摘要：This paper combines the theory of teams and data envelopment analysis（DEA） to design a mechanism to optimally allocate resources in public healthcare. A statutory authority and the public hospitals under its governance are interpreted as a team, the members of which seek to operate efficiently under the shared institutional constraint that public healthcare is a public good. The individual public hospital exploits DEA to maximize own-payoff, subject to the team-condition that the payoff of each other public hospital does not fall and thereby subtract from the external effects created by the public supply of healthcare. The resulting team-DEA solution, which is shown to be both an individuallyefficient and team-satisficing equilibrium and to be computable in terms of a convergent algorithm, can then be applied by the authority to determine the optimal allocation of resources in public healthcare.A case based on Chinese data is presented to illustrate the team-DEA model＇s ready operationalization and computation.

摘要：目的比较四药联用(四联)和三药联用(三联)的抗逆转录病毒疗法作为HIV感染者一线治疗的疗效和安全性,并探讨现有的临床试验对临床实践及未来研究的启示。研究设计随机对照试验的系统综述和meta分析。数据来源系统检索了PubMed、EMBASE、CENTRAL、Web of Science及Allied health Literature数据库2001年3月至2016年12月期间发表的研究(其中PubMed和EMBASE的检索更新至2018年6月),以及合格研究和相关综述的参考文献。入选标准比较四联与三联抗逆转录病毒疗法作为HIV感染者一线治疗的作用,并评价了至少一种疗效或安全性结局的随机对照试验。分析方法对检测不到的HIV-1 RNA、CD4 T细胞计数、病毒学失败、新发艾滋病事件、死亡和严重不良反应等结局的资料采用随机效应模型进行meta分析。结果共找到12项合格研究,包括4251例HIV感染者。四联和三联抗逆转录病毒疗法对各个疗效和安全性结局的影响均类似,且没有一个总体效应估计值的点值落在支持四联疗法的一侧。与三联疗法相比,四联疗法组出现不同结局的相对危险度及其95%可信区间如下:检测不到的HIV-1 RNA,0.99(0.93~1.05);病毒学失败,1.00(0.90~1.11);新发艾滋病事件,1.17(0.84~1.63);死亡,1.23(0.74~2.05);严重不良反应,1.09(0.89~1.33)。两组CD4 T细胞计数增加的平均差值为-19.55个细胞/μl(-43.02~3.92)。无论联合抗逆转录病毒疗法的具体方案如何,上述2组比较的结果大体不变,且在所有亚组和敏感性分析中均变化不大。结论在HIV感染者的一线治疗中,四联抗逆转录病毒疗法的效果并不优于三联疗法。这一发现为当前推荐三联疗法作为一线治疗的指南提供了支持。除非对现有证据的系统综述显示确有必要,否则不应该再开展新的临床试验去比较四联与三联疗法。当然,本系统综述并不排除当有新的抗逆转录病毒药物种类出现时包含该类药物的四联疗法比三联疗法更好的可能性。