Annual risks and severity of bleeds and haemorrhagic stroke, by anticoagulant treatment received
| No treatment | Aspirin | NOAC*† | Warfarin* | |
| ICH‡ | 0.2% | 0.5% | 0.3% | 0.8% |
| HS§ | 0.1% | 0.3% | 0.2% | 0.5% |
| Other ICH | 0.1% | 0.2% | 0.1% | 0.3% |
| ECH¶ | 2.0% | 2.5% | 2.7% | 3.1% |
| GI bleed | 0.8% | 1.0% | 1.1% | 1.3% |
| Other ECH | 1.2% | 1.4% | 1.6% | 1.8% |
| All major bleeds | 2.2% | 2.9% | 3.1% | 3.8% |
| CRNM bleed | 5.3% | 6.9% | 7.9% | 9.5% |
| HS severity | Mild: 28%; moderate: 23%; severe: 12% fatal: 37%14 15 | |||
| Case fatalities | Due to other ICH: 13%, due to ECH: 2%14 15 | |||
*NOAC was used as treatment in base-case analysis and warfarin was considered in sensitivity analysis.
†A class effect was assumed by taking the average across apixaban, dabigatran (low and high dose), rivaroxaban, edoxaban (low and high dose).11
‡ICHs (59.7%) were assumed to be HS.14 15
§HS risk was adjusted by a factor of 1.97 (95% CI 1.79 to 2.16) per decade.31
¶ECHs (41.8%) were assumed to be GI bleeds.14 15
CRNM, clinically relevant non-major;ECH, extracranial haemorrhage;GI, gastrointestinal;HS, haemorrhagic stroke;ICH, extracranial haemorrhage;NOAC, new oral anticoagulant.