Admission White Blood Cell Count and Neutrophil-to-Lymphocyte Ratio for Early Risk Stratification after Spontaneous Intracerebral Hemorrhage: A Retrospective Cohort Study.
To evaluate the adjusted associations and incremental prognostic value of admission white blood cell count (WBC) and neutrophil-to-lymphocyte ratio (NLR) for unfavorable discharge outcome after spontaneous intracerebral hemorrhage (sICH).
Consecutive adults with CT-confirmed sICH admitted to a tertiary academic hospital in northwestern China from August 2021 through August 2025 were retrospectively reviewed. After eligibility screening, 764 patients formed the study cohort. Admission WBC and log2-transformed NLR were analyzed as continuous exposures; the primary outcome was discharge modified Rankin Scale (mRS) score ≥ 3. Multivariable models were benchmarked against the ICH score and a clinical severity model. Internal validation used 1000 bootstrap resamples. A three-level outcome-derived WBC-NLR burden classification was treated as exploratory and evaluated with 1000 nested bootstrap resamples.
The primary complete-case analysis included 752 patients, of whom 358 (47.6%) had an unfavorable discharge outcome. Each 1 × 109/L increase in WBC was associated with higher odds of unfavorable outcome (adjusted odds ratio [OR], 1.116; 95% CI, 1.035-1.204; p = 0.004), and each doubling of NLR was associated with an adjusted OR of 2.187 (95% CI, 1.687-2.835; p < 0.001). Adding WBC and NLR increased the ICH score AUC from 0.757 to 0.848 (ΔAUC, 0.091; p < 0.001) and the clinical model AUC from 0.875 to 0.909 (ΔAUC, 0.034; p < 0.001). The optimism-corrected AUC of the full model was 0.900, with a corrected calibration slope of 0.932 and Brier score of 0.129. The nested-bootstrap-corrected AUC was 0.776 for the exploratory burden-only model and 0.899 after adding clinical covariates.
Admission WBC and NLR remained associated with unfavorable discharge outcome after adjustment and improved discrimination beyond established severity measures. They may support preliminary short-term risk assessment, whereas the outcome-derived burden categories require external validation. The discharge endpoint should not be interpreted as a substitute for 90-day functional outcome.
Consecutive adults with CT-confirmed sICH admitted to a tertiary academic hospital in northwestern China from August 2021 through August 2025 were retrospectively reviewed. After eligibility screening, 764 patients formed the study cohort. Admission WBC and log2-transformed NLR were analyzed as continuous exposures; the primary outcome was discharge modified Rankin Scale (mRS) score ≥ 3. Multivariable models were benchmarked against the ICH score and a clinical severity model. Internal validation used 1000 bootstrap resamples. A three-level outcome-derived WBC-NLR burden classification was treated as exploratory and evaluated with 1000 nested bootstrap resamples.
The primary complete-case analysis included 752 patients, of whom 358 (47.6%) had an unfavorable discharge outcome. Each 1 × 109/L increase in WBC was associated with higher odds of unfavorable outcome (adjusted odds ratio [OR], 1.116; 95% CI, 1.035-1.204; p = 0.004), and each doubling of NLR was associated with an adjusted OR of 2.187 (95% CI, 1.687-2.835; p < 0.001). Adding WBC and NLR increased the ICH score AUC from 0.757 to 0.848 (ΔAUC, 0.091; p < 0.001) and the clinical model AUC from 0.875 to 0.909 (ΔAUC, 0.034; p < 0.001). The optimism-corrected AUC of the full model was 0.900, with a corrected calibration slope of 0.932 and Brier score of 0.129. The nested-bootstrap-corrected AUC was 0.776 for the exploratory burden-only model and 0.899 after adding clinical covariates.
Admission WBC and NLR remained associated with unfavorable discharge outcome after adjustment and improved discrimination beyond established severity measures. They may support preliminary short-term risk assessment, whereas the outcome-derived burden categories require external validation. The discharge endpoint should not be interpreted as a substitute for 90-day functional outcome.