Taxpayer-funded research findings in the last decade have altered clinical knowledge about the best way to treat patients, but five obstacles, the biggest of which is perverse financial incentives, keep physicians from adopting them for their patients.
That's the message from RAND researcher Justin Timbie and colleagues, who report their conclusions in the October issue of Health Affairs.
"The difference between reimbursements in some of these treatments is so huge, it can sway decision-making at the point of care," Timbie says. "This misalignment of financial incentives that slow adoption of comparative effectiveness research evidence is probably the biggest, thorniest challenge we have today."
He adds, "financial interests can influence how physicians and patients interpret the results of new research, through marketing and through messaging of different sorts.
"Additionally," he says, "the industry has funded opinion leaders and has conducted marketing campaigns to spread particular interpretations of these studies' results." Biases also impact the formalization of guidelines that professional societies produce to guide physicians, he says.
Four other bottlenecks keep doctors from adopting best practices.
• Ambiguity of study results that hamper decision-making and lead some clinicians to contradictory conclusions.
• Cognitive biases in the interpretation of new information. These include "confirmation bias," the tendency to embrace evidence that confirms preconceived ideas; "pro-intervention bias," the tendency to do something over nothing even if the marginal benefit is small; and the "pro-technology bias," the tendency to believe that newer is better.
"Explaining this phenomenon, one cardiologist told us that by the time of angiography, a patient has become reduced to 'an anatomic appearance on a picture,' implying that the presence of a blockage dwarfs consideration of the relative benefits and harms of the procedure or patient preferences."
• Failure of the research to address the needs of end users, who may be more interested at the start in whether a drug is safe than whether it is effective.
• Limited use of decision support systems by clinicians and patients. The authors wrote that integrating support tools is challenging because clinicians fear spending too much time at their computers than with their patients, and false alerts may trigger too often.
The RAND researchers picked six major studies from the last decade that produced findings that should have dramatically altered the way patients with certain conditions, such as stable coronary artery disease, schizophrenia, hypertension, or lumbar spinal stenosis, should be treated.
They then identified several steps by which traditional research should be translated into practice, for example, from interpretation to formalization—when the findings are converted to guidelines and performance measures or clinical decision support tools—to dissemination to implementation.
But after three of the studies, there was "no change in practice," and in two there was merely "some" or "little change" in practice or "slow adoption."
For example, the ALLHAT (Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial), published in 2002, found that for patients with hypertension and at least one other cardiovascular disease risk factor, "less costly, traditional diuretics are more effective than newer medicines at lowering high blood pressure and preventing some forms of heart disease," according to the National Heart, Lung, and Blood Institute.
Timbie, however, found that these less expensive diuretics "are adopted much more slowly than heavily marketed treatments that are less effective."
Similarly, the 2007 COURAGE trial, (Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation) revealed that medical therapy and percutaneous coronary intervention (PCI) did not provide patients with better survival benefit and angina relief than medical therapy alone. Yet there has been "little or no change in practice," Timbie says.
Only after one of the trials examined, the Philadelphia Bone Marrow Transplant Group, did practice change. In that study, researchers discovered that bone marrow transplantation added to both conventional and high dose chemotherapy for women with metastatic breast cancer sixth study did not improve survival than chemotherapy.
In fact, there were higher risks of serious side effects from bone marrow transplantation.
"Treatment harms are a potent driver of changes in clinical practice," Timbie says. "That's sort of a common thread, when harms are on the table, usually that is one of the main drivers in changes of practice."
The RAND team suggests that comparative effectiveness researchers should design studies with an eye to successful translation of whatever the findings are.
And they criticized bias that often is found in specialty society guideline development committees, which lack multi-specialty viewpoints and balance. Such committees should have "robust" conflict of interest policies, more complete documentation of the strength of evidence supporting each guideline, and the use of more explicit statements regarding the recommended clinical actions to take," which few now do.
Such findings might provoke pushback from physicians, Timbie acknowledged. He says, however, that they shouldn't shoulder all the blame for not doing what's in the best interest of their patients.
"We can't let patients off the hook because patients usually want the latest technology, and they don't want to hear that their physician might choose a more conservative approach."
He emphasizes that "it's not as if the providers are solely at fault. There [are] just numerous factors in play, all pushing in the same direction toward aggressive treatment of these conditions...And [there are] few incentives to choose the more cost-effective treatment."
The RAND researchers say they hope that accountable care organization and bundled payment models may stimulate greater use of evidence in clinical choices as those programs role out in the years to come.