Thrombocytopenia and Anemia Among Patients with Cerebral Palsy: A Big-Data Analysis

Title: Thrombocytopenia and Anemia among patients with Cerebral Palsy: A big-data analysis

Introduction : The incidence of thrombocytopenia and anemia in patients with cerebral palsy (CP) is not well reported in the medical literature . Patients with CP, particularly those with the most severe phenotypes, may be repeatedly exposed to medications and infections associated with blood dyscrasias. The aim of this study was to assess the the prevalence of thrombocytopeniaand anemia in patients with CP comparing it to the general population. We also compared the prevalence of hospitalizations between patients with CP, with and without these blood dyscrasias.

Methods: Utilizing a commercial database (Explorys Inc, IBM), we identified a cohort of patients diagnosed with CP based on the Systematized Nomenclature of Medicine-Clinical Terms. We calculated the overall prevalence rate of diagnosis of CP, described age, and gender-based prevalence rates of CP, and identified associated diagnoses of thrombocytopenia and anemia associated with CP. Explorys is a largereservoir of de-identified, HIPAA-compliant aggregated data of more than 63 million unique patients. First we compared the incidence of thrombocytopenia and anemia between patients with CP and the general population. To generate a cohort of patients representative of the general population, a cohort of patients with any coded disease was pooled except for CP. Moreover, relevant covariates including hospitalization, use of antiepileptic drugs (AED), infections, age, and gender were recorded and will be included in a multivariate analysis. We further sub-grouped patients with CP based on diagnoses of wheelchair dependency and AED exposure as markers of morbidity and risk for blood dyscrasias respectively.

Results: We identified 49,492,350 unique patients of whom 93,850 were patients with CP. The prevalence of thrombocytopenia was higher in patients with CP than in the general population (Table 1; 4.91% vs 1.76%, RR 2.8, p<0.0000001). The same case was also noted for anemia (24,4% vs 11,4%, RR 2.1, p<0.0000001). When comparing the baseline characteristics of the two populations it seems that patients with CP are less aged compared to the general population (prevalence of patients >65 yoa , 53% vs 28%). Moreover, the infection rate in patient with CP surpasses that of the general population (infectious disease prevalence 93% vs 77%). On subgroup analysis, wheelchair dependence,AED exposure and any infectious process in patients with CP are risk factors for both anemia and thrombocytopenia. Nonetheless, an infectious process seems to be a stronger risk factor for both thrombocytopenia and anemia compared to AED exposure and wheelchair dependence (WD). More specifically the RR for thrombocytopenia in patients with CP is 7.4 compared to 4.8 and 2.4 for AED exposure and WD respectively. The same pattern is noticed for anemia (RR 4.7 vs 2.6 vs 1.8, p<0.001).Finally, we compared the total hospitalization rate in the past 5 years between patients with CP and with or withoutthrombocytopenia, anemia and cytopenia (defined as thrombocytopenia and anemia). Higher hospitalizations rates were found for patients with CP and with any of the above defined blood cell dyscrasias (Relative risk for hospitalization for patients with CP and thrombocytopenia/anemia/cytopenia vs without, RR 10/9/12.5, P<0.001).

Discussion: Our study using large, aggregated data from Explorys demonstrates there is a higher prevalence of thrombocytopenia and anemia in patients with CP. In our cohort, patients with CP with thrombocytopenia and/or anemia were more likely to be hospitalized. Further research is needed to confirm this observed association and to clarify the mechanism of the relationship between CP and cytopenias.