In-hospital stroke, myocardial infarction, and venous thrombosis in immune thrombocytopenia: Incidence, risk factors, and outcomes from a national cohort

Abstract Background: Immune Thrombocytopenia (ITP) is paradoxically associated with an increased risk of thrombotic events despite low platelet counts. Data on hospitalized patients—who face higher risks due to acute illness and comorbidities—remain limited. This study evaluates the incidence, risk factors, and outcomes of both arterial and venous thrombosis in hospitalized ITP patients. Prior research often focused on outpatients or single thrombotic types, leaving gaps in understanding combined thrombotic risks in acute care. We aimed to identify modifiable risk factors and quantify thrombosis-associated mortality and prolonged hospitalization. Methods: In this retrospective cohort study, we analyzed 23,046 hospitalized adults with ITP from the 2019–2020 Healthcare Cost and Utilization Project All Payer-US Inpatient Sample (NIS), stratified by thrombotic event occurrence (arterial or venous; n=1,817) versus no thrombosis (n=21,229). Thrombotic events were rigorously identified using validated ICD-10-CM codes, with exposure and outcome variables adjudicated through structured data abstraction. Baseline demographics, comorbidities (e.g., SLE, APS, solid tumors), and inpatient outcomes (mortality, LOS, total costs) were compared using Pearson's χ² tests for categorical variables and independent-sample t-tests for continuous variables. Multivariable logistic regression models, adjusted for age, sex, race, comorbidities, and hospitalization duration, estimated adjusted odds ratios (aORs) for thrombosis-associated mortality while controlling for confounding. All analyses incorporated NIS discharge weights to generate nationally representative estimates, with two-tailed p<0.05 considered statistically significant. Results: ITP patients with thrombosis were significantly older (mean 66.5 vs 60.7 years, p<0.001), more likely female (50.2% vs 43.0%, p<0.001), and showed racial differences with white majority (p=0.031) compared to those without thrombosis. In terms of comorbidities, obesity (p=0.016), systemic lupus erythematosus (SLE, p=0.034), solid cancers (p<0.001), hyperlipidemia (p<0.001), and antiphospholipid syndrome (p<0.001) were significantly more prevalent in the thrombosis group. Hematological cancers showed a trend (p=0.092). Smoking, income, steroid use, and hormone therapy showed no significant association among the groups. On univariate analysis, thrombosis was associated with significantly higher mortality (OR 2.67, 95% CI 2.25-3.18, p<0.001) and longer LOS (mean 9.3 vs 6.1 days, p<0.001). Upon adjusting for confounders using multivariate, independent predictors of thrombosis included: Older age (OR 1.02, p<0.001), Longer LOS (OR 1.03, p<0.001), Obesity (OR 1.20, p=0.004), SLE (OR 1.27, p=0.043), Antiphospholipid syndrome (OR 3.70, p<0.001), Hyperlipidemia (OR 1.26, p<0.001), Solid cancers (OR 1.45, p<0.001), Female sex was protective (OR 0.83, p<0.001). Conclusions: Thrombosis occurs in a significant proportion (7.9%) of hospitalized ITP patients and is associated with substantially worse outcomes, including higher mortality and longer hospital stays. Key independent risk factors include older age, longer hospitalization, obesity, SLE, antiphospholipid syndrome, hyperlipidemia, and solid cancers. Female sex was protective. These findings highlight the need for targeted management of modifiable thrombotic risk factors (e.g., hyperlipidemia) in ITP patients, particularly those with identified comorbidities like APS, SLE, or cancer to prevent Arterial and Venous thrombotic events.