Identifying Veterans with a Higher Risk of Social Needs Using Cluster Analysis

Area of Science:

• Public Health
• Health Services Research
• Social Determinants of Health

Background:

• Social needs screening is crucial for population health and reducing disparities.
• Barriers to effective social needs screening persist.
• Administrative data can identify at-risk patient populations for targeted interventions.

Purpose of the Study:

• To demonstrate cluster analysis for identifying patient segments at risk for social needs.
• To inform targeted social needs screening and referral strategies.

Main Methods:

• Clustering analysis of 2010 Veterans based on demographics, health, and healthcare utilization variables.
• Survey assessed nine types of social needs.
• Weighted logistic regression analyzed the association between identified clusters and social needs risk.

Main Results:

• A distinct Veteran population segment was identified through cluster analysis.
• This segment, characterized by younger age, more women, and racial minorities, showed a higher risk of experiencing one or more unmet social needs (OR 1.74).
• This group also had a significantly higher risk for utility needs (OR 3.78).

Conclusions:

• Identifying characteristics of populations with unmet social needs enables targeted screening.
• The identified cluster, younger and with fewer comorbidities, may be overlooked by traditional healthcare interactions.
• This highlights the need for proactive, data-driven approaches to social needs assessment.