Highest In-Hospital Glucose Measurements are Associated With Neurological Outcomes After Intracerebral Hemorrhage

Background and Purpose: The relationship between in-hospital hyperglycemia and neurological outcome after intracerebral hemorrhage (ICH) is not well studied. Methods: We analyzed the relationships between pre-hospital and hospital variables including highest in-hospital glucose (HIHGLC) and discharge Glasgow Coma Scale (GCS), discharge Modified Rankin Scale (MRS) and 3-month MRS using a single-institution cohort of ICH patients between 2013 and 2015. Results: There were 106 patients in our sample. Mean HIHGLC was 154 +/- 58 mg/dL for patients with discharge GCS of 15 and 180 +/- 57 mg/dL for patients with GCS < 15; 146 +/- 55 mg/dL for patients with discharge MRS 0-3 and 175 +/- 58 mg/dL for patients with discharge MRS 4-6; and 149 +/- 52 mg/dL for patients with 3-month MRS of 0-3 and 166 +/- 61 mg/dL for patients with 3-month MRS of 4-6. On univariate analysis, discharge GCS was associated with HIHGLC (P = .01), age (P = .006), ICH volume (P = .008), and length of stay (LOS) (P = .01); discharge MRS was associated with HIHGLC (P < .001), age (P < .001), premorbid MRS (P = .046), ICH volume (P < .001), and LOS (P < .001); and 3-month MRS was associated with HIHGLC (P = .006), discharge MRS (P < .001), age (P = .001), sex (P = .002), ICH volume (P = .03), and length of stay (P = .004). On multivariate analysis, discharge GCS only had a significant relationship with ICH volume (odds ratio [OR] .949, .927-.971); discharge MRS had a significant relationship with age (OR 1.043, 1.009-1.079), premorbid MRS (OR 2.622, 1.144-6.011), and ICH volume (OR 1.047, 1.003-1.093); and 3-month MRS only had a significant relationship with age (OR 1.039, 1.010-1.069). Conclusions: The relationship between in-hospital hyperglycemia and neurological outcomes in ICH patients was meaningful on univariate, but not multivariate, analysis. Glucose control after ICH is important.