Evaluating the Minimum Data Set 2.0 for Nursing Home Residents
Author Information
Author(s): Vincent Mor, Orna Intrator, Mark Aaron Unruh, Shubing Cai
Primary Institution: Brown University Medical School
Hypothesis
How valid and consistent is the Minimum Data Set (MDS) 2.0 data in U.S. nursing homes over time and across different locations?
Conclusion
The MDS shows reasonable consistency in diagnoses and internal functioning measures, but its discharge tracking should not be solely relied upon for indicating hospitalizations or mortality.
Supporting Evidence
- Match rates between MDS data and Medicare records exceeded 95% for residents aged 65 and over.
- Internal consistency of MDS items was high, with PPV exceeding .9 for many diagnoses.
- The CHESS index was highly predictive of one-year survival among nursing home residents.
- Discharge tracking records indicated a significant number of hospitalizations not substantiated by Medicare claims.
- PPV for major diagnoses like heart failure and diabetes was between .6 and .7.
Takeaway
The study looked at how well the data collected from nursing home residents matches up with hospital records, and found that while it's mostly accurate, there are still some issues.
Methodology
The study analyzed linked MDS and Medicare claims data for nursing home residents from 1999 to 2007, assessing internal consistency and predictive validity.
Potential Biases
Potential for underreporting or misclassification of hospitalizations and deaths in MDS records.
Limitations
The MDS discharge tracking records should not be used alone to indicate hospitalizations or mortality due to inconsistencies.
Participant Demographics
Most participants were nursing home residents aged 65 and older, with a stable average age of around 81 years.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website