Prediction of intracerebral haemorrhage expansion with clinical, laboratory, pharmacologic, and noncontrast radiographic variables

Int J Stroke. 2015 Oct;10(7):1057-61. doi: 10.1111/ijs.12507. Epub 2015 Apr 28.

Abstract

Background: Hematoma expansion confers excess mortality in intracerebral haemorrhage, and is potentially preventable if at-risk patients can be identified. Contrast extravasation on initial computed tomographic angiography strongly predicts hematoma expansion but is not very sensitive, and most centers have not yet integrated computed tomographic angiography into acute intracerebral haemorrhage management. We therefore asked whether other presentation variables can predict hematoma expansion.

Methods: We searched the electronic medical records of a large integrated healthcare delivery system to identify patients with a hospitalization discharge diagnosis of intracerebral haemorrhage between the years 2008 and 2010. Hematoma expansion was defined as radiographic increase by 1/3 or by 12·5 ml within 48 h of presentation. Pre-specified patient demographic and clinical presentation variables were extracted. Stepwise multivariable logistic regression was performed to model hematoma expansion. Because some patients may have died from hematoma expansion without a second head computed tomography, we constructed a separate model including patients that died without a second head computed tomography in 48 h, hematoma expansion or death.

Results: Ninety-one of 257 patients (35%) had hematoma expansion. Antithrombotic use (odds ratio = 1·9, P = 0·04) and initial mNIHSS (modified National Institutes of Health Stroke Scale; odds ratio = 1·06, P = 0·001) were significant predictors in the hematoma expansion model (area under the Receiver-Operator Characteristics curve, AUROC = 0·6712, pseudo-r(2) = 0·0641). 163 of 343 patients (48%) had hematoma expansion or death. Age (odds ratio = 1·02, P = 0·02), initial mNIHSS (odds ratio = 1·07, P < 0·001), and initial hematoma volume (odds ratio = 1·01, P = 0·03) were significant predictors of hematoma expansion or death (AUROC = 0·7579, pseudo-r(2) = 0·1722).

Conclusion: Clinical and noncontrast radiographic variables only weakly predict hematoma expansion. Examination of other indicators, such as computed tomographic angiography contrast extravasation (the 'spot sign'), may prove more valuable in acute intracerebral haemorrhage care.

Keywords: CT scan; acute; brain bleed; cerebral haemorrhage; haemorrhage; intracerebral haemorrhage.

MeSH terms

  • Aged
  • Aged, 80 and over
  • Cerebral Hemorrhage / diagnosis*
  • Delivery of Health Care / statistics & numerical data
  • Electronic Health Records / statistics & numerical data
  • Female
  • Hematoma / diagnostic imaging*
  • Hematoma / etiology*
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • Predictive Value of Tests
  • Retrospective Studies
  • Severity of Illness Index
  • Tomography, X-Ray Computed