When there is a statistical dependency between when the outcome occurs, and when the observation period ends, this can introduce bias. For example, if an outcome is likely to lead to death and therefore censoring this could introduce bias. We don't have a good diagnostic for this.
When there is a statistical dependency between when the outcome occurs, and when the observation period ends, this can introduce bias. For example, if an outcome is likely to lead to death and therefore censoring this could introduce bias. We don't have a good diagnostic for this.