Development of MDS-Based Predication Model for COPD Severity in Nursing Home Residents.

Area of Science:

• Gerontology
• Pulmonology
• Health Informatics

Background:

• Assessing chronic obstructive pulmonary disease (COPD) severity in nursing home (NH) residents is difficult due to incomplete data.
• Traditional assessment methods often lack detailed symptom and exacerbation history in this population.

Purpose of the Study:

• To determine if the Minimum Data Set (MDS), a standard clinical assessment tool, can predict COPD severity in NH residents.
• To map MDS variables to the Global Initiative for Chronic Lung Disease (GOLD) severity groups (A-D).

Main Methods:

• A prospective cohort analysis of 175 NH residents aged 60+ with COPD.
• Multivariate multinomial logit models were used to correlate MDS data with GOLD groups, derived from the COPD Assessment Test (CAT) and exacerbation history.
• Long-acting bronchodilator (LABD) use and dyspnea were identified as key predictors.

Main Results:

• The MDS successfully predicted GOLD A-D groups with 60.6% accuracy.
• Dyspnea and LABD use were significant predictors of COPD severity.
• A notable underuse of recommended LABD treatment was observed across GOLD B, C, and D groups.

Conclusions:

• The MDS is a viable tool for estimating COPD severity in NH residents.
• Validated MDS-based COPD severity prediction can guide evidence-based treatment and personalize care pathways.
• Findings highlight potential gaps in guideline-adherent pharmacotherapy for NH residents with COPD.