Data Dives

Medicare IDs Few Hospitals As Outliers In Readmissions

By Jordan Rau

July 23rd, 2012, 8:18 AM

Correction: An earlier version of this report incorrectly listed 10 hospitals as having better than average readmission rates for heart attack, heart failure and pneumonia patients. Only two hospitals, Citrus Memorial Hospital of Inverness, Fla. and Sarasota Memorial Hospital of Sarasota, Fla., had better than average readmission rates for those three conditions tracked by Medicare’s Hospital Compare website. Another 23 hospitals, including the 10 hospitals that were named in the blog post, had better than average readmissions in two of the three conditions.

Despite several years of concerted efforts, hospital readmission rates aren’t dropping, the latest Medicare data show. Readmissions cost Medicare $17.5 billion in inpatient spending, with nearly 10 million Mediciare beneficiaries readmitted within 30 days for any cause, a rate of nearly one in five Medicare patients who enter a hospital.

Nonetheless, Medicare continues to publicly single out very few hospitals as poor performers on its Hospital Compare website, even as the agency readies new financial penalties against those with too high rates. In the latest readmission data released Thursday, Medicare identified only:

  • 2 percent of hospitals which meet a case number threshold, or 41, as having worse than expected readmission rates for heart attack patients.
  • 4 percent of hospitals, or 159, as having worse than expected readmission rates for heart failure patients.
  • 3 percent of hospitals, or 123, as having worse than expected readmission rates for pneumonia patients. Medicare labels between 94 percent and 97 percent of hospitals as having average admissions.

Out of more than 4,000 hospitals, only eight hospitals were identified as having worse than average readmissions for all three diagnoses: Beth Israel Deaconess Medical Center in Boston; Florida Hospital in Orlando Fla.; Franciscan St. James Health in Olympia, Ill.;  Henry Ford Hospital in Detroit; Mount Sinai Hospital in New York; Olympia Medical Center in Los Angeles; San Juan VA Medical Center in San Juan, Puerto Rico; and Tampa VA Medical Center in Tampa, Fla.

Medicare is even more stingy in identifying hospitals as doing a superior job in avoiding readmissions. Only 1 percent of hospitals were labeled as better than average in heart attack (30 hospitals) or pneumonia cases (33 hospitals) and only 2 percent (94 hospitals) were labeled as better than average in heart failure.

Two hospitals got the coveted trifecta of having better than average readmissions for all three conditions: Citrus Memorial Hospital of Inverness, Fla. and Sarasota Memorial Hospital of Sarasota, Fla.

There are reasons behind Medicare’s caution: There actually aren’t that many readmissions for each of these three conditions at some individual hospitals, so it’s difficult to be statistically confident about calculating those hospitals’ readmission rates with enough precision to label those institutions outliers. (In fact Medicare doesn’t even try to categorize the readmission frequency of heart attack patients at 1,896 hospitals because there are too few cases.)

Medicare and researchers at Yale University have developed a readmission measure that includes all conditions bringing patients into a hospital, and that could allow for more accuracy in labeling hospital performance. But the hospital industry has been pushing back against the new effort by complaining to the National Quality Forum, a Washington, D.C., nonprofit that evaluates measures for the government, about whether the forum sufficiently takes providers’ opinion into account when endorsing measures. In their appeal letter, eight hospital organizations — including industry heavyweights Cedars Sinai Medical Center, Intermountain Healthcare and the Johns Hopkins Health System — wrote that:

It is noted that less than 20% of the more than 400 NQF members voted on this measure and a disproportionate number of Health Professional and Provider Organization members voted “No”, with the final total vote actually being less than 50% in favor of the measure.

These findings call into question serious concerns about whether the NQF Consensus Development Process achieves consensus among affected stakeholders as intended, and reflects decision making in a high stakes environment that is, in our view, neither fair or balanced.

Last month, NQF reaffirmed its decision. But NQF’s board also appointed a task force to consider its process of approving measures in response to the complaints.  It remains to be seen whether hospitals and other providers will end up getting more say–and thus sway–in how their performance is judged just as Medicare ramps up its Value-Based Purchasing program that will provide bonuses and penalties to hospitals based on how their patients fare.

jrau@kff.org

Comments are closed.

Share