Objective: Most of the claims file fit naturally in to a panel format since there are data on cost and utilization at multiple intervals such as months or years. In this paper we describe a method that can apply to longitudinal data when the outcome variable is censored. Method: It has been proved that inverse probability weighted random effect models produce consistent estimators with censored outcome variable. We briefly summarize how this consistency is achieved and how we can calculate the standard errors. Moreover, we described how to write the commands in statistical software programs to estimate relatively complicated and advanced formulas to achieve consistency. Results: We calculated two standard errors: unadjusted to censoring bias and adjusted to censoring bias for inverse probability weighted random effect models. We showed that adjusted standard errors always equal or less than unadjusted standard errors. Conclusion: One of the main difficulties to apply inverse probability of weighted random effect models is to estimate adjusted standard errors. Due to complications, in outcomes research, these models are not widely used. This paper shows that most of the times unadjusted standard errors would be enough to make strong conclusions about the effect of our variables.