IMPROVE Bleeding Risk Score
The IMPROVE Bleeding Risk Score estimates the risk of major bleeding in acutely ill medical inpatients, helping clinicians decide whether the benefit of pharmacologic VTE prophylaxis outweighs the bleeding risk.
inputs
✓ when to use
Use alongside the Padua VTE Prediction Score when deciding on thromboprophylaxis in medical inpatients. If Padua ≥4 (high VTE risk), assess IMPROVE bleeding risk. If IMPROVE <7, pharmacologic prophylaxis is favoured. If IMPROVE ≥7, consider mechanical prophylaxis or carefully weigh risks.
✗ when not to use
The IMPROVE score was developed for acutely ill medical inpatients, not surgical patients. Not all risk factors are independently validated — the score provides a framework for structured decision-making rather than a precise probability. Always incorporate clinical judgement alongside the score.
clinical pearls
- The Padua-IMPROVE pair is the standard approach: Padua assesses VTE risk (≥4 = high), IMPROVE assesses bleeding risk (≥7 = high). When both are elevated, the decision requires individualised clinical judgement — neither score alone provides the answer.
- Active GI ulcer (4.5 points) and recent bleeding (4 points) are the highest-weighted items. These alone can push a patient above the ≥7 threshold.
- A high IMPROVE score does not contraindicate ALL thromboprophylaxis — mechanical prophylaxis (intermittent pneumatic compression) carries no bleeding risk and is an alternative when pharmacologic prophylaxis is too risky.
- The score uses fractional points (0.5 increments), which is unusual and can be confusing. Some institutions round to whole numbers for simplicity, though this changes the threshold interpretation slightly.
- Reassess both VTE and bleeding risk throughout the admission. A patient who was high bleeding risk on admission (post-GI bleed) may become low bleeding risk after 48 hours of stability, allowing prophylaxis to be initiated.