目的：探讨群随机试验在艾滋病干预研究中的应用。方法：以某艾滋病干预项目“女性性工作者100％安全套推广”为例，探讨群随机试验在艾滋病干预研究中的适用性。结果：通过群随机试验研究调整的χ检验得到干预组最后一次安全套未使用率与对照组没有统计学意义（χ2＝1．20，P＞0．05）的结果，通过普通的个体对照试验四格表检验得出两者有统计学意义（χ2＝6．47，P＜0．05）的结果；这是由于女性性工作者个体之间的相互影响，使得数据存在相关性，由此产生了方差膨胀，因此通过普通的个体对照试验卡方检验产生了分析偏性，通过调整的χ卡方检验得到了相反的结论。结论：通过群随机试验开展艾滋病干预措施效果的分析和评价是可行的，在分析时要调整方差膨胀对分析结果的影响。% Objective:To describe the design and application of cluster randomization trial in HIV/AIDS intervention programs .Methods:Taking an HIV/AIDS intervention project of '100% female condoms promo‐tion'as an example ,the feasibility of cluster randomized trial in the application of HIV/AIDS intervention programs was tested .Results:Through the adjusted Chi‐Square method in cluster randomization trial ,the rate of not using condom in the last time intercourse with client had no significant differences (χ2 = 1 .20 ,P>0 .05) between intervention group and control group ,while it had significant difference (χ2 =6 .47 , P<0 .05) through the common Chi‐Square method in the individual randomized controlled trial .As the correla‐tion exists among the individuals of cluster ,there produced the variance inflation ,w hich caused the analysis bias w hen using the common Chi‐Square method .Conclusions :It is feasible that cluster randomization trial is used in HIV/AIDS intervention programs .
展开▼
