A mixed-methods approach to assessing actual risk readership on branded drug websites

Despite FDA requirements that drug benefits and risks be presented in a fair balance' manner, little is known about the extent to which individuals actually seek and process risk information. Self-report measures of drug risk reading show mixed results. This exploratory study investigates this topic within the context of a branded prescription drug website for a fictional seasonal allergy drug. We use a mixed-methods approach involving eye-tracking, survey and qualitative data. While approximately 80% of participants (N=29) claimed to have read half or more of the risk information, eye-tracking measures revealed limited to no risk reading when given the opportunity to freely view the site without explicit directions to read risk information. Survey results showed minimal unaided recall and modest aided recall of risks. To gain insight into why participants failed to seek drug risk information, we analyzed retro-active think aloud post-task interviews for potential explanations and looked for expressions of social desirability, information avoidance, optimism bias and familiarity. Perceived familiarity with the health condition, its risks and companion drug options surfaced as the primary explanation for failure to seek, and subsequently process, the risk information. This perceived familiarity prompted participants to ignore the risk information and discount the commercial information source. We offer practical recommendations to encourage risk reading in light of our findings.