I 新农合方案调整对农村老年人健康绩效的影响研究 —基于宁夏医改样本县的面板数据分析 摘要 目的：本研究利用“创新支付制度，提高卫生效益”项目的调查数据，分别从主观 健康和客观健康两个维度来分析新农合方案调整对试点县老年人健康状况的影响，从总 体水平和不同收入水平两个方面计算新农合政策调整对老年人健康的净效应，进而分析 政策调整对老年人健康绩效的影响，为促进宁夏统筹城乡医疗保障制度健康持续发展提 供政策依据。 方法：运用样本县 2009、2011 、2012 和 2015 年四年的面板数据，以样本县常驻 （居住 6 个月及以上）60 岁及以上老年人为研究对象，在倾向得分匹配的基础上运用倍 差法模型分析新农合方案调整对农村老年人健康状况的净效应。用多阶段 DID-Logit 模 型进一步分析宁夏医改试点项目在实施过程中对不同收入老年人健康水平的递进性影 响。通过计算边际效应分析新农合政策调整对老年人的健康绩效。 结果：（1）四次调查，干预组和对照组老年人的两周患病率均有降低趋势，干预 组老年人在 2009、2011、2012 和 2015 年的两周患病率分别为 41.7%、27.6%、29.3%和 26.0%，对照组分别为 31.4%、30.7%、25.5%和 22.0%，至 2015 年，干预组和对照组比 干预前分别降低了 15.7%和 9.4%。慢性病患病率在干预后的第一年两组均有所下降， 2012 年开始上升，至 2015 年，干预组与对照组较干预前分别上升了 9.2%和 12.8%。自 评健康方面，对照组的自评健康良好率呈不稳定性波动，干预组呈稳定上升趋势。（2） 双重差分结果显示，从两周患病率看，2011 年、2012 年和 2015 年，政策实施后干预组 老年人的两周患病率分别降低了 11.1%、5.5%和 5.7%，即政策的实施对干预组老年人的 两周患病率起到持续性降低的作用，且见效快。从慢性病患病率看，在政策干预一年后 和三年后，干预组老年人的慢性病患病率均有所降低，但效果并不显著，干预六年后，宁夏医科大学硕士学位报告 中文摘要 II 慢性病患病率有升高，但依然不显著，即政策的调整对干预组老年人的慢性病患病率无 显著影响。此外，新农合方案调整对干预组老年人的自评健康在政策实施初期效果不明 显，在实施三年后和六年后，干预组老年人的自评健康良好率比对照组老年人分别增加 了 8.8%和 9.2%，干预效果具有一定的滞后性。（3）对不同收入组老年人而言，新农合 支付制度改革政策的健康绩效主要体现在中低收入组、中等收入组和高收入组老年人， 在低收入组和中高收入组老年人中未体现。新农合方案调整对中低收入组老年人的健康 改善主要体现在对主观健康的影响，政策调整后，该收入组老年人的自评健康良好率在 政策实施一年、三年和六年后分别增加了 15.3%、25.0%和 14.9%。新农合方案调整对中 等收入组老年人的健康改善主要体现在对客观健康的影响，政策调整后，该收入组老年 人的两周患病率在政策实施一年、三年和六年后分别降低了 18.0%、17.6%和 12.4%。对 于高收入组老年人，新农合方案调整后，该组老年人的两周患病率仅在项目实施一年后 降低了 14.0%，此后不再显著。但政策的调整使得该组老年人自评健康良好率在政策实 施三年和六年后分别增加了 16.8%和 14.5%。 结论：（1）新农合政策调整后干预县老年人的两周患病率降低，政策效应快，且具有 持续性。（2）新农合政策调整对干预县老年人慢性病患病率的降低作用不明显，政策 效应不显著。（3）新农合政策的调整改善了干预县老年人的自评健康，但改善作用在 干预三年后逐渐显现，反映了政策调整对主观健康的滞后作用。（4）对于不同收入层 次老年人而言，新农合政策的健康绩效主要体现在中低收入、中等收入和高收入组老年 人中，在低收入组和中高收入组老年人中未体现。 关键词 新型农村合作医疗，农村老年人，健康绩效宁夏医科大学硕士学位报告 英文摘要 III Study on the Effect of the New Rural Cooperative Medical Scheme Adjustment on the Health Performance of Rural Elderly —Based on Panel Data in Sample County of Ningxia Medical Reform ABSTRACT Objective To investigate the impact of the New Rural Cooperative Medical Scheme (NCMS) adjustment on the health status of the elderly in the pilot counties, adopting the data of the “Innovative Payment System and Improving Health Benefits” project to analyze subjective health and objective health of the elderly, and also evaluate the health net effect of the policy adjustment from the overall and different income levels, so as analyzing the impact and its magnitude of the policy adjustment on the health performance of the elderly, and providing the useful policy evidence for promoting the sustained development by medical security policies of the overall plans on urban and rural in Ningxia. Methods Using the panel data of four counties in 2009, 2011, 2012 and 2015 years, the elderly in the sample counties （residential 6 months and above） aged 60 and older were studied. Based on the matching of propensity scores, the Difference-in-Difference model (DID) was used to analyze the net effect of the NCMS on the health status of rural elderly. The multi-stage DID-Logit model has been applied into the study to further analyzeits dynamic changes of its multiple stages and the progressive effects on the health of different incomes in the implementation process, and the health performance of the NCMS adjustment for the elderly analyzed by calculating the marginal effect. Results （1）During the four survey years, the two-week prevalence rate of the elderly in the intervention group and the control group decreased. The prevalence of the elderly宁夏医科大学硕士学位报告 英文摘要 IV among the intervention group in 2009, 2011, 2012 and 2015 were 41.7% ,27.6% ,29.3% and 26.0% , the control group were 31.4%, 30.7%, 25.5%, and 22.0%, respectively. By 2015, the two-week prevalence rate of the intervention group and the control group were reduced by 15.7% and 9.4% than before. The prevalence of chronic diseases decreased in both groups after the intervention, and began to rise in 2012. the intervention group and the control group increased by 9.2% and 12.8% in 2015. respectively, compared with the baseline survey. In terms of self-evaluation health, the self-evaluation health rate of the county was unstable, and the intervention group showed a steady upward trend. （2）The DID results showed that the implementation of the policy reduced the prevalence of the two-week elderly in the project counties by 11.1%, 5.5% and 5.7% respectively，in 2011, 2012 and 2015. That is to say, the implementation of the policy has a sustained reduction in the two-week prevalence rate of the elderly in the intervention group, and the policy effect is effective; In terms of the prevalence of chronic diseases, the chronic disease prevalence of the elderly in the intervention group decreased after one year and three years of policy intervention, but the effect was not significant. After six years of intervention, the prevalence of chronic diseases increased. However, it is still not significant, indicating that the adjustment of policies has no significant effect on the prevalence of chronic diseases in the intervention group. In addition, the NCMS adjustment effect on the self-rated health of the elderly in the intervention group was not obvious at the initial stage of the policy implementation. After three years and six years of implementation, the elderly in the intervention group increased by 8.8% and 9.2% respectively compared with the control group, in terms of self-rated health, the policy effect has a certain lag.（3）For the elderly in different income groups, the adjustment of the NCMS improved the health of the elderly in middle-low income group, the middle-income group and the high-income group, and had no significant effect on the low-income group and the middle-high income group. The effect of the adjustment of the NCMS on the health宁夏医科大学硕士学位报告 英文摘要 V improvement of the middle and low-income group is mainly reflected in the subjective health. As a result, the self-rated health of the group increased by 15.3%, 25.0% and 14.9% respectively one year, three years and six years after the implementation of the policy. The adjustment of the NCMS to the health of the middle-income group is mainly reflected in the objective health effects, which reduced the two-week prevalence rate of the group by 18.0%, 17.6% and 12.4% respectively one year, three years and six years after the implementation of the policy. For the high-income group of seniors, the two-week prevalence rate was reduced by 14.0% after the adjustment of the NCMS, and then no longer significant. However, as a result of the implementation of the policy, the rate of self-rated good health of the elderly in this group incre