A note on competing risks in survival data analysis