中文摘要：

目的选择恰当的均衡性检验方法,探讨均衡性检验的必要性。方法采用Monte Carlo模拟比较假设检验和标准化差异检验基线资料均衡性的能力。结果随机化研究中,大样本情况下出现不均衡的可能性很小;而在样本量相对较小的情况下,很难保证基线资料的均衡性。结论标准化差异是一种有效的均衡性检验方法。非随机化研究中,均衡性检验必不可少;随机化研究中,样本相对较小情况下也需要作均衡性检验。

英文摘要：

Objective The choose the suitable method of balance test and discuss the necessity of balance test.Methods Monte Carlo simulations were adopted to compare standardized difference and hypothesis test was used to assess balance in baseline variables.Results In small randomized trials,one was more likely to observe systematic differences in baseline variables between groups than in large randomized trials.Conclusion Standardized difference should be recommended in practice.Balance tests are necessary in non-randomized trials and small randomized trials.