i3 Innovus, Klarabergsviadukten 90 Hus D, 111 64, Stockholm, Sweden.
Eur J Health Econ. 2012 Oct;13(5):561-8. doi: 10.1007/s10198-011-0334-y. Epub 2011 Aug 6.
BACKGROUND/AIMS: Improved data and methods are needed for modeling disease progression in Alzheimer's disease (AD) for economic evaluation of treatments. The aim is to estimate prediction models for long-term AD progression and subsequently economic outcomes.
Three-year follow-up data on 435 patients treated with the cholinesterase inhibitor donepezil in clinical practise were analyzed. Regression models were estimated for long-term prediction of decline in cognitive function (ADAS-cog) and activities in daily living (ADL) ability, risk of institutionalization and costs of care.
The cognitive deterioration was estimated at between 1.6 and 4 ADAS-cog points per every 6 months, increasing with disease severity. Cognitive function was an important predictor of ADL-ability, which itself was the most important predictor of the risk of institutionalization and costs of care. Combining all models in a cross-validation process generated accurate predictions of costs of care at each 6 months follow-up.
The proposed methods for representing AD progression and economic outcomes can be used in micro-simulation models for the economic evaluation of new treatments.
背景/目的：需要改进数据和方法来对阿尔茨海默病（AD）的疾病进展进行建模，以对治疗方法进行经济评估。目的是估计用于长期 AD 进展和随后的经济结果的预测模型。
分析了 435 名在临床实践中接受胆碱酯酶抑制剂多奈哌齐治疗的患者的 3 年随访数据。对认知功能（ADAS-cog）和日常生活活动能力（ADL）下降、住院风险和护理成本的长期预测进行了回归模型估计。
认知功能每 6 个月恶化 1.6 到 4 个 ADAS-cog 点，随疾病严重程度而增加。认知功能是 ADL 能力的重要预测因素，而 ADL 能力本身是住院风险和护理成本的最重要预测因素。在交叉验证过程中结合所有模型可以准确预测每 6 个月随访的护理成本。
用于表示 AD 进展和经济结果的建议方法可用于新治疗方法的经济评估的微观模拟模型。
