Ratio of the clinical SCC. The clinical sickle cell crisis rate ratio was calculated in all patients (A), patients receiving HU treatment at baseline (B), and patients not receiving HU treatment at baseline (C). The analysis of the sum of events for each subject was performed using Poisson regression (Proc Genmod). The model statement included covariates for either treatment (all doses vs placebo) or for the 4 dose levels vs placebo. Each analysis was adjusted for disease severity at baseline, the subject’s age group, and the subject’s use of HU at the time of the baseline evaluation. The model was standardized to an annual rate, using a SAS offset variable. Each table shows the estimate, its standard error, the risk ratio, and the 95% CI for the risk ratio, which was done by exponentiation of the lower and upper CIs for the treatment β coefficients.