M Shakiba, Ma Mansournia, H Soori,
Volume 13, Issue 1 (6-2017)
Abstract
Standard methods for estimating exposure effects in longitudinal studies will result in biased estimates of the exposure effect in the presence of time-dependent confounders affected by past exposure.
In the present review article, we first described the assumptions required for estimating the causal effect in longitudinal studies and their structure regarding various types of exposure and confounders; then, we explained the bias of standard methods in estimating the causal effect.
Two types of bias, i.e. over-adjustment bias and selection bias, occur in estimating the effect of time-varying exposure in the presence of time-dependent confounders affected by previous exposure using standard regression analysis. Standard regression methods cannot sufficiently modify time-dependent confounders and estimate the total causal effect of the exposure.
N , , , , H Poustchi, M Yaseri,
Volume 13, Issue 4 (3-2018)
Abstract
One of the traditional methods used for the analysis of survival data is the Cox regression technique. This method calculates the conditional risk ratio. However, when the aim of the study is to estimate the effect of exposure in the total population level, using these conditional methods is not apposite. Furthermore, the hazard ratio has disadvantages of its own such as being non-collapsible, having the risk of structural selection bias and variability in time. Given the limitations, it is recommended to use the marginal hazard ratio, which estimates the average causal effect of exposure in the total population level.
This study introduces the inverse probability treatment weighting (IPTW) as a method of estimating the marginal causal effect. Finally, to illustrate IPTW method, we used Golestan Cohort Study and estimated the marginal causal effect of smoking on time to death due to the upper gastrointestinal cancer (esophageal-gastric).
