摘要
目的探讨不同的统计学方法在股骨头坏死随访资料疗效分析中的应用。方法利用Excel2003和SAS6.12软件包,运用CMHχ2检验、Ridit分析、两组等级资料的秩和检验和多分类有序反应变量的Logistic回归,对现有文献Harris评分资料的近期疗效评价进行统计分析。结果四种不同的统计分析方法皆能分析出两组优良率的差别有统计学差异,且其P值的趋势大小是一致的。结论四种方法皆适合于随访资料的疗效分析,只不过不同作者有不同的选择而已。
Objective To explore different statistical methods to analyze the effect on osteonecrosis of the femoral head follow up data. Methods CMH Х^2 test , Ridit analysis , Rank sum test in two groups and multicategory ordinal independent variable logistic regression analysis were used to make corresponding analysis of the effect of follow up data on osteonecrosis of the femoral head in certain current literature. Results The statistical significance of the effectiveness between two groups could be calculated by means of four different statistical methods, and the trends of P value in above 4 methods were congruoue as checked here by us. Conclusion Four kinds of statistical methods are all suitable in the analysis of the effect of follow-up data . Different authors might select their own preferable method.
出处
《同济大学学报（医学版）》
CAS
2007年第5期98-101,共4页
Journal of Tongji University(Medical Science)
关键词
股骨头坏死
统计学
随访
osteonecrosis of the femoral head
statistics
follow-up
