Methods: Men over 50 years (N=603) scheduled for a preventive physical examination were recruited and randomly assigned to one of four Internet tools prior to their appointment: a didactic decision aid providing information about PSA screening (Group 1); a decision aid using a chronic illness trajectory model for prostate cancer followed by a one item visual analog and time trade-off exercise (Group 2); both the didactic decision aid and the chronic illness trajectory (Group 3); or links to reputable public prostate cancer information websites (Center for Disease Control and American Cancer Society; Group 4). Participants completed questionnaires before and after their physical exam. Primary dependent measures were knowledge about prostate cancer, decisional conflict, PSA test choice and prostate cancer treatment preferences.

Results: There were no baseline differences between groups: average age was 59 years (SD = 5.4); the majority were Caucasian (86.1%), married (77.9%) and college educated (70.5%). Participants assigned to view public websites (n= 150) were significantly less likely to review information (76.7% reviewed) than those assigned to the didactic decision aid (n=155; 84.5% reviewed), the chronic illness trajectory decision aid (n=152; 86.2% reviewed) or the combination condition (n=146; 87.7% reviewed; p<.05). At the follow-up evaluation, knowledge scores were significantly lower among those randomized to review public websites, compared to all other groups (p<.001). Participants randomized to Groups 1 and 3 were more likely to feel that they knew the risks of prostate cancer screening (p<.001), knew how important these risks were to them (p<.001) and felt like they made an informed choice (p<.004). There was an overall reduction in preferences for PSA screening from pre- to post-test (p<.001), but no significant group differences in the rates at which men chose PSA screening. Men assigned to review public websites were less likely to indicate a treatment preference for watchful waiting (p<.03).

Conclusion: Public websites about prostate cancer provide less effective decision support than specially designed decision aids. Didactic decision aids lowered decisional conflict compared the chronic illness trajectory decision aid; however, screening decisions and treatment preferences were similar among the three experimental conditions.