Hundreds of times a day, doctors weigh what is the best next step with lab tests, imaging exams, blood transfusions, medications – the list can go on and on. But the goal is choosing the option that medical research best supports.
The reality, though, as the book “How Doctors Think” explains, is that doctors are as human as the rest of us, said Dr. Harry C. Sax, executive vice chair of surgery at Cedars-Sinai Medical Center in Los Angeles. Human beings tend to be influenced by their latest experience, he said, not necessarily by hard data or evidence.
The book talks a bit about how doctors are affected by the outcomes of their last three patients, Sax said, and if they decided on research-supported actions for those patients but the patients didn’t fare well, then the doctors likely will go to another tool in their kit for the next patient.
But what if doctors didn’t deviate? What if they stuck with only the evidence-based options? How would patients do?
The answers to those questions were revealed Wednesday in a study published in the American Journal of Managed Care by Sax and other physicians and research investigators with Cedars-Sinai and their partner, Optum Advisory Services, an information technology business based in Eden Prairie, Minnesota.
To get the answers, the research team embedded electronic reminders into the hospital’s electronic health records, showing physicians in real time when their care instructions deviated from evidence-based medical guidelines, said Sax, who was a senior author of the study. Then they assessed what would happen for patients whose doctors adhered to all guidelines and those whose doctors didn’t adhere to any of them.
They looked at nearly 26,500 patients admitted to Cedars-Sinai over three years.
The findings showed striking differences in outcomes. The odds of complications increased 29 percent for patients whose doctors did not uniformly follow the alerts, the study found, and the likelihood of hospital readmission was 14 percent higher for them.
Costs for patients whose doctors didn’t adhere to guidelines was 7.3 percent higher – $944 per patient – and their length of stay was 6.2 percent longer than for those whose doctors strictly followed the guidelines, the data showed. The investigators accounted for differences in the complexity and severity of the patients’ illnesses when considering the data.
“Sometimes the best care for certain patient conditions means doing less,” said another senior author of the study, Scott Weingarten, the chief clinical transformation officer at Cedars-Sinai. “We have seen that real-time aids for clinical decision-making can potentially help physicians reduce low-value care and improve patient outcomes while lowering costs.”
Once the study began, here’s what would happen: A doctor would type an order – a sedative for a senior patient who couldn’t sleep, for example – into the electronic health record, and when the doctor pushes the button to submit the order, a separate computer program sees it, gathers relevant vital statistics from the patients’ health record and weighs whether the order meets evidence-based guidelines. If it doesn’t, the program sends the doctor a real-time alert explaining why.
While a sedative might help an older patient sleep, investigators said, it also can increase their risk of falls. The program doesn’t say the doctor can’t order it, but it explains the guidelines and seeks the reason for overriding the guidelines.
“Anytime we provide good information in real time, we react to that more appropriately,” Sax said. “Think about when you’re driving your car and your GPS is telling you, ‘You’re going to make a right turn in 500 feet.’ You’re more likely to do the best thing than, if an hour or two earlier, you were looking at a map, and now you are trying to remember when to make the right turn.”
Doctors sometimes order tests or medications because they’re in defensive mode and want to avoid potential malpractice, but the embedded alerts put some firepower behind the decision to stick with evidenced-based recommendations and gives the doctor permission to stop second-guessing.
“When you have evidence-based information that says this test is not needed or this test has potentially a down side, that makes us more comfortable not ordering something purely as a defensive measure,” said Sax, a senior author of the study.
For the study, the investigators used medical guidelines from an initiative called Choosing Wisely, which indicates common tests and procedures that the American Board of Internal Medicine Foundation has agreed may not have a clear benefit.
One drawback of the study, Sax said, is that they didn’t look at outcomes for patients whose physicians followed some but not all of the guidelines. They weighed only patients whose doctors adhered to the guidelines against those whose doctors didn’t.
Cedars-Sinai and Optum chose to display only 18 alerts out of hundreds of possibilities from Choosing Wisely because they didn’t want doctors to get fatigue from constant pop-ups and start ignoring them. To select the real-time alerts they would use, the researchers let the program run invisibly in the background for a while and assessed which alerts triggered most often.
“In the U.S., we spend 18 percent of our GDP on health care,and our outcomes are not as good as many other countries that spend less,” Sax said. “There could be a variety of reasons for that. What we’re excited about here is that this is going to allow physicians to make good decisions in real time, provide better care at lower cost and better outcomes. Those are really the key takeaways here. We’re helping the clinicians.”