N.C. Baptist Hospital performed worse than the national average in four of six categories identified by the federal government as key measures of treatment for Medicare patients who suffered heart attacks, heart failure or pneumonia during a three-year period.
Baptist officials, responding to the results, vigorously defended the hospital's quality of care.
They pointed to a U.S. News & World Report-sponsored study that found better performance by Baptist using some of the same Medicare data in ways they say better reflect Baptist's overall performance. The magazine's study ranked Baptist the best hospital in the Triad and one of the best in the nation in four categories.
Baptist officials said that while the government's performance review has merit, the results don't fully take into account how many very sick Medicare patients from 20 counties end up at Baptist, often referred from other emergency rooms.
Medicare officials say the results in the government report are risk-adjusted to take into account the kinds of patients a hospital serves.
Compared with other hospitals in North Carolina, Baptist's poorest performance in the government report came in heart attack death rates for Medicare patients within 30 days of admittance. Out of 85 hospitals in the state fully assessed by federal regulators, Baptist was the only one with a death rate worse than the national average for heart attack patients.
The hospital also performed worse than the national average in Medicare patient death rates for pneumonia and for Medicare patients who were readmitted within 30 days of their release after treatment for heart failure or pneumonia.
Baptist performed the same as the national average in death rates for heart failure and readmission rates for heart attack patients.
The results come from the Hospital Compare report provided by the Centers for Medicare and Medicaid Services, or CMS, the agency under the U.S. Department of Health and Human Services that reimburses hospitals for Medicare expenses. The report measured July 1, 2007, through June 30, 2010.
In the CMS report, most of the other major medical groups had rankings that were no different than the national average in all six categories.
The medical groups include Forsyth Medical Center, Duke University Medical Center in Durham, UNC Hospitals in Chapel Hill, High Point Regional Hospital, Carolinas Medical Center in Charlotte and Presbyterian Hospital in Charlotte.
Rex Hospital in Raleigh was the same as the national average except for a better-than-average rate for heart failure deaths. WakeMed in Raleigh was the same except for a better rate for readmissions for heart failure.
Moses Cone Hospital in Greensboro performed the same as the national average in every category except the pneumonia death rate, which was worse than the national average.
Dr. John McConnell, chief executive of Wake Forest Baptist Medical Center, and Dr. Russell Howerton, the center's chief medical officer, stressed that the hospital's medical practices are good.
They cite the recent U.S. News & World Report survey that compares Baptist with its local and state peers. The survey found Baptist to be the Triad's top hospital, based on CMS data, national reputation and other factors. The survey shows Baptist performing at or above academic medical center peers, such as UNC and Duke, in cardiology, a category that combines Medicare heart attack and heart failure rates.
Unlike the Medicare report, the U.S. News & World Report survey does not include results on readmission rates, where Baptist underperformed for pneumonia and heart failure. But Baptist is working on that issue. The hospital has recently joined a broad, communitywide initiative aimed at keeping patients healthy once they leave the hospital.
The magazine's survey also does not attribute a death to a hospital if the patient was referred from another hospital's emergency room. The survey is done by the nonprofit Research Triangle Institute group, which, according to its federal financial report, is affiliated with UNC Hospitals and Duke, and is used for commercial purposes.
McConnell said medical center officials take "all quality measures extremely seriously and have put in place a large number of process improvements to address both the CMS reported outcomes and many others.
"We remain committed to doing our best to ensure high-quality performance and fulfill our mission to care for all patients, including the most gravely ill."
Medicare officials said the death and readmission rates are risk-adjusted, which means they take into account several factors, such as indigent care and emergency room situations, so that hospitals in low-income areas or serving as trauma centers, for example, do not unfairly rate worse than the national average.
The Medicare survey assigns a death to a hospital if the patient was referred from another emergency room, but not if the patient was admitted to another hospital, then referred. That death would be attributed to the first hospital.
"The idea is that the hospital they are first admitted (to) is the one that decides about transfer, and so if they send the patient to a hospital that is not good — they should own that outcome. If they transfer to a great hospital, then they should get credit," said Dr. Harlan Krumholz, a cardiologist at Yale School of Medicine who helped come up with the statistical modeling for Hospital Compare.
An official with the Joint Commission, the independent nonprofit organization that accredits hospitals in the United States, said the CMS report's statistical model has some weaknesses. But he added that any hospital that gets four "worse" ratings should find out what's going on.
"To be an outlier in that many categories, it suggests that there's more than chance playing a part. I'd say the hospital should look at their processes to see where they can improve their processes that affect care," said Stephen Schmaltz, the associate director for the Center for Data Management and Analysis in the commission's Division of Healthcare Quality Evaluation.
Medicare data tend to be considered the standard measuring stick in most categories, said David Meyer, a senior partner for Keystone Planning Group LLC of Durham.
"Hospitals are sensitive to how the quality of their services is portrayed in these surveys," Meyer said.
"There is increasing pressure among employers, payers and the public for greater accountability of hospitals, foremost for them to justify the value of the expenditure and show improving quality. The heart and pneumonia (categories) are two fairly typical areas to focus on in measuring quality of care," he said.
Dr. Anthony Shih, executive vice president for programs for The Commonwealth Fund, a private foundation focused on improving health-care standards, said the death rate and readmission measures used by CMS have been endorsed by the National Quality Forum. The forum is an independent nonprofit organization charged with endorsing national consensus standards for measuring and publicly reporting on performance.
"As such, I believe them to be valid and reliable. They are not comprehensive measures of quality. Therefore, patients should use in their assessment other measures as well, including the chart-based process-of-care measures also published by CMS."
By the numbers
The death and readmission rates in the CMS report refer strictly to Medicare patients —generally those 65 or older — who die within 30 days or are readmitted within 30 days after release.
Baptist treated 419 Medicare patients who had heart attacks during the three-year period covered by the report. Baptist's heart attack death rate was 19 percent, while the national average was 15.9 percent.
Because the rate is based on averages and is risk-adjusted to guard against unfair comparisons, it would be imprecise to apply 19 percent against 419, said Krumholz, the Yale cardiologist. Instead, he said, the 19 percent death rate means that for about every 100 patients, about three additional patients died compared with the national average.
Only 35 other hospitals in the United States out of the 2,877 fully assessed by CMS had death rates for heart attack patients on Medicare that were worse than the national average.
In North Carolina, most hospitals — 82 — had death rates for heart attack patients that were no different than the national average. Two hospitals were better than the national average. Twenty-seven hospitals were not included in the review because they did not treat enough patients.
The N.C. Hospital Association includes the six categories — death and readmission rates for heart attacks, heart failure and pneumonia — on its comparison link on its website (www.nchospitalquality.org).
In that comparison, Baptist was the only hospital in its peer group to be ranked in the lowest quartile — the bottom 25 percent — in all six categories. Forsyth Medical Center was ranked in the top quartile in two categories, second quartile in three and third quartile in one.
"We take the Medicare data and place it in a more stringent form to present the information as transparently as possible," hospital association spokesman Don Dalton said. "Health-care data is a point of reference and not a point of judgment."
Theresa Smiley, the publicly reported data manager for Novant Health Inc., said Forsyth takes the Hospital Compare information seriously because it provides a "level playing field."
"We scrutinize that data as part of determining where our processes may not be up to par and doing the right thing for patients," Smiley said.
Only a few hospitals in North Carolina performed worse than the national average.
Eight hospitals had worse-than-average rates for pneumonia, including Moses Cone, Baptist, Iredell Memorial Hospital and CarolinaEast Medical Center. Readmission rates for heart failure were worse than average for Baptist, Davis Regional Medical Center in Statesville, Sandhills Regional Medical Center in Hamlet and Southeastern Regional Medical Center in Lumberton.
Readmission rates for pneumonia were worse than the national average for Baptist, Martin General Hospital in Williamston and Wayne Memorial Hospital in Goldsboro.
None had worse-than-average death rates for heart failure or readmission for heart attacks.
Baptist also ranked poorly when compared with the five other hospitals in North Carolina certified as Level One trauma centers. The others were Moses Cone, UNC, Carolinas Medical Center, Duke and Pitt County Memorial Hospital in Greenville.
Of those six trauma-center hospitals outside Baptist, only Cone performed worse than the national average in any of the six categories — the death rate for pneumonia.
Cone performed "as expected" for "survival" rate for pneumonia in the U.S. News & World Report ranking. However, Cone spokesman Doug Allred pointed to the CMS report in explaining why the hospital has tightened its focus on caring for pneumonia patients.
"We take these results very seriously at Cone Health," Allred said. "Since Hospital Compare is a national database, that's the one Cone Health uses as a benchmark.
"Our entire team of physicians, nurses and pharmacists are analyzing our data and protocols against the best known evidence to determine how we can rapidly improve these outcomes.
"We have done this for heart failure and heart attack patients. Our pneumonia patients deserve no less."
Issue of patient transfers
Baptist's poor ratings are based in part on a statistical model that overly weighs the hospital's practice of taking care of the sickest of the sick from the region, particularly patients who pass through the emergency rooms of other hospitals, said Howerton, Baptist's chief medical officer.
The CMS report "is an imprecise measurement tool for our institution and the role we serve in our region," Howerton said.
Of the 76 Baptist patients who died over the three-year period of an acute myocardial infarction, or heart attack, 18 passed away within the first 24 hours of arrival, Howerton said.
Baptist got better ratings in the hospital-comparison survey done by U.S. News & World Report because that survey does not include any patient transfers, Howerton said. That would mean, for heart attack patients, for example, that 40 percent would be excluded from the death rate calculations, based on estimates in an email from Howerton. He said he supports the Research Triangle Institute's methodology.
Asked whether all patient transfers should be excluded in Baptist's performance results, Howerton said, "I would not suggest that any patient should not 'count' in this analysis.
"I remain completely confident in the quality of patient care delivered here."
Email
Facebook
Google
Twitter
Favorites
More
Advertisement