How Doctors Handle Contradictory Evidence

Based on research by Ajay Kalra, Shibo Li (Indiana) and Wei Zhang (Iowa State)

Key findings:

• People update their beliefs based on various information sources, but the information they receive is often biased and, on the whole, contradictory.
• When new negative information is not clear-cut, specialist physicians are less affected, because these traits improve their ability to discern the nature and quality of biased information.
• Firms could forfeit revenue when they don’t understand how different types of customers respond to contradictory information.

We’re surrounded by health headlines that contradict each other. One might say, “Organic food reduces cancer risk,” while another warns, “Organic food consumption is linked to higher cancer risk.” Faced with conflicting information, consumers must decide what — or whom — to believe.

Physicians face similar dilemmas. They must decide whether to continue prescribing a medication when new evidence is conflicting or unclear.

Consider Avandia, a once-leading diabetes drug. On the same day the New England Journal of Medicine published a 2007 study linking Avandia to increased cardiovascular risk, it also ran an editorial criticizing the study’s methods. Other major sources, including the FDA, published conflicting guidance — some supporting continued use, others raising concern.

With no clear consensus, physicians had to decide: keep prescribing, or switch to an alternative?

Under circumstances where new negative information about a drug is not clear-cut, how might a physician’s belief about the quality of the drug and, ultimately, future prescriptions be influenced? The answer may depend on whether the physician is a generalist or specialist.

To understand how physicians respond to unclear negative information, Rice Business marketing professor Ajay Kalra and co-researchers Shibo Li (Indiana) and Wei Zhang (Iowa State) studied physician behavior following the Avandia controversy. They developed a model showing how doctors draw on multiple, often biased, information sources — including news, sales reps and patient feedback — to update their beliefs about a drug’s safety and efficacy.

The researchers assumed — accurately — that physicians often confront biased and contradictory information. Journalists don’t just report facts; they often editorialize. Likewise, pharmaceutical sales reps are trained to promote their own products, offering free samples and persuasive talking points. In this noisy environment, it’s hard to separate fact from spin.

Kalra and his team collected data from January 2006 to October 2007 — before and after the Avandia controversy erupted. They tracked the prescribing patterns of 1,500 physicians — specialists, hospital-based primary care doctors and office-based PCPs — along with marketing activities from pharmaceutical reps.

They also analyzed 112 print and online articles related to Avandia, classifying 65 as negatively biased and 47 as positively biased.

The researchers found that different types of physicians responded differently to the conflicting news. Specialists were less likely to let the noisy information landscape influence their prescription decisions than generalists (PCPs).

Why the difference? The researchers speculate that specialists were better able to critically filter and assess new information because due to greater expertise, more peer interaction and higher self-efficacy. By contrast, generalists, particularly those in private offices, tend to have lower risk tolerance and are more likely to rely on outside sources, including sales reps and patients.

Drugmakers also responded to the Avandia controversy — but not in the most effective way. They reduced marketing to office-based PCPs, even though those doctors were the most influenced by the negative news. They maintained or increased marketing to specialists and hospital-based PCPs, who were less likely to change their prescribing behavior.

The result? Many office-based PCPs stopped prescribing Avandia and switched to metformin, an older competitor.

In short, specialists and generalists don’t process information the same way. What does this mean for patients? When new information about a drug emerges — especially if it’s contradictory — your doctor’s response may depend on their professional context and background. 

Kalra, A., Li, S., & Zhang, W. (2011). “Understanding responses to contradictory information about products.” Marketing Science 30(6), 1098-1114.