Glasgow Coma Scale Predicts 30-Day Mortality After ICH

Daniel M. Keller, PhD

June 06, 2012

June 6, 2012 (Lisbon, Portugal) — The Glasgow Coma Scale (GCS) score was highly predictive of 30-day mortality after acute intracerebral hemorrhage (ICH), based on a comparison with 3 other prognostic scoring systems in a large cohort of patients with ICH.

Adrian Parry-Jones, MBChB, PhD, honorary specialist registrar in neurology in the Brain Injury Research Group at the Salford Royal NHS Foundation Trust in Salford, United Kingdom, concluded that the GCS performs as well and is easier to do. He presented his findings here at the XXI European Stroke Conference (ESC).

Dr. Parry-Jones concluded that all 3 ICH scores were "highly predictive" in this cohort and "the GCS alone seemed to have a similar predictive ability... and given that the GCS alone is much simpler to apply, this may be preferred when prognosticating."

Poor Prognosis

ICH has a very poor prognosis compared with other forms of stroke. Many prognostic tests have been developed to estimate the risk for a poor outcome, information that may be very valuable to patients and families.

The researchers studied how well published prognostic tests performed and whether they were any more informative than readily available clinical measures. The Manchester Neurosciences Center in the United Kingdom sees about 450 cases of ICH annually. The study population comprised all ICH referrals to the center between January 2008 and October 2010. Survival status was determined from a national statistical database in October 2011 such that all patients had a minimum of 1 year of follow-up.

The final analysis included 1181 cases. At baseline, patients had a median age of 73.4 years, 52.7% were male, and the GCS score at referral was 13 (interquartile range, 8 - 15; range, 3 - 15). The median ICH volume was 24.8 mL, and intraventricular hemorrhage (IVH) was present in 37.6% of cases.

The investigators retrospectively used the first computed tomographic brain scan after presentation to determine ICH volume and location and to identify IVH. They calculated scores for 3 prognostic tests — the ICH score, the ICH grading scale (ICH-GS), and the modified ICH (MICH) score — using clinical, age, and stroke origin, location, and volume variables. The GCS is 1 of the variables from which the ICH, ICH-GS, and MICH scores are calculated.

Mortality was 41.3% at 30 days and 52.4% at 1 year. By log-rank test, all 3 scores and the GCS effectively discriminated cases according to mortality risk (P < .001).

The investigators generated receiver-operating characteristic (ROC) curves for 30-day mortality for each of the tests. The area under the ROC curve (AUC) is a measure of how well a variable can distinguish between 2 groups, in this case, between a prognosis of alive or dead.

"The area under the curve statistic is essentially the probability that if you're given 2 patients, 1 who will survive to 30 days and 1 who won't, that the score will generate a higher probability of death for the patient who goes on to die at 30 days," Dr. Parry-Jones explained. "So a perfect score would have an area under the curve of 1, and 1 which didn't discriminate at all would have an area under the curve of .5."

All 3 variants of the ICH score gave very similar AUCs; the differences between the methods were not statistically significant. "Interestingly, when we did the same analysis for just the Glasgow Coma Scale on its own, we found that the area under the curve statistic was very similar to the prognostic scores themselves," Dr. Parry-Jones reported.

Table. AUC by Score

Score AUC (95% Confidence Interval)
ICH 0.863 (0.841 - 0.885)
ICH-GS 0.872 (0.851 - 0.893)
MICH 0.859 (0.837 - 0.881)
GCS 0.871 (0.849 - 0.892)


Compared with the ICH scores or its variants and with the GCS, the ICH volume (AUC, 0.776; 95% confidence interval [CI], 0.749 - 0.803) and especially age (AUC, 0.570; 95% CI, 0.537 - 0.603) were much less predictive of mortality.

Score in 3 Minutes or Less

Session moderator Thorsten Steiner, MD, PhD, chairman of the Department of Neurology at Klinikum Frankfurt Höchst in Frankfurt, Germany, and a faculty member at the University Clinic, Heidelberg, commented to Medscape Medical News that he found it interesting that the GCS turned out to be very close to the other prognostic scores in its ability to predict 30-day mortality.

"If this is true, this would be a very interesting measurement because the Glasgow Coma Scale is such an interesting and easy to perform measurement." He said that he would like to know more details about the study cohort and that the GCS as a prognostic tool still needs to be validated in a prospective study.

"It was interesting to see that it performs as well as 3 other scores that go more into detail, for example, concerning the location of the bleeding. One score was even differentiating between infra- and supratentorial [location of bleeding], and the Glasgow Coma Scale does nothing of that and still is as good as the other scores," he said. "If it turns out to be as good as we have seen it here, that would be wonderful because then we have a very easy scale to make some prognostic statements on an ICH patient."

One can come up with a GCS score within 3 minutes or less, but for the other scores, brain imaging is necessary "because you need to decide is or is there no IVH, and you certainly need to decide about the size of the intracerebral or intraparenchymal bleeding to decide whether it has reached or has not reached a certain size or certain volume," Dr. Steiner noted.

Imaging will always be done at some point, but looking toward the future, he said, "If everything needs to be done fast, we would like to have very early the information on the prognosis, but certainly, it will take a long time until we can rely on only 1 scale for major decisions in an individual patient."

He also said that, if validated, the GCS could prove useful as a selection criterion for future trials, "particularly when it comes down to do we operate on a patient, do we give hemostatic agents, do we lower pressure, and so forth."

The study received no commercial funding. Dr. Parry-Jones has disclosed no relevant financial relationships. Dr. Steiner, who was not involved in the study, has disclosed no relevant financial relationships.

XXI European Stroke Conference (ESC). Presented May 23, 2012.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: