When I started reviewing hospital quality data professionally in 2011, most patients had no idea that the federal government collected detailed safety performance data on every Medicare-certified hospital in the country. Fourteen years later, the data is more detailed than ever — but patients still aren't using it as much as they should be.
This guide explains what "hospital safety" actually means in the context of CMS data, what the scores measure, and — critically — how to interpret them without making common mistakes.
What CMS Actually Measures
The Centers for Medicare & Medicaid Services (CMS) collects and publishes quality data through its Hospital Compare program. This data covers five main domains:
🏥 The Five CMS Quality Domains
- Mortality rates — 30-day death rates for conditions like heart attack, heart failure, pneumonia, COPD, hip/knee replacement, and CABG surgery
- Readmission rates — How often patients return to any hospital within 30 days of discharge (a proxy for care quality and discharge planning)
- Patient safety indicators — Complication rates for events that are largely preventable: blood clots, falls, pressure ulcers, sepsis following elective surgery
- Infection rates (HAI) — Hospital-acquired infections tracked through CDC's NHSN system: CLABSI, CAUTI, SSI, C. difficile, MRSA bloodstream infections
- Patient experience (HCAHPS) — Standardized survey asking discharged patients about communication, responsiveness, cleanliness, and discharge information
The Critical Thing Most People Get Wrong
Hospital safety scores are population-level statistics, not a guarantee about your individual care. A hospital with a score of 8.5/10 still has individual bad outcomes. A hospital scoring 4.0/10 still has many successful surgeries every week. What the scores tell you is: across thousands of patients, this facility tends to perform better or worse than average.
This distinction matters enormously when you're making a decision. If you're choosing between a Grade A and a Grade D hospital for a scheduled knee replacement, the Grade A hospital statistically gives you better odds of avoiding a complication. It is not a certainty — it's a probability.
What the Scores Don't Capture
After more than a decade reviewing this data, I'm acutely aware of what isn't measured. CMS scores don't tell you about:
- Individual surgeon or physician performance (only hospital-level data)
- Wait times in the emergency department beyond a basic "timely care" metric
- Nursing staff ratios (though these correlate with outcome scores)
- The quality of food, parking, or administrative efficiency
- Specialty-specific outcomes (a hospital might excel at cardiac care but underperform in neurology)
How SafeHospitals USA Calculates Its Score
Our composite Safety Score weights the five CMS domains with different emphasis based on their demonstrated relationship to patient outcomes in published research:
- Hospital-acquired infections: 30% weight
- Patient safety measures: 25%
- Mortality outcomes: 20%
- Readmission rates: 15%
- Patient experience: 10%
Scores are normalized to a 0–10 scale against the full national distribution of all Medicare-certified hospitals. This means a score of 5.0 is precisely average for the country, not "mediocre" — the distribution is weighted heavily toward the 4–6 range.
Practical Takeaways
Here is how I'd use these scores if I were making a personal hospital decision:
- For planned procedures, identify every hospital within a reasonable distance and start with those scoring 7.0+ or Grade B+
- Cross-reference with your insurance network — the best hospital in the world is less useful if you pay triple the out-of-pocket cost
- Ask your physician which hospital they have privileges at and where they believe you'd get the best outcome for your specific condition
- For emergency care, don't use these scores — go to the nearest ER