Headlines

As reported by The New York Times, August 21, 2005.

Hospitals Are Uneven on Basics

By Ford Fessenden

JUST about everyone in medicine agrees that a patient who shows up in the emergency room with a heart attack should be given aspirin. That simple treatment has been proved to cut death rates by nearly a quarter.

But dozens of patients at hospitals in Connecticut don't get that aspirin, or nine other crucial therapies for heart attack, heart failure and pneumonia, a review of recently published hospital data shows. Some hospitals are better than others, but none delivered all the treatments to eligible patients, in spite of their well-known benefit.

Last month, The New England Journal of Medicine reported that, although their performance had improved somewhat from 2002 to 2004, hospitals across the country were neglecting many lifesaving treatments. The journal also reported on another study that found that hospitals around Boston and Oklahoma City performed better than those in other large regions, but that even in those places hundreds of patients did not receive basic therapies.

A review of the data by The New York Times showed that hospitals in Connecticut tended to perform better as a group than those in New York, but not as well as hospitals in Massachusetts. Eight of Connecticut's 31 hospitals were in the top 10 percent in the country for at least one of the conditions.

Over all, Connecticut hospitals delivered appropriate treatment to heart attack victims 94 percent of the time, to heart failure patients 89 percent of the time and to pneumonia patients 79 percent of the time, all better than Manhattan hospitals. Only five Manhattan hospitals were in the top 10 percent.

The hospital performance data are collected by the federal Department of Health and Human Services under a program called Hospital Compare and cover 10 treatments considered essential for most patients. They include admissions from January to June 2004. Detailed information is online at www.hospitalcompare.hhs.gov.

"We really should be close to 100 percent on all of these measures," said Dr. Ashish Jha, a Boston doctor and author of one of the national studies published last month. "There's not much controversy about whether giving antibiotics to someone who has pneumonia is good or not."

Besides aspirin for heart attack and antibiotics for pneumonia, the measures include beta blockers for heart attack, angiotensin-converting enzyme inhibitors for heart failure and heart attack, and vaccinations for pneumonia.

Heart attack is an acute episode, usually requiring an emergency room visit, caused by constriction of flow in a coronary artery. Heart failure is a chronic condition, essentially the wearing out of the heart muscle to the point that it cannot pump enough blood.

Connecticut hospitals performed better as a group than those in most states, but they also had a head start. The state required hospitals to report their performance on these measures in 2003, six months before the federal program started. That has helped boost compliance, hospital officials said.

"What is measured improves, and what isn't measured doesn't improve," said Ken Roberts, spokesman for the Connecticut Hospital Association. "It's an incentive. Providers don't want to be at 85 percent when their neighbors are higher."

In spite of the head start, some hospitals still lagged. Midstate Medical Center in Meriden was below the national median for all three conditions.

"We think they're fair measures, and they represent what was happening at that time," said Barbara Kaplowe, director of quality improvement at Midstate. "We think we're doing much better now."

Ms. Kaplowe said the hospital had discovered that treatments, or the reasons for not giving them, were not being written in the charts. "We have this information available, and we go to the medical staff and present it to them," she said.

When they review them she said, "We find sometimes it's not treatment issues, it's documentation."

But authorities said failing to document is also a quality issue.

"If there are lifesaving treatments the patients have received and you don't document it in the chart, that's a failure as well," Dr. Jha said. "If I'm treating the patient later, and I have no idea they received an important treatment, or didn't receive it for a reason, that's not good."

At Day Kimball Hospital in Putnam, heart attack patients did not receive appropriate treatments 13 percent of the time, the highest in the state. Dr. Ann Errichetti, the hospital's president, said the numbers were improving; the proportion was 25 percent when the state first started to gather information in the third quarter of 2003.

"It isn't as simple as giving an order to doctors, you have to get physicians to partner with you," she said. "We have made progress and we're moving in the right direction, but we're not done."

Hospitals that performed well found they had to acknowledge there were multiple opportunities for even well-trained, well-intentioned people to allow important things to fall through the cracks.

"On average, for a general admission, 60 people touch a patient, and most of these with a high degree of independence," said Dr. Steven Strongwater, the director of John Dempsey Hospital in Farmington. "The system has to acknowledge that, even with the best of intentions, when someone writes an order, it might not happen in the way the order was written."

At John Dempsey Hospital, a teaching hospital that is part of the University of Connecticut Health Center, administrators have used a combination of checklists and what the hospital calls "patient safety rounds" to bring their failure rates down, nearly to zero for heart patients. The hospital ranks in the top 10 percent in the country in delivering treatments both to heart attack and heart failure patients.

Dr. Strongwater said everyone was trying to improve performance.

"It's a question of how vigorously they're doing it," he said.

The measures cover a narrow band of the range of care the hospitals deliver, and Connecticut hospital officials stressed that there were many other ways to judge hospital quality. Dr. Jha's study showed that a hospital that did well treating one of the conditions did not necessarily do well treating another, so a hospital's overall quality would be better judged by a wider range of measures.

But, he said, these are common ailments, and the treatments are widely acknowledged to be necessary and crucial.

"These are the bread and butter of medicine," Dr. Jha said. "We know a tremendous amount about how to take care of these people. We know that if you walk in with a heart attack and I give you an aspirin, your chances of surviving have improved dramatically."

And Connecticut hospital officials said the measures can be good indicators of a hospital's quality.

"If a hospital is at 100 percent on all the measures, that's not going to guarantee you a perfect outcome, but it's a pretty good indicator of their overall approach to care," said Mr. Roberts of the Connecticut Hospital Association. "I wouldn't base my decision solely on these measures, but I would look to them if I was to choose a hospital."

Dr. Jha agreed that patients should use the ratings to choose hospitals.

"They can say, 'I am not going to hospital x and instead go to hospital y,' " he said. If circumstances make it hard to go to another hospital, he said, "they can ask their doctors and nurses and administrators, 'Why are we not doing better?' "