Adaptation and Initial Validation of Minimum Data Set (MDS) Mortality Risk Index to MDS Version 3.0

Area of Science:

• Gerontology
• Health Services Research
• Clinical Epidemiology

Background:

• Accurate prediction of mortality risk is crucial for clinical decision-making and resource allocation in nursing home settings.
• The Minimum Data Set (MDS) Mortality Risk Index-Revised (MMRI-R) has been used to predict mortality, but an updated version based on MDS version 3.0 (MMRI-v3) is needed.
• Identifying residents with limited life expectancy (LLE) is important for palliative care planning.

Purpose of the Study:

• To evaluate the predictive validity of the MMRI-v3 for 6-month mortality in veterans.
• To compare the predictive validity of the MMRI-v3 with a single MDS item indicating LLE.
• To assess the combined predictive ability of MMRI-v3 and the LLE indicator.

Main Methods:

• Retrospective, cross-sectional study utilizing MDS assessments from Veterans Affairs (VA) nursing homes.
• Included veterans aged 65 and older newly admitted between July 2012 and September 2015.
• Dependent variable: death within 6 months. Independent variables: MDS items for MMRI-v3 and LLE indicator.

Main Results:

• The MMRI-v3 demonstrated strong predictive ability for 6-month mortality (c-statistic 0.81), comparable to the original MMRI-R (c-statistic 0.76).
• MMRI-v3 showed superior predictive ability compared to the single LLE indicator (c-statistic 0.76).
• Combining MMRI-v3 with the LLE indicator yielded the highest predictive ability (c-statistic 0.86).

Conclusions:

• The MMRI-v3 is a valuable tool for predicting 6-month mortality in VA nursing home residents.
• The MMRI-v3 and LLE indicator are useful for research on palliative care interventions.
• These tools can aid in guiding the allocation of palliative care services in clinical practice.