External validation of the IMPROVE Bleeding Risk Assessment Model in medical patients

Area of Science:

• Internal Medicine
• Clinical Risk Management
• Pharmacology

Background:

• The IMPROVE Bleed Risk Assessment Model (RAM) is the sole tool for assessing bleeding risk in hospitalized medical patients.
• It utilizes 11 clinical and laboratory factors to predict bleeding events.

Purpose of the Study:

• To externally validate the predictive accuracy of the IMPROVE Bleed RAM.
• To assess the model's ability to identify high-risk patients for bleeding events.

Main Methods:

• Retrospective chart review of 12,082 hospitalized medical patients between October 2012 and July 2014.
• Application of the IMPROVE Bleed RAM scoring system to calculate risk scores.
• Dichotomization of patients into low-risk (<7) and high-risk (≥7) groups to compare bleeding rates.

Main Results:

• A total of 2.6% of patients experienced any bleed within 14 days of admission.
• High-risk patients (score ≥7) had significantly higher rates of any bleed (4.68%) and major bleeds (3.2%) compared to low-risk patients (2.12% and 1.5%, respectively).
• The Odds Ratio for any bleed was 2.3 and for major bleeds was 2.2 in the high-risk group, with p<0.0001 for both.

Conclusions:

• The IMPROVE Bleed RAM was successfully externally validated in a large cohort of hospitalized medical patients.
• A score of 7 or higher effectively identifies patients at increased risk for major and any bleeding events.
• The validated IMPROVE Bleed RAM can aid in tailoring thromboprophylaxis strategies for hospitalized medical patients.