Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Identifying Homogeneous Patient Clusters in Swiss University Hospital Through Latent Class Analysis.

Gilles Cohen1, Pascal Briot2, Pierre Chopard2

  • 1Finance Division, University Hospital of Geneva, Switzerland.

Studies in Health Technology and Informatics
|May 25, 2022
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Integrating ARIMA and Deep Learning Models for Counterfactual Evaluation of Nirsevimab's Early Impact on RSV Infant Admissions.

Studies in health technology and informatics·2026
Same author

The Medication Safety Adventure Trail: An Educational Intervention to Promote Public Awareness on Medication Safety.

Pharmacy (Basel, Switzerland)·2025
Same author

A Hexagonal Aim as a Driver of Change for Health Care and Health Insurance Systems.

The Milbank quarterly·2024
Same author

When Illegitimate Tasks Threaten Patient Safety Culture: A Cross-Sectional Survey in a Tertiary Hospital.

International journal of public health·2023
Same author

Total arthrodesis for non-rheumatoid wrists: Outcomes at 5 and 20years of follow-up.

Orthopaedics & traumatology, surgery & research : OTSR·2022
Same author

How to improve hospital admission screening for patients at risk of multidrug-resistant organism carriage: a before-and-after interventional study and cost-effectiveness analysis.

BMJ open quality·2022

This study identified 8 patient subgroups within a hospital population using clinical data. These clusters help understand patient heterogeneity and optimize resource allocation for better care.

Area of Science:

  • Health Services Research
  • Clinical Informatics
  • Healthcare Management

Background:

  • Hospitalized populations exhibit significant heterogeneity in patient characteristics, disease severity, and treatment responses, leading to varied outcomes and costs.
  • Identifying distinct inpatient clusters can enhance personalized care quality and optimize clinical resource utilization.
  • Super-utilizers (SUs) represent a critical group, disproportionately consuming healthcare resources and contributing significantly to overall costs.

Purpose of the Study:

  • To segment a large hospitalized patient population into clinically homogeneous subgroups.
  • To identify patterns in demographics, medical conditions, service types, and costs within and between patient clusters.
  • To inform targeted interventions and resource allocation strategies for diverse patient groups, including super-utilizers.
Keywords:
ClusteringHospital EfficiencyInpatient SegmentationLatent Class AnalysisQuality ImprovementSuper-Utilizers

Related Experiment Videos

Main Methods:

  • Utilized cost, utilization metrics, and clinical information from 32,759 patients admitted annually to the University Hospitals of Geneva between 2017-2019.
  • Employed Latent Class Analysis (LCA) to statistically identify and define distinct patient subgroups based on shared characteristics.
  • Analyzed the distribution of specific patient populations, such as super-utilizers, pediatric patients, and orthopedic patients, across the identified clusters.

Main Results:

  • Identified 8 distinct patient subgroups characterized by high similarity within groups and significant differences between groups.
  • Demonstrated that 82% of super-utilizer patients, 99% of patients under 20, and 78% of orthopedic patients were clustered into just 3 specific groups.
  • Revealed a distinct cluster comprising 90% adult women aged 20-40, highlighting demographic-specific groupings.

Conclusions:

  • Latent Class Analysis effectively segments heterogeneous hospitalized populations into meaningful clinical subgroups.
  • The identified clusters provide a data-driven framework for understanding patient diversity and tailoring healthcare delivery.
  • Targeting interventions based on these patient profiles can improve care quality, personalize treatment, and optimize healthcare resource management.