基于马尔可夫微模拟模型的乳腺癌筛查策略优选

Breast cancer screening is generally accepted as an effective means of reducing breast cancer mortality and improving survival of breast cancer patients in many western countries. There are many differences between breast cancer in Asia compared with western countries. The incidence rates remain low in Asia, although increasing at a more rapid rate than in western countries, due to changes in the lifestyle and diet. Other than lower incidence rates than in the Western countries, asian women typically have denser breasts and smaller size of breast. The performance of breast cancer screening modalities may be also different in Asian women. Most screening programs are not population-based, and are either self-financed privately or partially subsidized by the government in Asian countries. Population screening is still controversial in China. There are only limited data available on breast screening in the Chinese population. It is worth investigating the cost-effectiveness of screening strategies for breast cancer in China which is still in resource-limited environments. This study explores methods for finding optimal screening strategies by modelling the costs and benefits of breast cancer screening programs through a simulation-based and a theoretical approach. A program that includes free breast cancer screenings were offered to 530 000 women aged from 35-69 during 2008 to 2010. Data on performance of clinical examination, breast ultrasound and mammography were collected. A state-transition Markov microsimulation model will be developed to simulate disease history of breast cancer, breast cancer screening, diagnosis and treatment in a hypothetical, population based cohort of one million Chinese women. Screening strategies would be specified in terms of combinations screening modalities include clinical breast examination, B ultrasound and mammography. Combinations of screening modalities in terms of testing frequency and target age group will be evaluated in addition. And then a probabilistic sensitivity analysis will be conducted involved specifying appropriate probabilistic distributions for clinical parameters and employing a Monte-Carlo simulation with 1000 runs to select values at random from those distributions. Based on the simulation results, cost-effectiveness acceptability curves will be constructed to present the uncertainty in the incremental cost-effectiveness ratio caused by sampling variation. The most cost-effective approach depends on how much society is willing to pay to save a year of life which is influence by many social factors. The probability of cost-effectiveness at different willingness-to-pay threshold will be presented. For different willingness to pay thresholds, the optimal strategies will be examined. A societal perspective will be adopted during the cost effectiveness analysis.

乳腺癌筛查在西方国家已被广泛接受是恶性肿瘤中最能有效地提高患者生存率和降低死亡率的群防措施。但是由于亚洲女性乳腺癌的发病率低于西方国家，而且乳腺生理结构如大小、致密度等也与西方女性不同，因此在亚洲女性中进行乳腺癌筛查的效果有待进一步研究。本研究拟以卫生部2008年启动的覆盖全国30个省53个县的乳腺癌筛查项目积累的实践数据和临床资料为基础，应用马尔科夫微模拟模型，构建中国女性乳腺癌筛查策略卫生经济学评价模型，模拟在100万35岁以上中国女性队列中乳腺癌筛查、诊断，治疗的过程，进行备选乳腺癌筛查策略卫生经济学效果预测和比较，结合蒙特卡洛随机模拟技术，进行各种乳腺癌筛查策略的卫生经济学效果的敏感性分析，从社会角度出发，计算在可能的意愿支付成本区间，各项乳腺癌随访策略具有成本效果（用）的可能性, 绘制成本效果（用）可接受曲线，探索针对中国不同地区经济水平和乳腺癌发病特点的最优筛查策略。

课题组按照研究方案设计完成了计划书中参数收集与估计，乳腺癌疾病过程-马尔可夫微模拟模型， 乳腺癌发病风险评估方程建立，乳腺癌筛查策略成本效益分析，蒙特卡洛随机模拟等全部内容。完成了申请书规定的研究目标，构建中国女性乳腺癌筛查策略卫生经济学评价模型，预测各种乳腺癌筛查策略的卫生经济学效果，探索我国现有经济水平确定最优乳腺癌筛查策略。另外还包括乳腺癌风险评估软件、单纯临床体检，乳房钼靶射线及乳房超声检查等方法在乳腺癌筛查项目中的效果分析，及其在延长乳腺癌患者生存期方面的效果预测。在大的研究方向上，未作调整。增加了研究的内容，作为研究的支撑；包括系统文献综述乳腺钼靶和乳腺B超在亚洲女性乳腺癌筛查中的效果，应用社区女性乳腺癌筛查数据库获取了社区乳腺癌筛查初期效果数据，对乳腺癌筛查项目在“真实世界”的效果进行了初步分析，为进一步开展乳腺癌社区筛查体系建设研究奠定基础。.本研究通过模型模拟推荐优势可选策略，推荐年龄段35-59岁人群为我国女性的优势筛查年龄段，符合我国女性乳腺癌年龄别发病趋势的特点；推荐每2年一次的筛查策略优于每3年一次的筛查策略。优势筛查策略中均包含乳腺B超的筛查方法，提示在我国当前的乳腺癌发病水平以及经济水平下，乳腺B超筛查具有一定的指导价值，但单纯采用B超或钼靶进行筛查的方法可能不在成本可接受范围之内。这些结论对于符合我国国情的乳腺癌筛查有非常强的指导作用。.本研究在现场调查中关注真实世界数据，提示无论是大规模的随机对照试验，还是专题开展的乳腺癌筛查研究专项，以及基于此而进行的模型模拟的结果很难在真实世界中重现，因为乳腺癌筛查工作在“真实世界”的开展，除了筛查策略方面，还涉及到更多复杂的项目系统运行问题，研究组将在这方面深入开展研究。研究并无乳腺癌风险评估模型提高筛查效果的明确证据，但是该模型必然能提高广大女性对于乳腺健康和乳腺癌筛查的关注度，理论上会优化乳腺癌筛查工作的效果。