Rate of 24-hour blood pressure decline and mortality after spontaneous intracerebral hemorrhage: A retrospective analysis with a random effects regression model

Objective: To study the effect of decline in blood pressure on mortality in patients with spontaneous intracerebral hemorrhage (ICH). Design: Retrospective chart review. Setting: University-affiliated teaching hospital. Patients: Consecutive patients admitted with spontaneous ICH over a 3-year period. Measures: Blood pressure recordings were obtained from the first 24 hrs. Patients (n = 105) with more than five blood pressure recordings and on average greater than one measurement per 2 hrs were included (mean measurements per patient = 20.3). Mean arterial pressure (MAP) recordings over the first 24 hrs after presentation were regressed on time for each patient. Each patient's MAP was calculated as a slope (change mm Hg/hr). We performed logistic regression analyses to determine the effect of MAP slope on mortality and functional outcome, adjusting for other predictive factors including Glasgow Coma Scale (GCS) score and hematoma volume. The effect of MAP slope on mortality was also evaluated in subsets of patients based on age, gender, initial GCS score, initial MAP, treatment status, hematoma volume, and presence of ventricular blood. Main Results: Mean slope of change in MAP was -2.0 mm Hg/hr (± 1.9, range -8.5 to +0.6). The slope of MAP (faster rate of decline) within the first 24 hrs was significantly associated with higher mortality (p = .04), independent of initial GCS score and hematoma volume. In subgroup analyses, MAP slope was significantly associated with mortality in men (p = .08), patients with hematoma volume <50 mm³ (p = .08), initial MAP ≤146 mm Hg (p = .006), and those with initial GCS score ≥10 (p = .07). MAP slope did not predict functional outcome among survivors. Conclusions: A rapid decline in MAP within 24 hrs after presentation is independently associated with increased mortality in patients with ICH. A large, prospective, randomized trial is required to confirm these findings.