PAD Amputation Risk & Intervention Service Evaluation Tool

Purpose: This project is to identify population segments with high PAD-related amputation risk and inadequate intervention services, enabling more targeted outreach and resource planning.

Materials and Methods: We developed a machine learning model to estimate PAD-related amputation risk using publicly available variables, including Area Deprivation Index (ADI), racial demographics, diabetes and obesity prevalence, smoking rates, food environment index, and uninsured rates. PAD intervention rates were calculated for every U.S. County using proprietary data sources. Each county was then categorized by amputation risk (“Well Below Average” to “Well Above Average”) and intervention rate (“No PAD Intervention Market” to “Well Above Average”).

Results: An interactive heatmap was created to visualize all 3,146 U.S. counties, allowing users to explore 30 population segments based on their amputation risk and intervention rate. Segment-based filtering dynamically updates the map, providing a detailed view of counties within each category and offering insights into potential contributing factors.

Conclusion: This tool offers a scalable and intuitive solution for identifying geographic and demographic areas where PAD interventions may be most urgently needed. Knowing where these regions are allows us to make more informed decisions around outreach with state and local organizations, and resource allocation regarding free PAD screening services and awareness education, which in turn, can contribute to improving patient outcomes and reducing preventable amputations. Future enhancements may include integrating patient-level data from specific organizations to help intervention providers assess and optimize their service performance.