Severity of Brazilian sickle cell disease patients: severity scores and feasibility of the Bayesian network model use

The integration of the several clinical and laboratory dimensions and the influence of each parameter on the sickle cell disease (SCD)-related mortality is useful for predicting the phenotype of an individual. This study evaluated the feasibility of the SCD severity calculator use to measure disease severity in Brazilian patients. The study group was composed of 500 SCD patients (440 HbSS and 60 HbSC) diagnosed by molecular biology. We observed a decrease in severity scores in 72 SCD patients assessed before and after the hydroxyurea (HU) use. Furthermore, the HU influenced the increase of mean corpuscular volume (MCV) and HbF concentration, and the decrease of leukocytes and total bilirubin. We found 180 (36.0%) patients with intermediate phenotype, 170 (34.0%) mild phenotype and 150 (30.0%) with severe phenotype. Patients with ages >40 years had higher mean score (0.778 +/- 0.177) than patients between 18 and 40 years (0.562 +/- 0.152) and patients between 5 and 17 years (0.322 +/- 0.145). We observe that there is a tendency of individuals with leg ulcers, avascular necrosis and cardiac complications with increasing age. Correlation analysis showed relations between severity scores with leukocytes, reticulocytes, bilirubin, lactate dehydrogenase, HbS, hemoglobin and hematocrit (p < 0.05). Several comparisons involving age groups, SCD genotype and phenotypic classification had satisfactory results and this classification will be used for future studies involving genetic polymorphisms, response to treatment with HU and oxidative stress markers in SCD.Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq