We have now shown that the assumption of similar mortality rates in and out of the study is incorrect, at least for the Jamaican SS disease population. We therefore have to try something else, as these early deaths will impact on our survival figures.