Following the Evidence

For instance, if a doctor was seeing a woman with a bladder infection, it used to be standard practice to order a urine sample and a culture, then prescribe 7 to 10 days of antibiotics. Now, the evidence shows that the most-effective way to diagnose and treat the condition is with a urine sample (no culture) and three days of antibiotics. If the doctor pulls up the patient’s record and orders both a sample and a culture, a prompt will note that a culture is not needed. When ordering the prescription, the computer will offer only options for three days of medicine that the doctor must override if it’s necessary.

These strictures are not meant to tie doctors’ hands. “Having guidelines is important, but they don’t change how medicine is practiced by themselves,” says Dr. Charles Fazio, Medica’s chief medical officer. “It’s important to change a clinic’s infrastructure so it relies less on people’s memories and more on systems that don’t forget to provide reminders about what care needs to be done.”

Those systems are not all electronic. For instance, diabetes patients should have their feet examined regularly for loss of feeling. Doing this can easily be forgotten, however. At one clinic Fazio knows of, nurses have been trained to remind diabetics to remove their shoes and socks because “doctors won’t forget to check a patient with bare feet.”

Working Out the Kinks

Those who object to EBM often call it “cookbook” medicine, because they think it takes the art out of their practice, says Dr. Nancy Jarvis, medical director of supporting best care for Park-Nicollet Health Services in St. Louis Park.

“Art is how you interpret the guidelines,” Fazio argues. “It’s how doctors relate to people and help them make choices.” For example, a doctor may not insist on strict compliance with dietary guidelines for a disabled nursing-home resident, Fazio says. It’s also necessary for doctors to navigate among multiple guidelines for patients with more than one condition.

Frenz says doctors can’t be dogmatic followers of guidelines, but must listen to “a sixth sense that can’t be replicated through clinical pathways. There are too many ambiguous cases and areas.” But, he adds, “You ignore the research at your peril.”

One doctor puts it more bluntly. “Why would you want to do it any way but the correct way?” asks Dr. Barry Bershow, medical director for quality and informatics for outpatient clinics at Fairview Health Services in Minneapolis. “Everybody does better the closer we get to delivering the right care at the right time.”

The biggest impediment to evidence-based medicine is not individuals but systems, according to Dr. Gordon Mosser, who recently retired as ICSI president. “Doctors are trained in medical school that individual performance leads to good patient outcomes, but that’s a stunted view of what happens today,” Mosser says. “To paraphrase Bill Clinton, ‘It’s the system, stupid.’ It’s not just hiring good doctors. Systems determine whether a facility has good outcomes, and systems can be improved. Most of what ICSI does today is work with hospitals and clinics to change organizational behavior.”

Fridley’s Unity Hospital recently made systemic changes that lowered its death rate. To prevent pneumonia in patients on ventilators, for instance, hospital staff made it standard practice to brush patients’ teeth to kill bacteria and make sure their heads were elevated at least 30 degrees.

“Defensive” medicine—practicing medicine in such a way to avoid criticism and legal troubles—can also hinder evidence-based medicine. “In theory, evidence changes medicine, but doctors often do things because they’re worried about lawsuits,” Frenz says. For instance, patients are admitted to hospitals for chest pain, and given MRI tests for back pain at rates not justified by research, he says.