AMPREDICT Mobility-4: A Novel Four Category Mobility Outcome Prediction Model for Patients Undergoing Vascular...

Area of Science:

• Orthopedics
• Vascular Surgery
• Rehabilitation Medicine

Background:

• Lower limb amputation (LLA) due to vascular aetiology necessitates accurate mobility outcome prediction.
• Existing models may not adequately capture the nuances of functional recovery across different amputation levels.

Purpose of the Study:

• To develop and validate the AMPREDICT Mobility-4 model for predicting functional mobility post-amputation.
• To aid shared decision-making regarding amputation level based on predicted mobility outcomes.

Main Methods:

• Retrospective and prospective cohort study of 704 patients undergoing transmetatarsal, transtibial, or transfemoral amputation.
• Utilized electronic health record data including demographics, comorbidities, and health behaviors.
• Employed multinomial logistic regression and lasso for variable selection to predict four mobility levels at 12 months post-amputation.

Main Results:

• The AMPREDICT Mobility-4 model identified 15 predictors of functional mobility.
• The model demonstrated good discrimination between wheelchair mobility and community ambulation (c-index=0.71), and advanced vs. less than advanced community mobility (c-index=0.73).
• Discrimination between household and basic community mobility was less precise.

Conclusions:

• AMPREDICT Mobility-4 effectively predicts functional mobility across different LLA levels using readily available data.
• The model enhances patient-surgeon communication and expectation setting regarding post-amputation mobility.
• It supports personalized decision-making for amputation level selection.