Healthcare utilization and chronic condition clusters in multimorbidity patients using weighted k-means: a...

Area of Science:

• Healthcare Management
• Public Health
• Clinical Informatics

Background:

• Multimorbidity presents a significant challenge to healthcare systems, marked by increased utilization and fragmented care.
• Identifying high-need patients with multiple chronic conditions is vital for optimizing care delivery and resource allocation.
• Understanding patient heterogeneity is key to addressing the growing burden of chronic diseases.

Purpose of the Study:

• To identify and characterize distinct patient clusters based on multimorbidity profiles and healthcare utilization patterns.
• To inform the development of targeted interventions for diverse patient groups with multiple chronic conditions.
• To enhance the efficiency of healthcare resource allocation by understanding patient needs.

Main Methods:

• A weighted K-means clustering algorithm was applied to a large dataset of 1,184,334 individuals.
• Chronic conditions were defined using diagnostic algorithms based on ICD-10 and ATC codes, encompassing 33 conditions.
• Sociodemographic variables were utilized to describe the characteristics of the identified patient clusters.

Main Results:

• Four distinct patient clusters emerged, differentiated by their chronic condition profiles and healthcare utilization levels.
• Cluster 1 exhibited high healthcare utilization, with a significant burden of somatic and mental health conditions and low social status.
• Cluster 2 comprised younger women with mental health conditions and high psychological service use; Cluster 3 had low utilization with common conditions; Cluster 4 included older individuals with complex somatic conditions.

Conclusions:

• Patient clusters with multimorbidity display significant variations in health conditions and healthcare engagement.
• Tailored management strategies are essential, particularly considering mental health conditions and social determinants of health.
• Effective resource management requires a nuanced understanding of patient needs within multimorbidity clusters.