One strategy to keep community-acquired pneumonia out of the hospital A look at how Banner Health created a guideline on how to treat the condition–and then convinced physicians to use it by Edward Doyle
Published in the August 2004 issue of Today's Hospitalist
While hospitalists tend to focus on changing systems in the inpatient setting, some are looking to make changes in the outpatient setting. One goal is to create evidence-based algorithms to keep patients from being admitted to the hospital in the first place.
The challenge, however, lies in convincing physicians who work in a different setting to use guidelines created by inpatient physicians. Most primary care physicians are overwhelmed by the daily grind of outpatient practice.
When the Phoenix region of Banner Health decided to try to improve how its physicians treat community-acquired pneumonia, it
For simple pneumonia in non-seniors, days per thousand dropped from 11.4 to 1.53.
tackled both challenges. As a result, the health plan combined science and communication to craft a guideline that would eventually slash its length of stay and its use of hospital resources.
Banner Physician Hospital Organization (PHO), a division of Banner Health, spearheaded the program in the late 1990s after discovering tremendous variation in how its physicians were treating community-acquired pneumonia. (Banner PHO manages commercial and senior care risk plans for the health plan.)
“We realized these patients coming into the hospital setting were either not treated or treated with wrong antibiotics on the outside,” explains Andre Abbate, MD, inpatient medical director for Banner PHO. “When they hit the ER, they were in dire straits, and they usually ended up in the telemetry unit or in the ICU on a ventilator.”
In 1998, Dr. Abbate began working with primary care physicians and subspecialists. They eventually create a guideline to improve how Banner’s outpatient physicians treated patients presenting with community-acquired pneumonia.
That guideline divides patients into three basic categories: mild, moderate and severe. For all three groups, the guideline calls for physicians to examine the patient, obtain a chest X-ray if necessary, document fever and any comorbid conditions, and try to obtain a sputum sample.
For the mild cases that are typically seen in the outpatient environment, the guideline urges physicians to prescribe one of four antibiotics: clarithromycin, doxycycline, amoxicillinclavulanate or levofloxacin. Dr. Abbate says that in choosing antibiotics for the guideline, physicians examined antibiograms to find out what’s working. “Our antibiograms gave us the know-how to say that amoxicillin was not helping these patients,” he explains.
For patients with moderate cases of community-acquired pneumonia, the guideline calls for physicians to administer 1 gram of ceftriaxone in the office setting either intramuscularly or intravenously. They are also urged to write a prescription for one of the same four antibiotics given to patients with mild cases of the disease.
If the physician feels that additional therapy with ceftriaxone is indicated, the patient can return to the physician’s office for up to three additional days of medication.
For severe cases of community-acquired pneumonia that present to the ER, the guideline calls for intravenous ceftriaxone with clarithromycin or doxycycline. Dr. Abbate says that patients can receive levofloxacin if there is a documented penicillin or beta-lactam allergy.
Dr. Abbate says that the guideline was implemented in 1998, but its full impact wasn’t immediately felt. Within two years, however, the guideline dramatically reduced the incidence of community-acquired pneumonia in the hospital setting.
“If you look at simple pneumonia on the commercial side,” he says, “we went from 11.44 days per thousand to 1.53 days per thousand. On the senior side, we went from 300.81 days per thousand to 52.56 days per thousand. That’s a tremendous drop, and we did nothing except implement the guideline.”
Banner PHO uses two main hospitals for its inpatient network. The use of guidelines has helped reduce overall length of stay at both facilities. In 2001, the total in-network days per thousand for seniors was 1,086. By 2002, days per thousand for that population were down to 1,065, and in 2003, it had dropped to 985.
“If you had told me in 1999 that our utilization would be under 1,000 days per thousand on the senior side in 2004,” Dr. Abbate says, “I would have thought you were kidding me.”
Dr. Abbate is quick to add that while the guideline has helped Banner drive down its length of stay and overall resource utilization, other initiatives have also made an impact. For example, Banner’s emergency rooms now use a protocol that allows them to deliver antibiotics to patients within two hours.
While the evidence behind Banner’s community-acquired pneumonia guideline is sound, Dr. Abbate spent hundreds of hours explaining it to physicians. “The biggest arguments you hear from physicians when you develop guidelines,” he says, “is that they don’t like change and they don’t like to be told what they can and can’t use in treating patients.”
“The guideline was considered pushy,” Dr. Abbate adds, “but physicians quickly realized that these patients weren’t coming back into the office and sitting in their waiting rooms. They realized that the guideline helped free up time for them.”
To get physicians’ attention, he regularly talked to everyone from intensivists to emergency department physicians to outpatient physicians. During a three-month period in 1999, for example, he paid short visits to more than 60 primary care offices.
On the inpatient side, Dr. Abbate would visit with the hospitalists and emergency physicians in the morning and at night so he would have a chance to talk about the guideline. “I would show up at 6 in the morning when the night doctor was still on,” he says, “and then I would show up at 4 in the afternoon at shift change. I would even show up at 10 at night when the ER swing shift came on.”
Today, when a new physician joins the PHO either as member of the hospital staff or an outpatient practice that contracts with Banner, Dr. Abbate visits that person to talk about Banner’s guidelines. “If you don’t make that time commitment,” he says, “you will fail.”
Despite all the time he spent talking to physicians, Dr. Abbate acknowledges that there were some rough patches. When physicians refused to follow the guideline, for example, he would visit or call to ask why. In many instances, he found that it was patients, not physicians, who balked at the drugs, and usually because of cost.
As a result, Dr. Abbate convinced some local drug company representatives to leave extra samples of antibiotics with physician offices. In many cases, drug detailers now leave one sample dose of an antibiotic per month for every doctor in the office.
Dr. Abbate also learned some hard lessons about how often physicians need to be reminded about guidelines. Immediately after the guideline was rolled out, community-acquired pneumonia admissions dropped. But a year later, to everyone’s surprise, the number of those admissions started to creep back up.
Dr. Abbate says the problem was easily corrected: He had stopped regularly talking to physicians about the importance of using the guideline.
“We didn’t make that phone call, we didn’t visit them,” he recalls. “I have learned that every six months, I need to make a phone call to the physicians I’ve already seen to remind them about the guideline. I need to let them know that we’ve updated it, especially when we change antibiotic choices.”
Dr. Abbate notes that Banner Health has rolled out sections of the community-acquired pneumonia guideline in other parts of the country. He has heard that in some areas, however, physicians have resisted the guideline.
“Some physicians think it is something the administration is trying to force on them,” he says. “We have learned the hard lesson that if we don’t have a champion on our team from every portion of the health care continuum, physicians won’t use the guideline.”