The Centers for Medicare & Medicaid Services provides “quality of care data on every Medicare and Medicaid-certified nursing home in the country” (Data.CMS.gov). The “Provider Information” dataset from CMS provides general information on nursing homes, including ownership type, staffing metrics, and health inspection scores, among other details. The dataset piqued my interest for numerous reasons:
The size of the dataset, with 14,752 observations of 120 variables, lends itself to numerous levels of analysis.
With its emphasis on performance outcomes and ratings, the dataset connects to the concept of accountability and performance management in public administration. The inclusion of variables such as citations received and fines levied also allows analysis of how CMS’s regulation of nursing homes influences performance.
Public administrators must respond to demographic changes, and the growth of the aging population in the United States requires strategic planning for staffing and financing elderly care. PBS reports, “Medicaid covers more than 60 percent of residents in nursing homes” (Grabenstein). The demand on the the program will continue to increase.
The Provider Information dataset is rich with variable options, and the scope of the data is vast. Narrowing down on a specific research question was difficult. After reviewing the data and performing some rudimentary statistics in R, I turned my attention to focus on human resources and performance outcomes.
Do nursing homes with lower nursing staff turnover demonstrate better performance?
Null Hypothesis: Nursing staff turnover rate is not associated with facility performance.
Alternative Hypothesis: Higher staff turnover rates are associated with worse performance.
Independent Variable: Total nursing staff turnover
Dependent Variable: Total weighted health survey score
In Introducing Public Administration, Shafritz et. al. note the role of human resources planning in “assessing what sets of skills [a larger organization] needs and how this translates into specific numbers for future replacements” (p. 405). The dataset contains nursing staff turnover statistics for individual nursing homes, but I believe the data could be illustrative about an entire sector of society. I selected total nursing staff turnover rather than more specific variables, such as registered nurse turnover, to better incorporate the various levels of licensed professionals who provide care, including licensed practical nurses and nurse aids.
As previously noted, the Provider Information dataset contains a huge number of variables. Many variables are ratings between 1-5, including health inspection, staffing, and quality measure ratings, which are the basis for the overall rating of a nursing home. While these ratings cleanly aggregate complicated data into easily-digestible scores, I wanted to utilize a continuous variable for more robust analysis.
The total weighed health survey score is a component of CMS’s Five-Star Quality Rating System that is incorporated into the overall health inspection rating score, represented by 1-5 stars. The Technical User’s Guide for the rating system outlines the process for the scores. In summary, the score is based based on the two most recent health inspections required of nursing homes that are part of Medicaid or Medicare; prior to July 2025, the score incorporated the three most recent surveys. Deficiencies identified during the inspections are assigned points based on the scope and severity. The rating also weights repeated visits to “confirm that correction of deficiencies have restored compliance.” Additionally, the most recent survey is weighted more than the second most recent survey (p. 1-7). A lower total weighed health survey score reflects better performance.
I selected the total weighed health survey score because of its reflection on the quality of care and safety for the residents of nursing homes that are part of Medicaid or Medicare. Additionally, unlike a broad summary statistic like the overall rating, the total nursing staff turnover variable is entirely separate.
In conclusion, I would like to explore the relationship between nursing turnover and facility performance by examining the relationship between the total nursing staff turnover and total weighted health survey score variables. I am particularly interested in how public administration’s focus on accountability and transparency may be reflected in the data. This topic is especially applicable as the population continues to age, and public administrators will be tasked with administering the programs that provide the majority of funding for nursing home care.
Centers for Medicare & Medicaid Services. (July 2025). Design for Care Compare Nursing Home Five-Star Quality Rating System: Technical Users’ Guide. https://www.cms.gov/medicare/provider-enrollment-and-certification/certificationandcomplianc/downloads/usersguide.pdf
Centers for Medicare & Medicaid Services (September 1, 2025). Provider Information. https://data.cms.gov/provider-data/dataset/4pq5-n9py
Grabenstein, H. (May 30, 2025). How cutting Medicaid could upend long-term care for many older Americans. PBS News. https://www.pbs.org/newshour/nation/what-republicans-possible-medicaid-cuts-could-mean-for-nursing-homes
Shafritz, J. M., Russell, E. W., Bordick, C. P., & Hyde, A. C. (2023). HR Management and Laobr Relations. In Introducing Public Administration (10th ed., pp. 388–428). Routledge, Taylor & Francis Group.