In a simulated cohort, we estimated the effect of a treatment of interest vs a comparator among those who switched to the treatment of interest using the originally proposed time-conditional propensity score (TCPS) matching, standardized morbidity ratio weighting (SMRW), disease risk scores (DRS), and several alternative propensity score matching approaches. For each analytic method, we compared the average RR (across 2000 replicates) to the known risk ratio (RR) of 1.00.
SMRW and DRS yielded unbiased results (RR = 0.998 and 0.997, respectively). TCPS matching with replacement was also unbiased (RR = 0.999). TCPS matching without replacement was unbiased when matches were identified starting with patients with the shortest treatment history as initially proposed (RR = 0.999), but it resulted in very slight bias (RR = 0.983) when starting with patients with the longest treatment history. Similarly, creating a match pool without replacement starting with patients with the shortest treatment history yielded an unbiased estimate (RR = 0.997), but matching with the longest treatment history first resulted in substantial bias (RR = 0.903). The most biased strategy was matching after selecting one random comparator observation per individual that continued on the comparator (RR = 0.802).
Multiple analytic methods can estimate treatment effects without bias in a PNU cohort. Still, researchers should be wary of introducing bias when selecting controls for complex matching strategies beyond the initially proposed TCPS.
在模拟队列中,我们使用最初提出的时间条件倾向评分(TCPS)匹配来估计感兴趣的治疗与比较者之间的效果,标准化发病率加权(SMRW),疾病风险评分(DRS),以及几种替代的倾向得分匹配方法。对于每种分析方法,我们将平均RR(2000次重复)与已知风险比(RR)1.00进行了比较.
SMRW和DRS产生无偏结果(RR分别为0.998和0.997)。与替换匹配的TCPS也是无偏的(RR=0.999)。当从最初提出的治疗病史最短的患者开始确定匹配时,不进行替换的TCPS匹配是无偏见的(RR=0.999)。但当从治疗历史最长的患者开始时,会导致非常轻微的偏倚(RR=0.983).同样,从治疗史最短的患者开始创建不更换的匹配池产生了无偏估计(RR=0.997),但与最长的治疗历史匹配首先会导致实质性偏差(RR=0.903).最偏倚的策略是在每个个体选择一个随机比较器观测值并继续在比较器上进行匹配(RR=0.802)。
多种分析方法可以在PNU队列中无偏倚地估计治疗效果。尽管如此,研究人员在为最初提出的TCPS之外的复杂匹配策略选择控件时应警惕引入偏差.