A controversial measure is revised, but is the entire measure flawed? by Deborah Gesensway
Published in the March 2008 issue of Today's Hospitalist
While experts generally agree that delivering antibiotics quickly to patients with community-acquired pneumonia (CAP) is important, a debate has raged about how to improve—and measure—antibiotic timing for these patients. Now there are signs that the debate is leading to significant changes in public reporting and pay-for-performance measures.
For several years, the Centers for Medicare and Medicaid Services (CMS) and the Joint Commission have decreed that all CAP patients should receive their first dose of antibiotics within four hours. Many payers followed suit by creating quality-of-care measures that tracked antibiotic timing in CAP patients.
"Some of us aren’t sure that the time to antibiotics is really that crucial. Changing the measure to six hours makes no more sense than keeping it at four.”
–Christopher Fee, MD UCSF
however, have long claimed that a four-hour antibiotic window for CAP patients is inappropriate because it is based on retrospective, not prospective, data.
Even more importantly, they say, the measure is being applied to a patient population—general ED patients—who may differ substantially from the Medicare patients in whom the benefit was retrospectively demonstrated. And they complain that the four-hour goal leads to many unintended and adverse consequences.
It appears that some of those concerns have been heard. Starting this year, the CMS is changing its reporting requirement to target how many CAP patients receive their initial antibiotic dose within six, not four, hours of arrival at the hospital.
And there are other changes on the horizon. In June 2008, a new quality-of-care reporting standard—”30-day pneumonia mortality”—is due to take effect.
Many physicians say the use of such an outcomes measure could be a step in the right direction. They point out that no data link the use of process measures—such as timing in antibiotic administration—to lower mortality. They also note, however, that it remains to be seen how 30-day mortality figures will be structured and adjusted.
While no one is predicting that quality measures based on antibiotic timing for CAP are going to vanish entirely, the changes being made illustrate the difficulties in bridging the gap between clinical practice and quality improvement. The changes may also make the lives of inpatient physicians like hospitalists a little easier in the near future.
The evolution of a measure
The four-hour rule was based on two large retrospective studies, one of which appeared in the March 22, 2004, Archives of Internal Medicine. That study concluded that receiving a dose of antibiotics within four hours of arriving at the hospital was associated with “a 15% relative reduction in 30 day mortality,” which calculates to a 1.1% absolute reduction.
“A 1% reduction in mortality from delivering antibiotics more quickly would save a substantial number of lives,” says Dale Bratzler, DO, MPH, co-author of the study.
Apply that percentage to the CAP population, and the numbers are impressive. More than 700,000 Medicare patients each year are admitted to hospitals with pneumonia, making it one of the most frequent causes of hospitalization. Pneumonia is also the principal reason for more than 500,000 emergency department visits by Medicare patients each year.
Looking back, it made sense to try to reduce morbidity and mortality from CAP by encouraging faster antibiotic administration. “With a variety of infections, it is dogma that if you give antibiotics quicker, patients have better outcomes,” adds Dr. Bratzler, who is medical director of the Oklahoma Foundation for Medical Quality. “We have no reason to suspect that it would be any different with community-acquired pneumonia.”
Problems crop up
The CMS made the four-hour window one of its core measures, but it wasn’t long before studies began uncovering problems.
One study in the July 2006 issue of Chest, for instance, noted that initial “diagnostic uncertainty” was associated with a significant number—22%—of CAP patients. Another Chest study, this one in June 2007, described how the four-hour reporting rule resulted in more patients being misdiagnosed with CAP and given antibiotics they didn’t need.
The latter study, conducted at St. John Hospital and Medical Center in Detroit, found that the number of admitting CAP diagnoses that matched the final diagnoses at the end of patients’ hospitalizations dropped from 76% in 2003, before the four-hour rule took effect, to 56% in 2005.
Over that same period, St. John also saw huge increases both in the number of patients mislabeled as having CAP and in the number of patients without pneumonia who were given antibiotics early in their hospitalization.
“The only thing that changed was that the four-hour rule was introduced,” explains Mohamad G. Fakih, MD, medical director of infection control at the hospital and senior author of the 2007 Chest study. “I think the push to comply with this guideline forces people to just give the antibiotic if they are not sure.”
Moreover, he adds, because current CAP guidelines call for two antibiotics—to cover both typical and atypical organisms—unnecessary antibiotic usage was exploding.
“Now someone who doesn’t have pneumonia is getting two antibiotics,” Dr. Fakih points out. “ER doctors are not trying to do the wrong thing, but they are being pressured, and there is still the perception that it doesn’t hurt to give the antibiotics.”
That observation is cropping up across the country. “Anecdotally, no question, we keep hearing from our cardiologists: ‘Why are all my heart failure patients on ceftriaxone?’ ” says hospitalist Bradley Sharpe, MD, assistant chief of the medical service at the University of California, San Francisco (UCSF).
Real world concerns
Overprescribing is only one of the problems, according to Dr. Sharpe. In most busy emergency departments, he explains, meeting a four-hour goal “turns out to be unbelievably complicated.”
He recalls a recent UCSF patient who counted against the hospital on the antibiotic timing measure, even though few would argue that the patient received suboptimal care. The man presented with chest pain and shortness of breath but no fever or cough.
Because his ECG “showed some possible changes consistent with a myocardial infarction,” the patient was quickly taken to the cardiac cath lab for a presumed MI—in an effort to comply with a time-to-balloon quality measure for acute MI, Dr. Sharpe points out. Only when the catheterization turned out to be normal did the patient return to the ED, have more history and a chest X-ray taken. He then received a diagnosis of community-acquired pneumonia.
“It’s appropriate to take care of a possible MI first, but we were dinged for not appropriately treating the patient,” says Dr. Sharpe, who chairs his hospital’s pneumonia performance improvement committee. “Atypical presentations are not uncommon, and sometimes making the diagnosis of pneumonia just takes time.”
Then there is the challenge of ED volume, points out UCSF emergency physician Christopher Fee, MD.
Dr. Fee co-authored a study published in the November 2007 Annals of Emergency Medicine, that found a consistent decline in the likelihood of a CAP patient receiving antibiotics within four hours with each additional patient in the ED. Only 61% of CAP patients in that study received antibiotics within the four-hour window.
But as Dr. Fee notes, feasibility is only part of the problem.
“Some of us aren’t sure that the time to antibiotics is really that crucial,” he says. “Changing the measure to six hours makes no more sense than keeping it at four.” Instead, Dr. Fee says, he agrees with 2007 guidelines released jointly by the Infectious Diseases Society of America and the American Thoracic Society. Those guidelines recommend antibiotics as soon as possible after a diagnosis is made, but without any specific timeframe.
“That’s an acknowledgement,” Dr. Fee says, “that the core measure for time to first antibiotic dose, the data upon which it’s based and the patient population to which it has been applied are flawed.”
There is also the unproven but widely rumored suspicion that at least some emergency departments triage to favor potential pneumonia patients—even over individuals with other serious illnesses—so they will do well on publicly reported measures.
“It is not clear,” Dr. Sharpe concludes, “that all this effort to reduce the time to administration of antibiotics has enough benefit in terms of mortality in pneumonia patients to outweigh all the hazards and risks.”
New tracking and reporting requirements
In the last few months, both the CMS and the Joint Commission—along with scores of private payers that have instituted pay-for-performance programs—have begun asking hospitals to send two sets of data: how many CAP patients receive their first dose of antibiotics within four hours and within six hours of arrival. And beginning this month, the CMS will stop reporting how well hospitals meet the four-hour measure on its public Hospital Compare Web site. The agency will instead report hospital performance on the six-hour measure.
According to Dr. Bratzler, the possibility that hospitals prioritize patients based on public reporting or pay-for-performance requirements is a concern that haunts all performance measures.
And even Dr. Bratzler, who is one of the authors of the four-hour timing rule, acknowledges that the measure needs to be tweaked. He points out that the rule was initially written as a quality improvement measure to help hospitals figure out how to reduce disturbing delays in care, not as a tool to reshape care. The four-hour window, he adds, was not intended to be used as a test of a hospital’s quality or to determine either its public reputation or reimbursement.
While Dr. Bratzler says he’s concerned about “unintended consequences” like hospitals prioritizing patients based on performance measures instead of acuity, he notes that some hospitals have benefited tremendously from the four-hour measure. That’s particularly true, he says, when hospitals “focus on changing their systems of care within the ED, on throughput and on timely receipt of test results.”
Once those improvements take place, Dr. Bratzler adds, “I think the pressure comes off the antibiotic timing measure.”
Deborah Gesensway is a freelance writer reporting on U.S. health care from Toronto, Canada.