Incidence and Risk Factors for Intracranial Hemorrhage in Californians with Immune Thrombocytopenia

Intracranial hemorrhage (ICH) is thought to be an uncommon event in patients with immune thrombocytopenia (ITP), with reported cumulative incidences of approximately 1%. Most studies have drawn from single institution experiences. We determined the incidence using a large population-based administrative database.

Using the California Office of Statewide Health Planning and Development Patient Discharge Dataset, we identified all the records of Californians discharged with a diagnosis of ITP from 1991–2009. We then limited analysis to cases with ITP as the principal diagnosis. Similarly, all cases with a diagnosis of ICH (see types below) in any of the first 5 diagnosis fields were identified and merged with the ITP cohort. Univariate analysis and Cox proportional hazard modeling were then done to determine associations with ICH.

A total of 37,780 unique patients were initially identified with an ITP code, of which 13,651 were in the principal position and comprise the analysis cohort. This cohort was 60% female, 60% White, 22% Hispanic, 8% Asian, and 7% African-American with a median age of 53 years. Three hundred sixty-four cases had an ICH (cumulative incidence 2.67 %) at a median of 844 days after the first admission for ITP; 109 occurred with the first hospitalization for ITP. The median age of patients with ICH was 67 years. One hundred thirty-five cases with ICH had a splenectomy a median of 296 days prior to their hemorrhage. Of the ICH cases, 65 were known to have died (death statistics available only through 2007); 41 of these deaths were during the admission for the intracranial hemorrhage.

The cumulative incidence of ICH in this large cohort of Californians with ITP was 2.67%, higher than that reported from another population-based study with a smaller number of patients, and from single institution studies. The median age in this study is older than in other reports. The death rate for ICH hospitalization was 11.3% (41/364). Increasing age and certain medical co-morbidities were associated with increased hazard of ICH, as was diagnosis of lupus. Splenectomy was not associated with either increased or decreased hazard, but as this procedure was likely performed only in more severely affected but otherwise healthier patients this result is difficult to interpret. The finding of increased hazard of ICH in Hispanics and Asians with ITP is novel and unexplained. Intracranial hemorrhage may be a more common complication of ITP than previously appreciated.

Wun:Daiichi Sankyo Company, Ltd. and Eli Lilly and Company: Research Funding.