Estimating the Burden of Intracranial Hemorrhage in Persons with Hemophilia A using Administrative Claims Data


Intracranial hemorrhage (ICH) is a serious, life-threatening complication of hemophilia A (HA). Previous studies have reported an incidence rate of ICH between 2-4% in persons with hemophilia A (PwHA); however, a single episode of ICH can result in significant morbidity and mortality. Few studies have assessed the long-term consequences of ICH among PwHA. We aimed to examine the downstream clinical, healthcare resource utilization (HRU), and cost burden of ICH in PwHA.


We utilized commercial administrative claims data from US MarketScan Commercial Research Database and PharMetrics Plus Database from 01/01/06 to 12/30/18 (MarketScan data available until 9/30/18 at the time of analysis). PwHA aged ≤65y were identified using a validated claims-based HA algorithm (Lyons et al. 2018). Further, PwHA with ICH were identified as those with ≥1 claim involving ICD-9-CM/ICD-10-CM diagnosis codes for ICH. The first ICH event was identified as the index event. PwHA with ICH were required to have ≥6 months of continuous enrollment from the index event (follow-up period). Clinical conditions were identified using ICD-9-CM/ICD-10-CM diagnosis codes, and major bleeds were identified using a previously developed algorithm (Shrestha et al. 2017). Physical therapy (PT), durable medical equipment (DME) use, factor VIII (FVIII), and bypassing agents (BPAs) were identified using NDC, CPT, or other HCPCS codes. Patient demographics, clinical characteristics, HRU, and costs were descriptively examined during the follow-up period.


Of the total 4539 PwHA, 94 (2.1%) had ≥1 ICH diagnosis. Of these, 69 individuals with ≥6 months of continuous enrollment constituted the current study cohort. The majority (59.4%, n=41) had their index ICH event recorded in an ICU or inpatient setting. The mean age was 21.1y (SD±18.9) with 33 adults (range: 19-65y) and 36 children (range: 0-16y); 16 (23.2%) were aged <5y; all were male; 27 (39.1%) were in the northeast/east US; 37 (53.6%) had a Preferred Provider Organization health plan.

In the ICH cohort, 26 (37.7%) individuals had evidence of ≥1 major bleed in the 6 months following the index ICH event, while 11 (15.9%) had evidence of inhibitors (i.e. claim for a BPA). The mean Charlson Comorbidity Index score of those in the ICH cohort was 0.9 (SD±1.8); 23 (33.3%) had a diagnosis of joint/musculoskeletal conditions including arthropathy, 22 (31.9%) had chronic pain diagnosis, 12 (17.4%) had evidence of fracture, 11 (15.9%) had a mental health-related diagnosis including depression/anxiety, and 10 (14.5%) had evidence of paraplegia/hemiplegia.

In the 6 months following the index ICH event, 26 (37.7%) of the individuals had ≥1 ICU visit (mean no. of visits=0.6, SD±1.0), 23 (33.3%) had ≥1 non-ICU in-patient stay (mean=0.6, SD±1.5) and a mean length of stay of 5.5 (SD±14.9) days, 44 (63.8%) had ≥1 ER visit (mean=1.6, SD±2.3), 55 (79.7%) had ≥1 outpatient hospital visit (mean=6.2, SD±11.8), 64 (92.8%) had ≥1 office visit (mean=7.9, SD±7.4), 19 (27.5%) had evidence of DME or PT, and 47 (68.1%) had ≥1 claim for FVIII prescription/administration with an average of 5.2 (SD±11.4) claims in the post 6-month period. The average all-cause healthcare costs among PwHA with ICH was $201,595 (SD±205,388, median=$135,374) over 6 months following the ICH event, and medical costs accounted for 25.3% ($50,937) of these costs. The average healthcare costs associated with the index ICH episode were estimated to be $63,369 (SD±97,925, median=$16,813). To contextualize these findings, we estimated the average 6-month healthcare costs of PwHA without any ICH diagnosis to be $102,548 (SD±214,065, median=$40,401).

Among the 69 PwHA in the ICH cohort, 51 had continuous enrollment for 6 months prior to the index event. Of these, 31 (60.8%) had evidence of any FVIII or BPA use while 15 (29.4%) had a diagnosis of ≥1 non-ICH related major bleed in the 6 months preceding the index event.


Although rare, ICH still occurs in PwHA including younger children. ICH was associated with high comorbidity burden, health care service utilization, and costs. Due to unavailability of mortality data, the current analysis may still be an underestimate of the true burden of ICH in PwHA. Based on our findings, two out of five PwHA with ICH had no evidence of HA treatments prior to the incidence of ICH. Early management and treatment of HA may reduce the burden of ICH in this population.