Guide
How to Choose a Nursing Home: A Data-Driven Guide
Skip the brochures. Use the same government data that regulators use to evaluate every Medicare-certified nursing home in America.
Published Feb 18, 2026 · 12 min read
TL;DR
Evaluate nursing homes across four data dimensions: health inspection deficiencies, payroll-verified staffing levels, clinical quality measures, and penalty history. Eliminate any facility with Special Focus Facility status, abuse citations, or immediate jeopardy findings. Use CareGrader's search to compare A-F grades, then visit your top three to five candidates in person.
Why Data Should Drive Your Decision
Most families choose a nursing home based on proximity, appearance, and word of mouth. These are reasonable starting points, but they miss the most important information: how well the facility actually delivers care, as measured by independent government auditors.
The Centers for Medicare & Medicaid Services (CMS) conducts unannounced inspections of every Medicare-certified nursing home in the country. CMS also collects payroll-verified staffing data, tracks clinical quality outcomes for every resident, and records every fine and enforcement action. This data is public, but it is scattered across multiple government databases and buried in technical jargon.
CareGrader synthesizes all of this into a single A-F grade for each facility. But whether you use our grades or go straight to the CMS data, the framework below will help you evaluate any nursing home systematically.
The Four Data Dimensions to Evaluate
Every nursing home generates data across four measurable dimensions. Together, these give you a comprehensive picture of care quality that no brochure, tour, or online review can provide.
1. Health Inspections (35% of CareGrader score)
CMS conducts unannounced health inspection surveys every 12 to 15 months. Trained surveyors spend multiple days on-site observing care delivery, reviewing medical records, interviewing staff and residents, and checking compliance with federal regulations. When they find problems, they issue deficiency citations.
Each deficiency is classified on two axes: scope (isolated, pattern, or widespread) and severity (rated on a scale from A through L). Levels A through C represent potential for minimal harm. Levels D through F involve actual harm or potential for more than minimal harm. Levels G through I indicate actual harm. Levels J through L represent immediate jeopardy to resident health or safety.
What to look for: Count the total number of deficiencies across the most recent three survey cycles. Pay close attention to any deficiency rated F or higher, which indicates actual harm or potential for more than minimal harm. Any citation at the J, K, or L level (immediate jeopardy) is a serious red flag.
2. Staffing Levels (25% of CareGrader score)
Since 2017, CMS has required nursing homes to submit staffing data through the Payroll-Based Journal (PBJ) system. Unlike the self-reported estimates used previously, PBJ data is verified against actual payroll records. This makes it the most reliable measure of how many nurses are actually working at a facility on any given day.
The key metric is RN hours per resident per day. Research consistently shows that higher RN staffing is associated with fewer pressure ulcers, lower rates of urinary tract infections, and reduced hospitalizations. The national average is approximately 0.63 RN hours per resident per day, but experts recommend a minimum of 0.75 hours for adequate care.
What to look for: Total nursing hours per resident per day (including RNs, LPNs, and CNAs), weekend staffing levels (which often drop significantly), and nurse turnover rates. High turnover, particularly above 50% annually for RNs, indicates instability that directly affects care continuity.
3. Quality Measures (25% of CareGrader score)
CMS tracks clinical outcomes for both long-stay residents (those living in the facility permanently) and short-stay residents (those receiving rehabilitation after a hospital stay). These quality measures are derived from the Minimum Data Set (MDS), which nursing homes are required to submit for every resident.
For long-stay residents, critical metrics include: percentage of residents with new or worsening pressure ulcers, percentage with urinary tract infections, percentage who received antipsychotic medications, percentage physically restrained, and percentage who experienced one or more falls with major injury.
For short-stay residents, focus on: rehospitalization rates within 30 days of admission, emergency department visit rates, and whether residents were successfully discharged to the community. High rehospitalization rates suggest the facility is not managing post-acute care effectively.
What to look for: Compare each metric to state and national averages. A facility that is significantly worse than average on multiple quality measures is showing a pattern, not a fluke. Pay particular attention to antipsychotic medication use, which can indicate chemical restraint rather than appropriate treatment.
4. Penalty History (15% of CareGrader score)
When CMS finds serious or repeated violations, it can impose enforcement actions: monetary fines (per-instance or per-day), denial of payment for new admissions, state monitoring, temporary management, and ultimately termination from the Medicare and Medicaid programs.
Fines are a lagging indicator. By the time a facility is fined, the problem has already been identified, investigated, and adjudicated. But the pattern of fines over time reveals whether a facility treats violations as one-time events to correct or as a cost of doing business.
What to look for: Total fine amounts over the past three years, the number of separate enforcement actions, and whether the facility has ever had payment denied for new admissions. A facility with recurring fines across multiple survey cycles is demonstrating a systemic problem, not an isolated incident.
Red Flags That Should Eliminate a Facility
Some data points are not just concerning — they should remove a facility from your consideration entirely. These are the indicators that correlate most strongly with unsafe care.
- !Special Focus Facility (SFF) Status
CMS designates the worst-performing facilities as SFFs. These are selected from the bottom 2% nationwide. SFF candidates are nearly as concerning. Check the CMS SFF list or look for the SFF flag on CareGrader facility pages. - !Abuse or Neglect Citations
Deficiencies tagged as abuse, neglect, or exploitation are the most serious category of violation. These indicate that a resident was harmed through intentional action or failure to act. CareGrader flags any facility with abuse citations prominently on its facility page. - !Immediate Jeopardy Citations (J, K, L)
An immediate jeopardy citation means a surveyor found conditions that have caused, or are likely to cause, serious injury, harm, or death to a resident. These are severity levels J (isolated), K (pattern), and L (widespread) on the CMS severity scale. Even a single immediate jeopardy citation warrants extreme caution. - !Payment Denial for New Admissions
When CMS denies payment for new Medicare or Medicaid admissions, it is one of the most severe enforcement actions short of decertification. It signals that CMS has determined the facility is not providing care at the minimum acceptable level.
CMS Star Ratings vs. CareGrader Grades
CMS publishes a 1-to-5 star rating for every nursing home through its Care Compare tool. The overall star rating is a composite of three separate domain ratings: health inspections, staffing, and quality measures. Each domain receives its own 1-to-5 star rating, and the overall rating is primarily driven by the inspection domain, with adjustments from the other two.
This system has a critical limitation: a facility can receive a high overall star rating despite having serious problems in one domain. A facility with a 1-star inspection rating can still achieve 3 stars overall if its staffing and quality measure ratings are high enough. The star system also does not explicitly flag SFF status, abuse citations, or penalty history.
CareGrader's grading methodology addresses these gaps. We use four components instead of three (adding penalty history as a separate dimension), weight health inspections at 35% of the composite score, and apply automatic score reductions for SFF status, abuse citations, and immediate jeopardy findings. The result is an A-F grade on a 0-to-100 scale that is harder for a fundamentally unsafe facility to game.
| Feature | CMS Stars | CareGrader |
|---|---|---|
| Scale | 1 to 5 stars | A to F (0–100 score) |
| Components | 3 (inspections, staffing, quality) | 4 (+ penalty history) |
| Penalty data | Not factored into rating | 15% of composite score |
| SFF flagging | Listed separately | Automatic score reduction |
| Abuse citations | Embedded in inspection score | Separately flagged + score penalty |
Step-by-Step: How to Compare Nursing Homes
Follow this sequence to systematically narrow your options from thousands of facilities to a short list of three to five strong candidates.
- 1
Define Your Geographic Area
Start by identifying the city or region where you need care. Use the care assessment tool to determine whether a nursing home is the right care level, or whether assisted living or home care might be more appropriate.
- 2
Filter by Grade and Red Flags
Search for nursing homes in your area and immediately eliminate any facility with an F grade, SFF or SFF candidate status, abuse citations, or immediate jeopardy findings in the past three years. This typically removes 10 to 20 percent of facilities from consideration.
- 3
Compare the Data Across Dimensions
For your remaining candidates, compare performance across all four dimensions. A facility that is strong in one area but weak in another may be masking problems. Look for facilities that perform consistently above average across inspections, staffing, quality measures, and penalties.
- 4
Check the Financials
Use the cost calculator to understand median costs in your state. Confirm that your top candidates accept your payment method (Medicare, Medicaid, private pay, long-term care insurance, or VA benefits). Higher cost does not reliably predict higher quality — the data frequently shows expensive facilities with poor inspection records.
- 5
Visit in Person
Visit your top three to five facilities at different times of day, including evenings and weekends. Observe how staff interact with residents, whether call lights are answered promptly, the cleanliness of common areas and resident rooms, and whether residents appear engaged or isolated. Ask to see the most recent state inspection report — it should be posted publicly in the facility.
How to Use CMS Data Directly
CMS publishes nursing home data through two primary channels: the Medicare Care Compare website (consumer-facing) and the CMS Provider Data Catalog (raw datasets). Care Compare is easier to navigate but shows limited information. The Provider Data Catalog contains the complete datasets that CareGrader uses.
Key datasets to review include: Provider Information (facility basics and star ratings), Health Deficiencies (individual deficiency citations with severity codes), Penalties (fines and enforcement actions), and MDS Quality Measures (clinical outcomes). Each dataset can be filtered by state, provider number, or facility name.
For most families, using CareGrader's facility search is faster and more accessible than navigating raw CMS data. We pull from the same government sources and present the information in a format designed for families, not researchers. Our data sources page links to every CMS dataset we use, so you can verify any data point independently.
What the Data Cannot Tell You
Government data is the best available proxy for care quality, but it has gaps. Inspections occur periodically, not continuously. Staffing data is averaged over reporting periods and may not reflect day-to-day fluctuations. Quality measures are clinical metrics that do not capture every dimension of the resident experience.
An in-person visit fills in what the data misses: the culture of the facility, the morale of the staff, the quality of the food, the availability of activities, the attentiveness to individual preferences. Data helps you eliminate bad options and identify strong candidates. Your own observation and judgment make the final call.
We recommend using CareGrader grades and CMS data as the foundation of your decision, supplemented by at least one unannounced visit to each finalist facility. For a detailed checklist of what to observe during your visit, see our nursing home visit checklist.
Ready to Start Comparing?
Search every Medicare-certified nursing home in America. See A-F grades, inspection histories, staffing data, and penalty records — all from official government sources.
Frequently Asked Questions
What is the most important factor when choosing a nursing home?
Health inspection results are the single most predictive factor of care quality. CMS conducts unannounced surveys every 12 to 15 months, and the resulting deficiency reports reveal problems that marketing materials never will. Look at the number of deficiencies, their severity (especially anything rated F or higher on the A-through-L scale), and whether the facility has repeat violations across survey cycles. A facility with few deficiencies and no immediate jeopardy citations is statistically more likely to deliver safe, consistent care.
How reliable are CMS star ratings for nursing homes?
CMS star ratings are a useful starting point but have significant limitations. The overall star rating is a composite that blends three separate domains (inspections, staffing, and quality measures), which can mask serious problems. A facility with a 1-star inspection rating can still earn 3 stars overall if its staffing and quality measure ratings are high. CareGrader addresses this by weighting health inspections more heavily (35%) and separately flagging critical red flags like SFF status, abuse citations, and immediate jeopardy findings that the star system can obscure.
What is a Special Focus Facility (SFF)?
A Special Focus Facility is a nursing home that CMS has identified as having a persistent record of poor care. CMS selects SFF candidates from the bottom 2% of facilities nationwide based on a point system that weighs severity and frequency of deficiencies. Once designated, SFF facilities undergo inspections every 6 months instead of the standard 12 to 15 months. If they fail to improve, they face progressive enforcement including fines and potential termination from Medicare and Medicaid. There are roughly 88 active SFF facilities and approximately 400 SFF candidates at any given time.
Should I trust online reviews of nursing homes?
Online reviews capture subjective experience but are unreliable as a primary evaluation tool. They skew toward extreme opinions, can be manipulated by facilities, and rarely reflect systematic care quality issues. Use reviews to supplement data-driven analysis, not replace it. Government inspection data, payroll-verified staffing levels, and clinical quality measures are objective, standardized, and cover every Medicare-certified facility. Start with the data, then use reviews and in-person visits to fill in the qualitative picture.
How many nursing homes should I visit before making a decision?
Visit at least three facilities, ideally five. But narrow your list before visiting by using data. Filter out any facility with an F grade, SFF status, abuse citations, or immediate jeopardy findings. Then rank the remaining options by their composite score across inspections, staffing, quality measures, and penalties. Visit your top candidates at different times of day, including evenings and weekends, when staffing levels tend to drop. This data-first approach ensures you spend your limited time visiting facilities that have already passed objective quality thresholds.