Death rate data raise eyebrows

Officials question some numbers and the interpretation of others

The Post and Courier

New government data places death rates at most of Charleston County hospitals well within the average, except for Trident Medical Center, whose rates were above average for Medicare patients admitted for heart failure.

But administrators and health experts say that patient choices and odd discrepancies make the regional numbers misleading.

The numbers released Thursday by the Centers for Medicare and Medicaid Services are based on patients discharged between July 2006 and June 2007 for three conditions: heart failure, heart attack and pneumonia.

Allison Walters, administrator of Trident's cardiovascular service line, said that of 255 Medicare patients treated for heart failure, 40 died.

Those who died of heart failure at Trident either had a do-not-resuscitate order upon admission or, during their stay, made their wishes known, Walters said. "Every single one of these patients were into end-of-life issues," she said. "We did a detailed review of every death we had. We did not find any care issues."

The numbers, which hospitals got to preview in May, surprised Walters because Trident currently doesn't have a system for tracking 30-day mortality rates, she said.

Medicare calculated the rates by comparing hospital discharge rates with Social Security death records. That means that if a patient is discharged after being treated for heart failure and dies for some other, unrelated reason within 30 days, that patient will be included on the hospital's mortality rate.

"It's information, but you have to be careful with that information," said Steven Shapiro, chief medical officer of Roper St. Francis Healthcare.

In the 30-day mortality measure for pneumonia, Bon Secours St. Francis Hospital, while not significantly different from the national rate of 11.4 percent, came in near the bottom of that average range at 8.4 percent and was among the lowest in the country.

Roper Hospital also boasted the lowest local mortality rate in heart attack and heart failure. Shapiro credited Roper Hospital's 63-minute average door-to-balloon time, which is the time from when a patient enters the hospital to the time he reaches the catheterization lab. The national average is 90 minutes.

Coming .2 percent behind Roper Hospital in the heart attack mortality measure was the Medical University of South Carolina, which has a door-to-balloon average time of 53 minutes, said Bill Spring, administrator for heart and vascular services at MUSC.

Lynn Bailey, a Columbia-based health care consultant, said she has doubts about the numbers.

The total number of Medicare heart failure patients for MUSC totaled 88, compared to Trident's 255 and Roper's 291.

"That doesn't make any sense," she said. "It makes me question the data of the whole region." Sometimes with large facilities such a MUSC, data may be reported by large sub-units, she said, but where to find an explanation is not easy.

MUSC's Spring was surprised by the discrepancy, too. "Heart failure is the No. 1 disease among Medicare recipients," he said.

Ultimately, the answer is with those at the Centers for Medicare and Medicaid who slice and dice the data, he said.

The S.C. Office of Research and Statistics, Health and Demographics reported that there were 800 discharges in Charleston County for Medicare patients with heart failure in 2006. The government data totals 762 for the period July 2006 to June 2007.

Dr. Rick Foster, vice president for quality and patient safety with the S.C. Hospital Association, said the low number of patient discharges recorded for MUSC could be related to how conditions are recorded at the hospital.

If heart failure is related to other conditions, the system may not be picking it up, he said.

Seeing outcome related measures, such as mortality rates, is a trend Foster expects will continue.

"In general, we're supportive of outcomes measures. It is still somewhat new territory," he said.

Reach Jill Coley at 937-5719 or jcoley@postandcourier.com.

Copyright © 1997 - 2007 the Evening Post Publishing Co.