In this tutorial, we demonstrate how inverse probability weighted Cox models can be used to account for multiple measured confounders, while concentrating inferences on the treatment or exposure effects of central interest and providing graphical summaries of these effects.
Survival analysis methods such as Cox regression can be used in infectious disease research to compare the timing of clinical events between treatment or exposure groups.Randomized clinical trials are the gold standard for estimating the effect of a treatment or exposure on a survival time endpoint. However, clinical trials are not always ethical or feasible. In that case, inference about the effect of interest might be attempted using data from observational studies. Unfortunately, observational studies may be riddled with confounding which can cast doubt on the validity of the results. In this tutorial, we demonstrate how inverse probability weighted Cox models can be used to account for multiple measured confounders, while concentrating inferences on the treatment or exposure effects of central interest and providing graphical summaries of these effects. This approach is illustrated using an example that estimates the effect of injection drug use on AIDS-free survival among HIV-infected women.
Worth the Weight: Using Inverse Probability Weighted Cox Models