Estimating Individual Treatment Effects with Time-Varying Confounders