The Cox regression model has been widely used for the analysis of time to event data with their associated risk factors, it assumes a constant hazard ratio over time and that the risk factors are independent of time. When the assumptions are violated, the estimates of the hazard ratio of the Cox regression estimates of the hazard ratios becomes misleading. In this study, we use a modified Cox regression model that incorporates time dependent covariate which measures the interaction of exposure with time.
Birth interval between marriage and first birth for the ever married women after marriage, taken from NDHS 2013 women data is fitted using the Cox regression model with time varying covariates due to the failure of existence of proportionality assumption. This model performs better compared to Cox regression model.