Data-driven modeling of diabetes care teams using social network analysis

Area of Science:

• Health Services Research
• Social Network Analysis
• Diabetes Care Management

Background:

• Understanding collaboration dynamics is crucial for effective diabetes care.
• Previous studies often overlook the network structures within healthcare teams.

Purpose of the Study:

• To analyze working relationships and collaborations within and between diabetes healthcare provider teams.
• To identify influential providers and community structures using social network analysis and multi-scale community detection.

Main Methods:

• Retrospective analysis of claims data from a large employer over two years.
• Social network analysis (SNA) applied to identify provider networks for 827 diabetes patients.
• Multi-scale community detection to identify densely connected provider groups.

Main Results:

• Medical laboratories and mail-order pharmacies were identified as central providers.
• Primary care physicians were in 76% of communities, specialists in 97%.
• Pharmacists were central in 24% of identified provider communities.

Conclusions:

• SNA metrics effectively identify central and influential providers in diabetes care networks.
• These insights can guide the implementation of health promotion programs by leveraging key influencers.
• The proposed framework offers multi-scale metrics for assessing care team relationships and their impact on patient outcomes.