LAUSR

Introduction: Observational evidence from single center studies indicates that lower admission hemoglobin (Hb) levels are associated with poor outcome after spontaneous intracerebral hemorrhage (ICH). We combined data from three multicenter studies to test the hypothesis that Hb levels inversely correlate with functional outcome in ICH. Methods: We conducted a meta-analysis of individual patient data from the clinical trials ATACH-II and FAST and the multi-ethnic study ERICH. We included participants with available Hb and outcome data. We used multivariable logistic regression to test for association between admission Hb levels and 3-month dichotomized (0-3 versus 4-6) modified Rankin Scale (mRS), adjusting for the variables contained in the ICH score. We pooled study-specific estimates using inverse-variance weighted, fixed effects meta-analysis. Results: A total of 4106 ICH patients were included in the analysis. Each additional g/dL of admission Hb was associated with a 12% (OR 0.88, 95%CI 0.85-0.91; p<0.001) and 8% (OR 0.92, 95%CI 0.88-0.96; p<0.001) reduction in the odds of poor outcome in unadjusted and adjusted analyses, respectively (Table 1). Dose-response analyses indicated a linear relationship between Hb levels and poor outcome across the entire evaluated range (Figure 1, test-for-trend p<0.001). In metanalysis, there were not significant associations between Hb and ICH volume or expansion (both p>0.05). Conclusion: Lower hemoglobin levels are associated with poor outcome in ICH. Further studies of the underlying biological mechanisms are warranted. If replicated, this pathway could become an appealing target to be tested in clinical trials.