Neuroradiology · Post-reperfusion haemorrhage · 1.0

Heidelberg Bleeding Classification

A combined radiological–clinical system for classifying intracranial haemorrhage after ischaemic stroke reperfusion therapy, extending ECASS by adding remote, intraventricular, subarachnoid and subdural bleeding categories and symptomatic/asymptomatic criteria.

Purpose

Standardises the anatomic description and clinical relevance of intracranial haemorrhage occurring after ischaemic stroke treated with thrombolysis and/or endovascular therapy, for use in trial reporting, registries and clinical practice.

ClassTypeDescription
1aHI1Scattered small petechiae, no mass effect
1bHI2Confluent petechiae, no mass effect
1cPH1Haematoma within infarcted tissue, occupying <30% of the infarct, no substantive mass effect
2PH2Haematoma occupying ≥30% of the infarcted tissue, with obvious mass effect
3aRemote PHParenchymal haematoma remote from the infarcted brain tissue
3bIVHIntraventricular haemorrhage
3cSAHSubarachnoid haemorrhage
3dSDHSubdural haemorrhage

How to use it

Common mistake

Reporting only the ECASS-style HI/PH grade and omitting the class 3 subtypes (remote PH, IVH, SAH, SDH), or labelling any intracranial bleed as 'sICH' based on imaging appearance alone without applying the NIHSS-based symptomatic criteria.

Exam pearl

Only PH2 (class 2) is the anatomic category consistently linked to clinical deterioration - HI1/HI2/PH1 and isolated SAH are typically asymptomatic. When asked to define sICH under this system, anchor your answer to the NIHSS thresholds (≥4 total or ≥2 in one subcategory), not just 'any haemorrhage with worsening.'

Viva questions

What does the Heidelberg Bleeding Classification classify, and why was it developed?
It classifies intracranial haemorrhage after ischaemic stroke reperfusion therapy (thrombolysis or thrombectomy), extending the older ECASS HI/PH system by adding categories for remote parenchymal haematoma, intraventricular, subarachnoid and subdural haemorrhage, plus explicit symptomatic/asymptomatic criteria for trial reporting.
List the Class 1 subtypes and their imaging definitions.
1a (HI1): scattered small petechiae, no mass effect. 1b (HI2): confluent petechiae, no mass effect. 1c (PH1): haematoma within the infarct occupying less than 30%, no substantive mass effect.
How is Class 2 (PH2) defined, and why is it clinically important?
PH2 is a haematoma occupying 30% or more of the infarcted tissue with obvious mass effect. It is the only anatomic category consistently and independently associated with clinical deterioration and worse prognosis.
What are the four Class 3 subtypes?
3a: parenchymal haematoma remote from the infarcted tissue. 3b: intraventricular haemorrhage. 3c: subarachnoid haemorrhage. 3d: subdural haemorrhage.
How is symptomatic intracranial haemorrhage (sICH) defined under this classification?
A new intracranial haemorrhage associated with a total NIHSS increase of 4 points or more, or a 2-point increase in one NIHSS subcategory, or a haemorrhage leading to major medical or surgical intervention such as intubation, hemicraniectomy, or external ventricular drain placement, with no alternative explanation for the deterioration.
How does this system grade relatedness of a haemorrhage to clinical deterioration for trial reporting?
Class 2 (PH2) is graded as probably related when symptomatic. Classes 1b, 1c and 3 are graded as possibly related. Class 1a is graded as unlikely related.
What is a common pitfall when applying this classification in practice or reporting?
Reporting only the ECASS-equivalent HI/PH grade and omitting the class 3 subtypes, or calling a haemorrhage symptomatic based on imaging appearance alone rather than applying the defined NIHSS thresholds.
How does the Heidelberg Bleeding Classification relate to the older ECASS HI/PH terminology?
It retains the ECASS HI1, HI2, PH1 and PH2 labels within its Class 1 and Class 2, but adds Class 3 (remote PH, IVH, SAH, SDH) and formal symptomatic/asymptomatic and relatedness criteria that ECASS did not include.

Sources