrators to describe and understand better chronic disease burden of their treated populations. Tailoring to the VHA permits assessment of disease burden specific to this population.