Derivation and Validation of a Risk Factor Model to Identify Medical Inpatients at Risk for Venous Thromboembolism

Area of Science:

• Medical research
• Clinical epidemiology
• Health informatics

Background:

• Venous thromboembolism (VTE) prophylaxis is crucial for hospitalized medical patients at high risk.
• Existing VTE risk assessment models lack comprehensive comparison in large patient datasets.

Purpose of the Study:

• To develop and validate a novel multivariable prediction model for identifying patients at high risk of VTE.
• To compare the performance of the new model against established VTE risk assessment tools.

Main Methods:

• A derivation cohort of 155,026 patients from 6 years of data across 12 hospitals was used to identify VTE risk factors.
• VTE was identified using a combined administrative codes and clinical data algorithm.
• The developed model was validated in temporal, cross-validation, and external cohorts, with performance assessed by the C-statistic.

Main Results:

• The final model included 13 patient risk factors and achieved an optimism-corrected C-statistic of 0.79.
• In the temporal cohort, the Cleveland Clinic Model (CCM) outperformed the Padua (0.76 vs. 0.72) and IMPROVE (0.68) models.
• The CCM also outperformed the IMPROVE model in the external cohort (0.59) and showed similar performance to Padua (0.70 vs. 0.66).

Conclusions:

• The newly developed Cleveland Clinic Model (CCM) is an effective tool for VTE risk assessment in hospitalized medical patients.
• The CCM demonstrates superior or comparable performance to existing models, supporting its clinical utility for VTE prophylaxis decisions.