Medical Decision Making (MDM)
September/October 2011; 31 (5)
http://mdm.sagepub.com/content/current

Editorials
Katrina Brown and Nick Sevdalis
Lay Vaccination Narratives on the Web: Are They Worth Worrying About?
Med Decis Making September/October 2011 31: 707-709, doi:10.1177/0272989X11419664

Extract
The Internet allows us more access than ever before to the unadulterated anecdotes and opinions of our fellow laypeople. Our decisions—about health care or parenting, for example—were once based on advice from experts, plus perhaps testimonies from a small pool of friends and family, or a finite number of narratives filtered through the press or television. Now the proliferation of social networking and user-generated content in the age of Web 2.0 1 puts at our disposal a huge and often unmoderated bank of online material. Google searches in June 2011 for “health discussion forum” and “mothers discussion forum” yielded 310 million and 62 million hits, respectively—and we know that decision makers do access these online resources. 2, 3

What is less clear is whether, why, and exactly how lay narratives from online forums are associated with real-life decisions. In this issue of Medical Decision Making ( MDM), Betsch and colleagues 4 report an investigation of how lay narratives are used in decision making about vaccination. Using a fictional disease and vaccine context, a mock Internet bulletin board setup, and an undergraduate sample, the authors varied the relative frequency, emotionality, richness, and correlation with official risk estimates of narratives reporting vaccine adverse events and assessed perceived adverse event risk and vaccine uptake intention. Betsch and others found that a higher frequency of narratives reporting vaccine adverse events increased perceived vaccine risk and decreased vaccine uptake intention, that narrative frequency affected risk perception and uptake intention to a greater extent than did statistical information, and that emotionality in narratives increased risk perception, whereas richness had no impact.

Cornelia Betsch, Corina Ulshöfer, Frank Renkewitz, and Tilmann Betsch
The Influence of Narrative v. Statistical Information on Perceiving Vaccination Risks
Med Decis Making September/October 2011 31: 742-753, first published on March 29, 2011 doi:10.1177/0272989X11400419

Abstract
Background. Health-related information found on the Internet is increasing and impacts patient decision making, e.g. regarding vaccination decisions. In addition to statistical information (e.g. incidence rates of vaccine adverse events), narrative information is also widely available such as postings on online bulletin boards. Previous research has shown that narrative information can impact treatment decisions, even when statistical information is presented concurrently.

Objectives. As the determinants of this effect are largely unknown, we will vary features of the narratives to identify mechanisms through which narratives impact risk judgments. Methods. An online bulletin board setting provided participants with statistical information and authentic narratives about the occurrence and nonoccurrence of adverse events. Experiment 1 followed a single factorial design with 1, 2, or 4 narratives out of 10 reporting adverse events. Experiment 2 implemented a 2 (statistical risk 20% vs. 40%) × 2 (2/10 vs. 4/10 narratives reporting adverse events) × 2 (high vs. low richness) × 2 (high vs. low emotionality) between-subjects design. Dependent variables were perceived risk of side-effects and vaccination intentions.

Results. Experiment 1 shows an inverse relation between the number of narratives reporting adverse-events and vaccination intentions, which was mediated by the perceived risk of vaccinating. Experiment 2 showed a stronger influence of the number of narratives than of the statistical risk information. High (vs. low) emotional narratives had a greater impact on the perceived risk, while richness had no effect.

Implications. The number of narratives influences risk judgments can potentially override statistical information about risk.