Long-term services and supports (LTSS) are a foundational component of the United States healthcare system, specifically from the perspective of an aging population. These services support older adults who need assistance with personal and medical needs. They are delivered in the home, community, and institutional settings. Nursing homes serve about 1.3 million individuals per year and represent a major component of LTSS. Medicaid plays a central role in financing this care, which accounts for 36% of long-term care spending, with over half of nursing home residents relying on it for coverage. There is substantial variation in how states finance and reimburse nursing home care because Medicaid is jointly funded and administered by federal and state governments.
These differences can influence both state budgets and the quality of care delivered. Increased Medicaid payments have been linked to improved outcomes like better pain management, fewer pressure ulcers, and maintenance of the functional status of residents. There are limited publicly available and up-to-date estimates of state-level Medicaid nursing home payment rates. This gap has become increasingly significant after the One Big Beautiful Bill Act of 2025, which introduced funding cuts and policy variations that affect Medicaid financing. A study published in Health Services Research aimed to develop and validate a new and transparent method to estimate state-level Medicaid nursing home spending rates.
Researchers used data from the Centers for Medicare & Medicaid Services (CMS) Medicaid LTSS Annual Expenditure Reports, along with LTCFocus data from Brown University. The proposed measure calculated average Medicaid spending per nursing home resident day by dividing total state Medicaid nursing home costs by the total number of Medicaid nursing home days. The denominator was estimated by using three components: total nursing home beds, occupancy rate, and proportion of residents supported by Medicaid. The method was applied to data from 2004 to 2019. It was validated against existing benchmarks from LTCFocus (2004) and the Medicaid and CHIP Payment and Access Commission (MACPAC) (2019). Statistical analyses included Loess plots and correlation coefficients, with sensitivity analyses conducted after exclusion of outliers.
The proposed method estimated a median daily Medicaid nursing home rate of $122.30 (interquartile range (IQR): $105.73 to $142.60) compared with $149.44 (IQR: $126.64 to $170.37) reported by LTCFocus in 2004. The median difference was -$26.78, with an absolute difference of $26.91. Correlation analysis showed a moderate to strong relationship between the two estimates (Pearson 0.70, Spearman 0.80). It strengthened to 0.87 and 0.85 after the exclusion of outliers like Arizona and Pennsylvania. The new method produced a median estimate of $194.02 (IQR: $159.22–$233.13) compared with $201.73 (IQR: $185.85–$227.42) from MACPAC. The difference was $18.14, with an absolute difference of $30.91. Correlations were moderate (Pearson 0.68, Spearman 0.63). These results indicate that the proposed method aligns with both historical and more recent estimates.
The study shows that this method provides an accessible and practical way to estimate state-level Medicaid nursing home payment rates by using publicly available data. Limitations include inconsistencies in managed care reporting, reliance on aggregate instead of beneficiary-level data, and variations in the inflation adjustments. The method enables timely cross-state comparison without requiring restricted datasets. This makes it a valuable tool for researchers and policymakers to estimate the impact of Medicaid funding changes on long-term care. This method can support future analyses that aim to improve the availability, quality, and organization of nursing homes in the United States by offering a consistent and replicable metric.
Reference: Ratliff HC, Petzold KR, Maust DT, Thomas KS. An approach to estimating state-level Medicaid nursing home spending. Health Serv Res. 2026. doi:10.1111/1475-6773.70112
