Derivation of FAST CLOTS: A Multivariable Risk Assessment Model for Postoperative Venous Thromboembolism.

Area of Science:

• Medical Informatics
• Surgical Outcomes Research
• Thrombosis Prevention

Background:

• Venous thromboembolism (VTE) is a significant cause of postoperative complications.
• Current risk assessment models (RAMs) for VTE are often complex and lack specificity.
• There is a need for a more practical and clinically applicable VTE RAM.

Purpose of the Study:

• To develop and validate a new, parsimonious risk assessment model for postoperative VTE.
• To improve upon the limitations of existing VTE RAMs.
• To enhance adherence to risk assessment protocols.

Main Methods:

• Utilized multivariable logistic regression on the 2019 National Surgical Quality Improvement Project dataset.
• Incorporated patient demographics, comorbidities, and procedural details (CPT codes, minimally invasive surgery).
• Compared the new model's performance against established RAMs (Caprini, ACS-NSQIP, etc.).

Main Results:

• The FAST CLOTS model includes eleven key variables (e.g., age, BMI, functional status, comorbidities, operative factors).
• Achieved a c-statistic of 0.753, demonstrating good predictive accuracy.
• Showcased 89% sensitivity for VTE outcomes at a cut-off score of 6.

Conclusions:

• The FAST CLOTS model offers a clinically relevant and potentially more user-friendly alternative to existing VTE RAMs.
• Its design emphasizes biological plausibility and clinician face validity.
• Further validation could lead to improved VTE prevention strategies and patient care.