Moussa M A
Faculty of Medicine, Kuwait University.
Methods Inf Med. 1990 Jul;29(3):243-6.
Various approaches are considered for adjustment of clinical trial size for patient noncompliance. Such approaches either model the effect of noncompliance through comparison of two survival distributions or two simple proportions. Models that allow for variation of noncompliance and event rates between time intervals are also considered. The approach that models the noncompliance adjustment on the basis of survival functions is conservative and hence requires larger sample size. The model to be selected for noncompliance adjustment depends upon available estimates of noncompliance and event rate patterns.
针对患者不依从情况调整临床试验规模，人们考虑了各种方法。此类方法要么通过比较两种生存分布或两个简单比例来模拟不依从的影响。还考虑了允许不依从和事件发生率在不同时间间隔之间变化的模型。基于生存函数对不依从调整进行建模的方法较为保守，因此需要更大的样本量。选择用于不依从调整的模型取决于不依从和事件发生率模式的可用估计值。
