The ICH Score is the most widely used bedside grading scale for spontaneous intracerebral hemorrhage, combining five variables available within minutes of a head CT into a single number that correlates strongly with 30-day mortality. It was designed by Hemphill and colleagues to standardize severity assessment and communication among clinicians, not to make treatment decisions in isolation.
How the ICH Score works
The ICH Score sums points from five independently weighted variables assessed at presentation: Glasgow Coma Scale (tiered 0-2 points, the strongest single predictor), ICH volume at or above 30 cm³ (1 point), intraventricular hemorrhage (1 point), infratentorial location (1 point), and age 80 or older (1 point). The total ranges from 0 to 6 and maps directly to a 30-day mortality rate observed in the original 161-patient derivation cohort published in Stroke in 2001. Each cutoff — the GCS tiers, the 30 cm³ volume threshold, and the 80-year age threshold — was chosen because it independently predicted mortality in multivariable regression during derivation, and the resulting score has since been externally validated in multiple subsequent cohorts.
Inputs and what they mean
GCS is scored from the patient's best response at presentation and is the only variable worth more than 1 point, reflecting how strongly depressed consciousness predicts poor outcome. ICH volume is typically calculated from a non-contrast CT scan using the ABC/2 method and crosses a single 30 cm³ threshold — a hemorrhage just above or below this line moves the score by a full point, so accurate volume estimation matters. IVH presence and infratentorial location are both binary radiographic findings read directly off the CT. Age uses a single 80-year cutoff rather than a continuous scale, which keeps the score simple at the cost of some precision at the margins (a 79-year-old and an 81-year-old differ by 1 point despite near-identical biological age).
Limits and edge cases
No patients in the original derivation cohort scored 6, so the ~100% mortality figure at that score is an extrapolation, not an observed rate — the calculator flags this explicitly. The ICH Score was derived and validated primarily in cohorts receiving standard aggressive care; because early withdrawal of care strongly influences observed mortality, using the score to justify withdrawing care risks becoming a self-fulfilling prophecy, a limitation widely discussed in the subsequent literature. The score also does not capture anticoagulant use, hematoma expansion risk, or comorbidities that can meaningfully change an individual patient's trajectory. It should inform — never replace — a full clinical assessment and multidisciplinary discussion of goals of care.