Patient outcomes

Our Ratings of patient outcomes focus on three measures:

• Central-line-associated bloodstream infections (CLABSIs). Central lines are catheters, or tubes, used to deliver fluids, medication, and nutrition to patients. Bloodstream infections are caused by a mishandling of those central lines and are the most deadly kind of hospital-acquired infection; up to 25 percent of patients who get those infections die from them. Our data are based on CLABSIs that affect patients while they're in a hospital's intensive-care unit (ICUs).
• Surgical-site infections (SSIs). These are surgery-related infections that occur on or close to the skin surface, deeper in the body, or in any part of the body that is opened and manipulated during surgery. They can occur within 30 days of most surgical procedures or, in the case of implants, within a year. They're the second most common hospital-acquired infection and have been linked to up to 8,000 deaths a year.
• Chance of readmission. This shows the chance that a patient who has had a heart attack, heart failure, or pneumonia will have to be readmitted to a hospital within 30 days of his or her initial discharge. The need for such readmission can indicate problems in care while the patient was first hospitalized, including a hospital-acquired infection.

Central-line-associated bloodstream infections

What is the source of the data for the bloodstream-infections Ratings?

Some states require hospitals to report data publicly, often through the federal Centers for Disease Control and Prevention (CDC). And some hospitals voluntarily report data to The Leapfrog Group, an independent organization that collects information on hospitals. Less than half of the states now report the data publicly, and less than a quarter of hospitals voluntarily report to Leapfrog.

How often are the bloodstream-infections Ratings updated?

Several times a year based on when the various state agencies and The Leapfrog Group post new data.

What do the bloodstream-infection scores mean?

Each state and The Leapfrog Group report data on bloodstream infections for a particular selection of ICUs, such as cardiac, medical, and surgical. We include data for all ICUs reported (except burn, trauma, and neonatal ICUs), which often differ from one source to another.

Since the risk of infection can vary among different kinds of ICUs, we use a measure that compares each ICU to national data on infection rates for that type of ICU, and then combine those scores. That method takes into account the different risk of infections in different types of ICUs, and then gives a single score for each hospital.

We report our findings as a hospital's percentage difference from national rates, given the particular set of ICUs reported for that hospital. The scores for each hospital are based on those percentage differences from national rates.

Why doesn't my hospital receive a bloodstream-infections Rating?

You might not live in a state that currently reports bloodstream infections to the public in a form we can use for our Ratings, and your hospital might not voluntarily report data to The Leapfrog Group. In addition, we only rate hospitals that report at least 1,000 central-line days for its ICU patients or at least three infections.

What are the limitations of the bloodstream-infections Ratings?

The Ratings come from recent data but it's possible that updates will show improvements or declines in performance. The Ratings are based on only those ICUs that states or Leapfrog require hospitals to report on, not every bloodstream infection in the hospital. Though research suggests that bloodstream infections can be reduced and even eliminated, comparisons between hospitals that care for very different patient populations should be done cautiously. Data are reported by each hospital to either its state or to the Leapfrog Group. Although some states, such as New York and Tennessee, audit the data, most don't.

For more details, see our Hospital Ratings Technical Report.

Surgical-site infections

What is the source of data for the Ratings on surgical-site infections?

About one-fifth of the states in the U.S. report data on surgical-site infections that occur after one or more of the following procedures: caesarean section, colon surgery, coronary bypass surgery, gallbladder surgery, hernia repair, hip or knee replacement, hysterectomy, and spinal fusion. We gather this data and use it as the basis for our Ratings.

How often are the Ratings of surgical-site infections updated?

Several times a year, depending on when the various state agencies and organizations from which we collect this information post new data.

What do the surgical-site infection scores mean?

We report our findings as a hospital's percentage difference from national rates, given the particular set of surgeries reported for that hospital. The scores for each hospital are based on those percentage differences from national rates.

Since the risk of infection varies among surgical procedures, we compare a hospital's infection rate for each procedure to national averages for those procedures, and then combine those scores. That method takes into account different risk of infections in different types of surgery, and then gives a single rating for each hospital.

Why doesn't my hospital have a Rating for surgical-site infections?

You might not live in a state that currently reports surgical-site infections to the public in a form we can use for our Ratings. In addition, we only rate hospitals that report at least 100 surgical procedures or at least three infections.

What are the limitations of the Ratings for surgical-site infections?

We make statistical adjustments for differences among hospitals, and research suggests that every hospital can sharply reduce and even eliminate infections regardless of the patients they treat, but comparing hospitals with very different populations should be done with caution. The Ratings are based on only those surgeries that states are required to report on, not every procedure performed in the hospital. A hospital's overall performance at preventing the infections might be better or worse than it is for the surgical procedures they do report. Not all hospitals do the same procedures. Data are reported by each hospital, and while some states, such as New York and Tennessee, audit the data, most don't. Ratings come from recent sources but it's possible that updates will show improvements or declines in performance. States report data on different schedules, so comparisons among hospitals in different states might not be reliable. Most surgical-site infections aren't identified until after patients leave the hospital, and patients with infections don't always return to the hospital where the surgery was performed, so many might go unreported.

For more details, see our Hospital Ratings Technical Report.

Chance of readmission

What is the source of the data for the Ratings on the chance of readmission?

The information is collected by the Centers for Medicare and Medicaid Services (CMS), which tracks patients who have had heart attacks, heart failure, and pneumonia and are readmitted to a hospital within 30 days of their initial discharge. Patients can be readmitted to the same or different facility, and for the same or different condition.

How often are the Ratings updated?

As soon as possible after the government releases the data. The government's August 2011 release covers readmissions following discharges between July 2007 and June 2010.

What do the scores for chance of readmission mean?

We provide them for heart-attack, heart-failure, or pneumonia patients in any hospital that has at least 25 cases for any of those conditions. For hospitals that have readmissions data for all three conditions, we provide a composite Rating. The chance of readmission is calculated from readmission rates after a statistical adjustment for how sick patients were when they were first admitted to the hospital and for the amount of data available for each hospital. We combine adjusted readmission rates across the three conditions. Since heart attack is the least common, it receives a proportionally lower weight in the score.

Why doesn't my hospital receive a Rating?

Since the information comes from Medicare, we can only rate hospitals that see Medicare patients. In addition, we only rate hospitals that have at least 25 discharges for each of the three conditions.

What are the limitations of the Ratings regarding the chance of readmission?

These data come from billing and other administrative data submitted by hospitals to Medicare, and such records are less reliable than data based directly on the measurement of patient outcomes. In addition, while we think they provide a good indication of readmission rates overall, they reflect only readmissions for Medicare patients treated for three conditions. Ratings come from recent data but it's possible that updates will show improvements or declines in performance. The percentages reported are not exact numbers but estimates based on the statistical model used, and have a margin of error. Hospitals that have relatively low numbers of discharges have wider margins of error, and because of the statistical model CMS uses, are statistically adjusted to be closer to the average of all hospitals. Finally, while the data are adjusted for the health status of the patients discharged by each hospital, comparisons among hospitals with very different patient populations should be done with caution.

The statistical methods used by CMS for risk adjustment are documented here.

For more details, see our Hospital Ratings Technical Report.