• Joshua VanDerWerf, Donna Kurowski, James Siegler, Taneeta Ganguly, and Brett Cucchiara.
• Department of Neurology, Duke University, Duke University Medical Center, 3116 N Duke St, Durham, NC, USA. joshua.vanderwerf@duke.edu.
• Neurocrit Care. 2018 Aug 1; 29 (1): 40-46.

BackgroundHematoma expansion (HE) occurs in 1/3 of ICH patients and is associated with poor outcome. Intra-hematomal hypodensity (IHH) on CT has been reported to predict HE, as has the "BRAIN" score. We sought to assess the predictive value of these markers alone and in combination.MethodsWe performed a retrospective single-center study of ICH patients with CT < 6 h from onset. Two blinded neurologists assessed IHH on initial CT. Two HE definitions were examined: > 6 ml and > 6 ml or > 33%. Multivariable logistic regression was used to determine the relationship between IHH and HE. Predictive value of the BRAIN score alone and integrated with IHH was assessed.ResultsIn 122 included patients, median ICH volume was 13 ml, median time to CT 2.0 h; HE > 6 ml occurred in 31% and > 6 ml/> 33% in 43% of subjects. IHH were identified in 61% of patients with moderate inter-rater agreement (κ = 0.59). In multivariable analysis, IHH was associated with HE using > 6 ml definition (OR 8.3, 95% CI, 2.6-32.8, P < 0.001) but not using the > 6 ml/> 33% definition (OR 1.9, 95% CI 0.84-4.3, P = 0.12). Rate of HE (> 6 ml) increased across increasing BRAIN score quartiles (Q1:11%, Q2:23%, Q3:43%, Q4:57%, P for trend < 0.001). Rate of HE > 6 ml in patients with BRAIN score ≥ 10 and IHH was 55%, with either alone was 33%, and with neither was 3%.ConclusionsCombining IHH on non-contrast CT and a simple clinical BRAIN score is a potentially powerful way to predict those patients at very high and very low risk of HE.

21 keywords...

hide…