What Are CMS Core Measures and How Do They Work?

CMS core measures are standardized quality metrics that the Centers for Medicare & Medicaid Services uses to evaluate how well hospitals deliver care for common, high-stakes conditions like heart attacks, pneumonia, sepsis, and surgical procedures. They emerged in the early 2000s as a joint effort between CMS and The Joint Commission, and they now influence hospital reimbursement, public reporting, and accreditation decisions. The evidence on whether these measures actually improve patient outcomes, though, is more complicated than the policy ambition behind them suggests.

What the Measures Cover and Why They Exist

At their core, CMS core measures are process measures. They track whether a hospital does a specific thing for a specific condition: Did the heart attack patient get aspirin on arrival? Was the right antibiotic given before surgery, and was it given on time? Did the sepsis patient receive a blood culture and intravenous fluids within the required window? The logic is straightforward: evidence-based guidelines say these interventions save lives, so tracking compliance should push hospitals toward better care.

Over the years, CMS has introduced and retired dozens of individual measures across clinical domains. The major categories have included acute myocardial infarction (heart attack), heart failure, pneumonia, surgical care, perinatal care, sepsis, and patient experience. Some measures have been dropped as compliance became nearly universal, which is itself a sign of success. Others have been added as clinical evidence evolved or as new policy priorities emerged. CMS has increasingly moved toward outcome measures (did the patient survive, get readmitted, develop an infection?) rather than purely process measures, though process measures remain embedded in many reporting programs.

Heart Attack Measures and What They Showed

The acute myocardial infarction (AMI) measures were among the earliest and most studied. Hospitals were scored on whether they delivered aspirin at arrival, beta-blockers at arrival, timely reperfusion therapy, and several discharge interventions. The results looked encouraging: a quality-improvement study found that as compliance with AMI core measures improved over several years, in-hospital mortality dropped substantially. Patients eligible for the aspirin-at-arrival measure, for instance, saw adjusted odds of in-hospital death fall by more than half by year six compared to baseline, and patients eligible for a bundle of aspirin, beta-blockers, and timely reperfusion showed a similarly large mortality reduction.¹

But a broader look at the data tempered that optimism. A study of thousands of hospitals found that while some AMI process measures correlated with 30-day mortality rates, the measures together explained only about 6% of the hospital-level variation in mortality.² In other words, a hospital could score perfectly on every AMI checkbox and still have higher mortality than a hospital with middling compliance, because so many other factors (staffing, ICU capacity, patient complexity, local demographics) drive outcomes. This gap between process compliance and actual results became a recurring theme across nearly every clinical domain where core measures were applied.

Longer-term follow-up showed a similar mixed picture. Among AMI and pneumonia patients, high adherence to core measures was associated with better survival and fewer readmissions compared to very low adherence. But for heart failure patients, the results were inconsistent and sometimes even paradoxical, with certain measures not translating into the expected outcome improvements.³

The Sepsis Measure Controversy

No single CMS core measure has generated as much debate as SEP-1, the severe sepsis and septic shock early management bundle introduced in 2015. SEP-1 requires hospitals to complete a tightly timed sequence of interventions within three or six hours of recognizing sepsis: blood cultures, lactate measurement, broad-spectrum antibiotics, intravenous fluids, and vasopressors when needed. On paper, these are all things clinicians agree are important. In practice, the measure became a lightning rod for controversy over whether rigid time windows and documentation requirements actually help patients.

The evidence is genuinely split. One study found that SEP-1 compliance was associated with a large mortality benefit for patients in septic shock: those who met the bundle criteria had a mortality rate of about 26%, compared to roughly 42% in those who did not.⁴ But the same study found no survival benefit at all for patients with severe sepsis (the less critically ill group), where mortality was essentially identical whether or not the bundle was completed.

A multicenter retrospective study painted a more cautious picture. Hospitals that failed SEP-1 had higher raw mortality rates, but once researchers adjusted for how sick the patients actually were, the association between SEP-1 failure and death was no longer statistically significant.⁵ The implication is that hospitals with sicker patients were both more likely to miss the bundle’s time windows and more likely to have deaths, and the bundle itself may not have been the decisive factor.

The most damning finding came from a large national analysis comparing mortality before and after SEP-1 was implemented. Short-term mortality rates were essentially flat: about 20.3% in the period before SEP-1 and 20.4% afterward, with no significant change in either level or trend after adjusting for patient risk.⁶ This does not mean the individual interventions in the bundle are useless. It may mean that measuring compliance with strict time windows is not capturing what actually matters, or that hospitals were already doing most of these things before the measure was mandated.

Surgical Care and Infection Prevention

The Surgical Care Improvement Project (SCIP) was launched with an ambitious goal: reduce surgical site infections by 25% through standardized protocols covering antibiotic timing, antibiotic selection, hair removal methods, and perioperative glucose and temperature management.⁷ The idea was sound: surgical infections are common, costly, and often preventable.

The results, however, were mixed in a now-familiar way. When researchers looked at an all-or-none composite (meaning a hospital had to pass every single SCIP item for a given patient), full compliance was associated with roughly half the rate of postoperative infections compared to noncompliance. But when individual SCIP items were examined one by one, none of them showed a statistically significant association with infection reduction on their own.⁸ This suggests that it is the bundle approach, getting everything right for a given patient, that drives the benefit rather than any single intervention checked off in isolation. It also means that a hospital performing well on most items but missing one could look bad on paper without necessarily providing worse care.

Perinatal Care and a Clear Win

Not all core measure stories are ambiguous. One area where standardized measurement appears to have driven clear improvement is perinatal care, specifically the push to eliminate early elective deliveries. For years, a significant number of babies were being delivered by scheduled induction or cesarean section before 39 weeks of gestation without a medical reason, which increases the risk of respiratory problems and NICU admissions. CMS and state Medicaid programs began publicly tracking and penalizing early elective deliveries, and rates dropped sharply. Across twelve states, early-term elective deliveries fell by about a third between 2007 and 2011.⁹ This is one of the clearest examples of a core measure achieving exactly what it was designed to do: changing provider behavior in a way that directly benefits patients.

When Measures Do Not Move the Needle

Pediatric asthma care provides a useful counterexample. Children’s hospitals were measured on a composite of asthma care quality metrics (the CAC-3 measure), including whether patients received reliever medications, systemic corticosteroids, and a home management plan of care. A study across children’s hospitals found no significant association between hospital-level CAC-3 compliance and post-discharge emergency department visits or readmissions at 7, 30, or 90 days.¹⁰ Hospitals that scored high on the measure did not have meaningfully better downstream results than hospitals that scored low.

This is not necessarily an indictment of good asthma care. It may mean the measure was tracking the wrong things, or that the factors driving readmissions (home environment, outpatient follow-up, medication access, air quality) lie outside the hospital’s control. Either way, it illustrates a limitation of the core-measure framework: a process metric can achieve near-universal compliance and still fail to move the outcome it was supposed to influence.

Gaming Through Diagnostic Coding

One of the more uncomfortable findings in the core-measure literature is how easily hospitals can improve their scores through legitimate changes to diagnostic coding rather than actual improvements in care. A study simulated what would happen if hospitals with above-median pneumonia mortality rates recoded eligible patients from pneumonia to sepsis or respiratory failure, both of which are clinically defensible alternative primary diagnoses in many cases. The result: 90 out of 100 such hospitals would improve their pneumonia mortality rate, and 41 of them would drop below the national median entirely. For readmission rates, the effect was smaller but still meaningful, with 66 out of 100 hospitals improving their numbers.¹¹

This is not fraud. A patient with pneumonia who develops sepsis can reasonably be coded either way. But the fact that coding choices can swing a hospital from “worse than average” to “better than average” without any change in bedside care raises real questions about how much public reporting of these measures tells consumers. A hospital that looks like it has low pneumonia mortality may simply have aggressive coders, while one that looks worse may be coding more conservatively.

Disproportionate Impact on Safety-Net Hospitals

CMS core measures feed into payment programs like the Hospital Value-Based Purchasing Program and the Hospital Readmissions Reduction Program, which adjust Medicare reimbursements based on performance. These financial consequences do not land evenly. Safety-net hospitals, those serving large shares of low-income and uninsured patients, are penalized at higher rates and with larger dollar amounts than other hospitals. A cohort study found that safety-net hospitals were more likely to be penalized under value-based purchasing (roughly 63% versus 51% for non-safety-net hospitals, depending on how “safety-net” was defined), and they faced larger payment penalties on a per-bed basis.¹²

The same pattern shows up in sepsis. An analysis of SEP-1 performance found that safety-net hospitals struggled more with the measure, and the researchers concluded that existing sepsis policies may actually widen health disparities rather than narrow them.¹³ The problem is structural: hospitals caring for sicker, poorer patients with fewer resources are being measured against the same benchmarks as well-funded suburban medical centers, and then losing money when they fall short, which makes it even harder to invest in the improvements that would close the gap.

Researchers in health equity have called for performance measurement and payment systems that explicitly account for disparities rather than accidentally punishing the hospitals trying to serve the most vulnerable populations.¹⁴ Some proposals include stratifying performance data by race, income, or insurance status so that improvement within disadvantaged populations gets rewarded rather than overlooked.

The Reporting Burden

Complying with core measures requires extensive documentation, chart abstraction, and electronic reporting. For large hospital systems with dedicated quality departments, the administrative load is manageable. For smaller and rural practices, it can be overwhelming. A study connecting over a hundred small-to-medium primary care practices to an electronic clinical quality measure platform found that the capabilities and limitations of different electronic health record systems created significant barriers to accurate reporting.¹⁵ Data that should flow automatically from EHRs to quality-reporting systems often did not, requiring manual workarounds that consumed staff time and introduced errors.

The documentation demands also spill over into clinical workflow. Physicians already spend a large portion of their day in electronic health records, and the pressure to document compliance with quality measures adds another layer of charting. Research on physician documentation practices found that those who wrote the longest notes and used the most copy-and-paste spent substantially more time in the EHR after hours and closed fewer patient visits on the same day.¹⁶ Core-measure compliance is not the sole driver of documentation bloat, but it is part of a broader ecosystem in which the demand for trackable, auditable data competes with the demand for efficient patient care.

Patient Experience as a Quality Measure

CMS also incorporates patient experience into its quality framework through HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems), a standardized survey that asks patients about communication with nurses and doctors, hospital cleanliness, pain management, and discharge information. HCAHPS scores feed into value-based purchasing, meaning that how patients feel about their care directly affects hospital payments.

Research has found that lower clinical quality measures tend to track with lower patient satisfaction, suggesting that the two are not independent: hospitals that struggle with clinical processes also tend to deliver a worse patient experience.¹⁷ Whether satisfaction surveys genuinely capture care quality or simply reward hospitals that are newer, better-staffed, and located in wealthier communities remains a point of ongoing debate. Critics worry that tying reimbursement to satisfaction incentivizes amenity-driven care over medically appropriate but uncomfortable decisions, like limiting opioid prescriptions.

Where Core Measures Stand Now

CMS has been consolidating and streamlining its quality programs under what it calls the Universal Foundation, an effort to reduce the sheer number of measures hospitals must track and to align metrics across Medicare, Medicaid, and commercial insurance programs. The goal is to focus on a smaller set of high-impact, outcome-oriented measures rather than maintaining dozens of process checklists that may not move the outcomes they are supposed to influence.

The shift reflects lessons learned over two decades. Process measures were useful for establishing minimum standards and driving early improvements in conditions like heart attack care and perinatal safety. But their ability to differentiate truly excellent hospitals from merely compliant ones has proven limited. Outcome measures, patient-reported outcomes, and equity-stratified reporting are all gaining ground as supplements or replacements. Whether these newer approaches will avoid the same pitfalls (gaming, disparate impact on safety-net hospitals, administrative burden) remains an open question that the next generation of health-services research will have to answer.