|
Abstract:
|
With the advance of medicine, many drugs/treatments become available to treat patients. On the one hand, combination treatments have been developed to treat severe diseases such as cancer. On the other hand, polydrug use (i.e., using more than one drug at a time) may cause severe side effect. The observational data such as electronic health records may provide very useful information for drug interactions. In this article we develop statistical methods for evaluating the causal effect and causal interactions of two drugs. In particular, we propose a marginal structural model to assess the causal interaction of two drugs by controlling confounding variables. The joint effect of the two treatments are assessed using the weighted likelihood approach with weights being the inverse probability of treatment assigned. Simulation studies were conducted to examine the performances of the proposed method. Case study was conducted to investigate the joint effect of antecedent statins and opioids use on biomarkers for COVID-19 hospitalized patients.
|
