Are Admission Hemoglobin and Platelet Count Independent Prognostic Markers in Spontaneous Intracerebral Hemorrhage? A Retrospective Cohort Study.
Abstract
Aim
To examine whether admission hemoglobin (HB) and platelet (PLT) levels are associated with 90-day mortality independently of established severity markers in spontaneous intracerebral hemorrhage (ICH), and whether these associations relate to hematoma burden. MATERIAL AND
Methods
We retrospectively analyzed 100 consecutive adults with predominantly hypertensive spontaneous ICH admitted to a single tertiary center between 2021 and 2025. Mortality associations were assessed using univariate and parsimonious multivariable logistic regression. Incremental prognostic value was evaluated using nested likelihood-ratio (LR) tests, pseudo-R2 changes, and supplementary receiver operating characteristic analysis. Spearman correlations assessed relationships with hematoma volume. Sensitivity analysis excluded patients receiving comfort care and additionally adjusted for surgical status.
Results
Ninety-day case-fatality was 38%. Non-survivors had lower HB (11.4 vs 12.4 g/dL; p=0.002) and PLT (171 vs 195 ×103/μL; p=0.013). Both were associated with mortality in univariate analyses, but neither showed a statistically significant independent association after adjustment for Glasgow Coma Scale (GCS) plus hematoma volume or ICH score, while GCS and ICH score remained associated. Adding HB and PLT did not significantly improve model fit (LR p=0.93 for ICH score; p=0.71 for GCS plus volume), with essentially unchanged pseudo-R2 and minimal discrimination improvement (AUC 0.870 versus 0.865). HB and PLT correlated inversely with hematoma volume (ρ=-0.38, p<0.001; ρ=-0.28, p=0.004). Sensitivity analysis yielded similar findings.
Conclusion
Admission HB and PLT were associated with mortality in unadjusted analyses but provided limited additional prognostic information beyond hematoma burden and clinical severity. Larger studies are needed to confirm these findings.