Computerized Adaptive Assessment of Disease Impact

描述（由申请人提供）：该SBIR二期项目将开发和评估一种更实用和全面的方法，以标准化疾病特异性健康相关生活质量（HRQOL）的自我评估，该方法适用于日益增长的老年（45岁以上）成年人，并有并发慢性疾病的风险。具体而言，我们建议简化疾病影响筛查（用于预防和早期预警），并通过采用以消费者为中心的“智能测量”系统来监测HRQOL结果，该系统使用项目反应理论（IRT）开发以改进结果测量，利用计算机化自适应测试（CAT）软件来改进数据捕获和评分估计，并通过互联网提供内容。第一阶段证明了行政上的可行性，并支持对特定疾病影响项目的内容进行标准化及其跨疾病校准的假设，以实现对结果更敏感的措施，并在疾病之间具有有意义的可比性。其他以单一疾病为重点的补充性SBIR项目表明，有可能在广泛的严重程度范围内进行更实际和更精确的评估，并有可能消除“上限”和“下限”影响。在第二阶段，我们将结合这些努力，并通过管理来自新的标准化疾病特异性物项库的项目以及先前验证的疾病特异性和通用HRQOL措施，在具有多种合并症的老年人的更大样本中扩展我们的标准化疾病特异性和通用测量模型的测试，以实现以下目标：(1)使用(a)定性方法和专家/消费者评估来改进仪器和项目库，(b)对最普遍和最繁重的慢性病（如哮喘、慢性心力衰竭、慢性肾病、抑郁症、糖尿病、头痛、肥胖、骨关节炎）的大样本进行定量研究，以及(c)对更有效的静态和动态简短形式与“遗留”（先前验证的）工具进行正面比较；(2)开发和评估适合消费者（目前是或不是“患者”）及其临床医生的特定疾病和通用反馈汇总报告，并提供用户友好的规范/基准解释性信息；(3)在纵向前瞻性面板研究中评估自我评估系统，以重复心理测量特性、测试预测效度（医疗支出、角色/社会参与、死亡率）和反应性的测试，并解决数据质量、环境和模式效应、互联网可访问性和被调查者特征的影响；(4)通过对有和无慢性疾病的具有代表性的美国人口样本实施通用和特定疾病措施来收集规范。II期产品将是一个完全可操作的自我评估系统，用于筛查和监测具有多种合并症的老年人口的疾病影响，包括：改进的“静态”和动态（CAT）简短形式，用于评估疾病影响和一般健康结果，用户友好的反馈报告和文档。公共卫生相关性：疾病影响评估工具将测量身体功能和精神健康范围内多种合并症的影响。它将简化疾病负担筛查，为早期预警和卫生保健分层提供更多可操作的信息，并能够更实际地监测健康结果，供提供者、付款人、政府、研究人员以及最重要的是鼓励自我管理和以患者为中心的护理提供的消费者使用。

DESCRIPTION (provided by applicant): This SBIR Phase II project will develop and evaluate a more practical and comprehensive approach to standardizing the self-assessment of disease-specific Health-Related Quality of Life (HRQOL) in the growing population of aging (45+ years) adults with, and at-risk of, co-morbid chronic conditions. Specifically, we propose to simplify disease impact screening (for prevention and early warning), and monitoring of HRQOL outcomes by adapting a consumer-focused Smart Measurement" system, developed using item response theory (IRT) to improve outcomes measures, utilizing computerized adaptive testing (CAT) software to improve data capture and score estimation, and deliver content via the Internet. Phase I demonstrated administrative feasibility and supported the assumptions underlying the standardization of the content of disease-specific impact items and their calibration across diseases to achieve measures that are more responsive to outcomes as well as meaningfully comparable across diseases. Other complimentary SBIR projects focusing on single diseases have demonstrated the potential for more practical and precise assessments over a wide range of severity levels - and the possibility of eliminating "ceiling" and "floor" effects. In Phase II we will combine these efforts and expand our tests of standardized disease-specific and generic measurement models among larger samples of aging adults with multiple co-morbid conditions by administering items from new standardized disease specific item banks as well as previously-validated disease-specific and generic HRQOL measures to address the following aims: (1) Improve the instrument and item pools using (a) qualitative methods and expert/consumer evaluation, (b) quantitative studies among large samples of the most prevalent and burdensome chronic conditions (e.g., asthma, CHF, CKD, depression, diabetes, headache, obesity, osteoarthritis), and (c) head-to-head comparisons between more efficient static and dynamic short forms versus "legacy" (previously-validated) tools; (2) Develop and evaluate aggregated disease-specific and generic feedback report(s), suitable for consumers (who are and are not currently "patients") and their clinicians, with user-friendly normative/benchmark interpretive information; (3) Evaluate the self-assessment system in a longitudinal prospective panel study to replicate tests of psychometric properties, test predictive validity (medical expenditures, role/social participation, mortality) and responsiveness, and address data quality, context and mode effects, accessibility of the Internet, and effects of respondent characteristics; (4) Gather norms by administering generic and disease specific measures to representative US population samples with and without chronic conditions. The Phase II product will be a fully operational self-assessment system for screening and monitoring disease impact for the aging population with multiple co-morbid conditions, including: improved "static" and dynamic (CAT) short forms for assessing disease impact and generic health outcomes, user-friendly feedback reports and documentation. PUBLIC HEALTH RELEVANCE: The Disease Impact Assessment Tool will measure the impact of multiple co-morbidities across the spectrum of physical functioning and mental well-being. It will simplify disease burden screening and provide more actionable information for use in early warning and health care stratification, and enable more practical monitoring of health outcomes for use by providers, payors, governments, researchers and most importantly, consumers that encourage self-management and patient-centered care delivery.