Study: Side effect warnings may deter prescription compliance

Failure to follow prescribed drug regimens costs society nearly $100 billion a year and leads to an estimated 125,000 deaths, according to statistics cited by the Department of Health and Human Services' Office of the Inspector General.

Sean Young, a doctoral student in psychology at Stanford, and alumnus Daniel Oppenheimer, an assistant professor of psychology at Princeton University, have found that the way side-effect warnings are worded may exacerbate the problem. Their paper, "Presenting Risk Information," was published last month in the journal Clinical Therapeutics.

"We found that the way drug information is presented to people leads to a consistent overestimation of the likelihood of negative side effects, which in turn increases the likelihood that they will not take their medication as prescribed," Young said.

When advertising drugs, pharmaceutical companies are required by law to disclose side effects associated with their products. Since direct-to-consumer advertising became legal in the 1990s, companies have consistently preferred "semantic" indicators, which name the side effects without being specific about how often they occur, Young said. The problem is that consumers are not trained to interpret pharmaceutical literature, he continued, and drug company advertising practices are of little help for making informed decisions. "Right now we don't take into account people's perceptions when they make these decisions," he said. "Information needs to be tailored to people's psychologies."

Research findingsThe paper presents the results of three studies, carried out at Stanford, that were designed to isolate factors that might cause people to avoid taking their medication. In the first experiment, 40 people were asked to estimate the occurrence of "commonly reported" side effects, like headaches and nausea, associated with certain drugs. Participants typically estimated that as many as 20 percent of drug takers would experience such side effects, when in fact the reported incidence in clinical trials was closer to 2 percent, the researchers said.

In a second study, Young and Oppenheimer provided 31 people with percentage-risk statistics instead of stock phrases such as "side effects may include" and "some people may experience." The result was that participants indicated they would be much more likely than those in the first study to take the medications as prescribed.

In a third study, 120 participants were given actual magazine advertisements for medications. Bolstering the researchers' earlier findings, the results showed that a perception of risk led to decreased intention of complying with drug prescriptions.

"Health care providers assume that as long as you get information out there to people, it's less important how it's communicated," Young said.

Another problem with the practice of listing "most common" side effects is that it gives a distorted picture of the relative safety of different drug options, Young said. In the absence of hard numbers, people overestimate the dangers of drugs on the basis of the seriousness of side effects they hear about, he explained. But while alarming conditions such as hemorrhage or heart attack might be the most commonly reported side effects for a particular drug, they may still be extremely rare. "A 'most common' side effect could be fatal, yet be one in a million," he noted.

The researchers assert that drug company marketing practices may be designed to increase people's purchase of prescription drugs but are actually making people fearful of the drugs. "This research suggests that framing affects estimates of risk, and that current methods of disclosing information are non-optimal," Young and Oppenheimer write in the article.

Recommendations to increase complianceThe results of the study lead Young and Oppenheimer to make several suggestions that might contribute to increasing medical compliance. Pharmaceutical companies distributing drug information should give statistics in percentages and be specific about who is at risk, they said. For example, elderly people and patients on other medications typically report side effects more often than other patients. While this information is available, it is not widely distributed in advertising, Young said.

Another option is to standardize the wording of side-effect warnings, as is done in the European Union. Side effects could be rated anywhere from "extremely rare" to "common," with labels indexed to percentage risks. A third method, often used in psychology, is known as a "ladder scale," which compares medication risks with risks of other rare events. "To give someone an idea of just how uncommon something is, you might compare it to the chances of getting struck by lightning," Young said.

In the future, Young argued, the discipline of social psychology, which looks at how people interact with one another, has much to offer policymakers aiming to create a better-informed public. "One of the big questions in medical decision-making is whether we want a paternalistic system, which gives people less choice," he said. "Our society has steered away from that—we take the stance that there should be communication between patient and provider to reach a joint decision."

Melissa Fusco is a science-writing intern at Stanford News Service.