Patient classification and resource allocation in Veterans Administration nursing homes

Area of Science:

• Gerontology
• Health Services Research
• Nursing Informatics

Background:

• Effective patient classification is crucial for managing long-term care resources.
• Existing systems may not fully capture the complexity of nursing needs in long-term care settings.
• Veterans Affairs (VA) healthcare systems require robust methods for patient categorization.

Purpose of the Study:

• To compare the Resource Utilization Groups (RUGs) system with an independently generated classification system for VA long-term care patients.
• To examine the impact of detailed nursing time data (by provider and activity) on patient classification.
• To review the role of diagnoses in classifying long-term care patients.

Main Methods:

• Utilized data from 290 patients in two VA nursing homes.
• Employed the Automatic Interaction Detection (AID) Program for generating patient classes, a method similar to that used for Diagnosis Related Groups (DRGs).
• Analyzed disaggregated nursing times by provider type and nursing activity.

Main Results:

• The study generated patient classes using both RUGs and an independent AID-based system.
• The analysis revealed the influence of specific nursing activities and provider types on patient classification outcomes.
• The findings highlight the variability in patient needs within long-term care.

Conclusions:

• Patient classification in long-term care is complex and benefits from detailed nursing activity data.
• The AID program offers a viable method for developing classification systems in long-term care.
• Understanding nursing time allocation is essential for accurate patient categorization and resource management in VA facilities.