State makes push for death rate data

Correction: Because of a reporting error, a Page One story on Sunday about heart bypass surgery death rates incorrectly identified the school that works with the state Department of Public Health. Harvard Medical School's Department of Health Care Policy analyzes hospital and surgical morality rates under contract with the public health department.

Massachusetts health officials are concerned that a 21-month lag in public reporting of death rates for heart surgery patients could mislead consumers about hospitals' performances and are pushing surgeons and hospitals to make their results public more quickly.

The problem was highlighted last fall when a state website showed that Caritas St. Elizabeth's Medical Center in Brighton had a normal mortality rate for cardiac bypass surgery, though the hospital several months earlier had closely monitored the work of one of its busiest heart surgeons because an unusually high number of his patients died in 2004, according to a hospital report.

The ratings posted by the state last fall were almost two years old -- from 2003. Meanwhile, in 2004, the hospital's mortality rate for bypass surgery rose to an unusually high 5.3 percent, according to two hospital reports and executives. In March and April of 2005, the hospital had a senior surgeon assist on all cases performed by Dr. Robert Moses and also that year asked Moses to make numerous improvements in his work, according to the hospital reports that the Globe obtained from the state Department of Public Health.

The hospital also made changes to its cardiac surgery program, according to interviews with hospital executives and the reports, and the mortality rate for both the hospital and the surgeon returned to normal in 2005.

Similar delays occurred in disclosing an unusually high percentage of cardiac bypass deaths at UMass Memorial Medical Center in Worcester since 2003.

Doctors at the hospital knew for two years about problems in the heart surgery program, including, at one time, a high infection rate. But they did not publicly announce the problems or suspend operations until after state public health officials presented them with a detailed analysis last September showing that the hospital's death rate was nearly twice the state average.

At the time, another state website showed UMass Memorial as having a normal mortality rate. The hospital has since made numerous improvements, reopened its program, and hired a new chief of cardiac surgery.

The situations highlight the limits of public reporting, as the administration of Governor Mitt Romney pushes for greater transparency about the quality and cost of medical care, and the conflicts erupting between doctors and state officials over how fast to report performance measures.

Earlier this month, the Division of Health Care Finance and Policy quietly decided to post information on 2004 heart surgery death rates for all hospitals on its website over the objections of physicians, who said the type of data the division used is not reliable.

''We felt it would be a disservice to consumers not to put this out there," said Amy Lischko, the division's assistant commissioner. ''Timeliness is really important with this kind of material."

Surgeons favor data analyzed by the Harvard School of Public Health under contract with the state Department of Public Health, which they say is highly reliable in adjusting for a hospital's mix of patients. But those data have a 21-month lag time; these are the data the state posted in October. Lischko had her department analyze its own data using different methods, which took 13 months.

Lischko and the doctors involved said they are now working toward a compromise that would speed up the posting of data surgeons accept as reliable.

Paul Dreyer, director of the Division of Health Care Quality at the public health department, said he did not ask St. Elizabeth's to close its cardiac surgery program, because the high mortality rate at St. Elizabeth's occurred during a single year and apparently has been addressed.

During 2004, 13 of the hospital's 243 cardiac bypass patients died soon after surgery, more than expected at a hospital with St. Elizabeth's type and number of patients, said Dr. John Chessare, senior vice president for quality and patient safety.

During an internal review of the cases in early 2005, according to one of the hospital reports obtained by the Globe, the hospital discovered that eight of Moses' 109 bypass patients died.

That report, dated Jan. 25, 2006, said the hospital made numerous recommendations to the surgeon in 2005, including ''consult with and request assistance of colleagues for difficult cases," ''prompt infectious disease consultation," ''eliminate extraneous conversation, focus on expeditious surgery," and ''need to decrease pump and cross-clamp times," a reference to the time patients are on the heart-lung bypass machine.

''It is noteworthy and gratifying that this surgeon has had no CABG (coronary artery bypass graft) mortalities thus far in 2005," according to a report sent to the health department on Dec. 28, 2005. The department asked the hospital for information on Dec. 19, after being notified by the hospital of its 2004 mortality rate, Dreyer said.

The January report referred to Moses as ''an outlier surgeon," a term Moses' lawyer, Paul Cirel, objected to in an interview. He said the hospital found no problems with Moses' skills; he simply had a period in 2004 when he operated on many sick and complex patients less likely to survive. ''His results before and after that suggest there hasn't been a problem," Cirel said.

A spokesman for the state Board of Registration in Medicine, which licenses doctors, said yesterday that it had been notified by St. Elizabeth's that the hospital had suspended Moses on Jan. 9, but could not disclose details of the action. Neither Cirel nor hospital executives would comment on his status.

During its review, the hospital found that the rates of infections after cardiac surgery were ''higher than desired and in several instances these were ultimately related to patient mortality."

In 2004, for example, 2.8 percent of patients got a type of infection called a ''deep sternal wound infection," including one patient who died.

The hospital began giving all cardiac surgery patients antibiotics within an hour before surgery, ''improved traffic patterns in the OR," and made other changes. As a result, just two patients got deep sternal wound infections in 2005, the report said.

In an interview, Chessare said that in 2005, just four patients died from cardiac surgery and that in the past 11 years, except for 2004, the program has had an outstanding record. ''We think we have an exemplary program," he said. ''Every program everywhere will have blips."

Still, Lischko and Dreyer said they want data reported faster, and in December they met with several physicians who helped develop the public health department's cardiac surgery monitoring program to discuss the issue. She said they did not discuss St. Elizabeth's situation.

Lischko asked the surgeons, as well as the analyst from Harvard, whether she could post preliminary data more quickly. The surgeons, who helped develop the Harvard program, said no, so Lischko decided to post on Feb. 1 data analyzed by her department, which showed St. Elizabeth's high mortality rate -- something confirmed by the hospital's own analysis.

Dr. David Torchiana, a cardiac surgeon and executive at Massachusetts General Hospital, attended the meeting. He said in an interview that he argued against posting data from a source other than Harvard.

He said the data made public by Lischko have ''very poor credibility" because they are based on hospital billing records and may not sufficiently adjust for factors such as a hospital having sicker patients, which could contribute to an above-average mortality rate.

''The state agenda is confusing," he said in a follow-up e-mail. ''One state agency is rushing to put out a report based on administrative data, which is of marginal value while another state mandated effort is working on an accurate state of the art report on the same topic that involves major effort by hundreds of people at significant expense to the state and its hospitals. It doesn't make any sense."

Dr. David Shahian, chairman of surgery at St. Elizabeth's, argued at the meeting against posting cardiac surgery death rates this month. He said in an e-mail that ''it's notoriously unreliable."

Shahian, who joined the hospital in November 2004, is leaving next month because, staff members said, he and the hospital's president, Dr. Robert Haddad, did not agree on various issues, including how best to improve quality.

But in the end, Shahian said in the e-mail, the doctors agreed to compromise with state officials.

The group has agreed to cut down the lag time on the Harvard data, partly by getting hospitals to review and correct the results more quickly.