A new Modified Intracerebral Hemorrhage Score for treatment decisions in basal ganglia hemorrhage - a randomized trial

Objectives: To develop a Modified Intracerebral Hemorrhage (MICH) score to determine optimal cut-offs for conservative treatment vs surgical intervention for basal ganglia hemorrhage and to predict outcomes. Design: Prospective randomized trial. Setting: A 1,720-bed medical center affiliated with a university. Patients: In all, 226 patients with basal ganglia hemorrhage who presented at our hospital from 2001-2005. Interventions: Group A (n = 113) underwent endoscopic surgery; group B (n = 113) underwent conservative treatment. Score differences on the Glasgow Outcome Scale and 1 -yr Barthel Index were analyzed by chi-square test and Student's t-tests. Cut-offs for MICH scoring were evaluated using receiver operating characteristic curves for calculating the Youden Index. The treatment odds ratio was analyzed by univariate, multivariate, and multiple logistic regressions. Measurements and Main Results: The optimal cut-off point for mortality was a MICH score >= in which the Youden Index is 0.66 (sensitivity, 76.3%; specificity, 89.8%; area under the receiver operating characteristic curve, 0.897). The positive and negative predictive values were 81.8% and 86.3%, respectively. The treatment odds ratio for surgical treatment was 6.87 (95% confidence interval, 3.13-14.5) at MICH scores >= 3. The best cut-off for good functional outcomes (Glasgow Outcome Scale :4 or Barthel index 55) was MICH >= 2. Conservative treatment achieved a better mean Barthel Index at MICH = 0 or 1 than surgical treatment, p < .01. At MICH scores = 3 or 4, 6-month mortality for conservative treatment was higher than for surgical treatment, p < .01 and p = .04, respectively. At MICH scores of 5, all patients died. Conclusions. MICH scoring provides a simple, reliable system for treatment decisions regarding basal ganglia hemorrhage and may accurately predict functional outcomes. Conservative treatment is recommended for basal ganglia ICH patients with low MICH scores (0, 1) to preserve neurologic function. Surgery is recommended for patients with a midlevel MICH score to obtain better functional outcomes (MICH = 2) and to reduce mortality (MICH = 3 or 4). At MICH scores = 5, there are no indications for surgery.