社区人群他汀干预策略预防心血管病效果的马尔可夫模型评价

目的: 在中国鄞州电子健康档案研究(Chinese electronic health records research in Yinzhou, CHERRY)的队列人群中，评价西方发达国家指南普遍推荐的基于风险评估的他汀干预策略对我国发达地区人群心血管病一级预防的效果。方法: 采用马尔可夫模型比较基于风险评估的他汀干预的不同策略，具体包括：(1)不采取基于风险评估的他汀干预的常规策略，作为本研究的对照(策略0)；(2)采用2019年世界卫生组织(World Health Organization, WHO)心血管病风险评估简易模型(无实验室指标)进行风险分层，并对高危人群(风险20%及以上)进行他汀干预(策略1)；(3)采用WHO心血管病风险评估复杂模型(含实验室指标)进行风险分层，并对高危人群(风险20%及以上)进行他汀干预(策略2)；(4)采用中国动脉粥样硬化性心血管病风险预测研究(prediction for atherosclerotic cardiovascular disease risk in China, China-PAR)风险评估模型进行风险分层，并对高危人群(风险10%及以上)进行他汀干预(策略3)。在上述策略的情境下，依据指南对中危人群采取生活方式干预，高危人群采取生活方式加他汀药物干预，研究期限设为10年，马尔可夫模型的循环周期设为1年，模拟10个周期并计算心血管病事件发生数和死亡数等结局事件数，以及每预防一例心血管病事件或死亡的需干预人数(number need to treat, NNT)等效果评价指标。马尔可夫模型的参数主要来源于CHERRY队列人群、公开发表的中国人群研究数据、meta分析及系统综述。采用单因素敏感性分析探讨一般人群心血管病发病率的不确定性对结果的影响，采用概率敏感性分析探讨干预措施效应风险比参数的不确定性。结果: 纳入的225 811名基线未患心血管病的40~79岁的研究人群中，与不采取基于风险评估的他汀干预的常规策略相比，采用WHO简易模型的策略1可预防的心血管病事件为3 482例95%不确定性区间(uncertainty interval, UI): 2 110~4 661]，采用WHO复杂模型的策略2为3 685例(95%UI: 2 255~4 912)，采用China-PAR模型的策略3为3 895例(95%UI: 2 396~5 181)。每预防一例心血管病事件使用他汀的NNT在策略1、2和3分别为22人(95%UI: 14~54)、21人(95%UI: 14~52)和27人(95%UI: 17~67)。策略3能够预防更多的心血管病事件，而策略1和2每预防一例心血管病事件使用他汀的需干预人数更少。单因素敏感性分析及概率敏感性分析的结果与主分析一致。结论: 在我国发达地区人群中采用西方国家心血管病一级预防指南普遍推荐的基于风险评估的他汀干预策略能够取得更好的健康效果；在风险评估工具的选择上，采用China-PAR模型可以获得更多的健康收益，而采用预测变量更少的WHO模型则更有效率。

Effectiveness of statin treatment strategies for primary prevention of cardiovascular diseases in a community-based Chinese population:A decision-analytic Markov model

Objective: To evaluate the effectiveness of statin treatment strategies based on risk assessment for the primary prevention of cardiovascular diseases by the Western guidelines in a community-based Chinese population from economically developed areas using data from the Chinese electronic health records research in Yinzhou (CHERRY) study. Methods: A Markov model was used to evaluate the effectiveness of the following statin treatment strategies, including: (1) usual care without cardiovascular risk assessment(Strategy 0); (2) using the World Health Organization (WHO) non-laboratory-based risk charts with statin treatment for high-risk group (risk ≥ 20%) (Strategy 1); (3) using the WHO laboratory-based risk charts with statin treatment for high-risk group (risk ≥ 20%) (Strategy 2); and (4) using the Prediction for Atherosclerotic cardiovascular disease Risk in China (China-PAR) model with statin treatment for high-risk group (risk ≥ 10%, Strategy 3). According to the guidelines, adults in the medium-risk group received lifestyle intervention, and adults in the high-risk group received life-style intervention and statin treatment under these strategies. The Markov model simulated different strategies for ten years (cycles) using parameters from the CHERRY study, published data, meta-analyses and systematic reviews for Chinese. The number of cardiovascular events or deaths, as well as the number need to treat (NNT) with statin per cardiovascular event or death prevented, were calculated to compare the effectiveness of different strategies. One-way sensitivity analysis on the uncertainty of incidence rate of cardiovascular diseases, and probabilistic sensitivity analysis on the uncertainty of hazard ratios of interventions were conducted. Results: Totally 225 811 Chinese adults aged 40-79 years without cardiovascular diseases at baseline were enrolled. In contrast to the usual care without risk assessment-based statin treatment strategy, Strategy 1 using the WHO non-laboratory-based risk charts could prevent 3 482 95% uncertainty interval (UI): 2 110-4 661] cardiovascular events, Strategy 2 using the WHO laboratory-based risk charts could prevent 3 685 (95%UI: 2 255-4 912) events, and Strategy 3 using the China-PAR model could prevent 3 895 (95%UI: 2 396-5 181) events. NNTs with statin per cardiovascular event prevented were 22 (95%UI: 14-54), 21 (95%UI: 14-52), and 27 (95%UI: 17-67), respectively. Strategy 3 could prevent more cardiovascular events, while Strategies 1 and 2 required fewer numbers need to treat with statin per cardiovascular event prevented. The results were consistent in the sensitivity analyses. Conclusion: The statin treatment strategies based on risk assessment for the primary prevention of cardiovascular diseases recommended by the Western guidelines could achieve substantive health benefits in adults from developed areas of China. Using the China-PAR model for cardiovascular risk assessment could prevent more cardiovascular diseases while using the WHO risk charts seems more efficient.