Methods: A randomized controlled trial was conducted of the decision-aid verses a control intervention. Postmenopausal subjects were recruited from 4 primary care clinics affiliated with the Department of Veterans Affairs in the Midwest region of the United States between May, 2002 and October, 2003. The decision-aid used a self-navigated computer modality and included background information, a value elicitation exercise, tailored risk estimates, videotaped patient testimonials, and a new research web site link. Results from the Women's Health Initiative (WHI) clinical trials were on the new research link. The control consisted of an educational pamphlet that was reviewed with the participant and included printed results from the WHI clinical trials. Baseline assessments of knowledge and current HT use were obtained prior to the intervention. All participants were scheduled to see their primary care provider 3 weeks post- intervention to have the opportunity to discuss treatment options. The following outcomes were assessed 3 months post-intervention; current use of HT; knowledge about menopause and HT, satisfaction with decision (Satisfaction with Decision Scale), and decisional conflict (Decisional Conflict Scale and Sub-Scales).

Results: There were 177 participants in the study (22% participation rate). Of these, 27% were minority race, 24% had no college experience, and 32% had an annual income of <$0,000. A temporal trend to discontinue HT was noted (35% were on HT at baseline and 27% at follow-up) but did not differ between groups, p=0.83. Knowledge improved for both groups but greater improvement was seen in the control group (p=0.03). There was no difference in satisfaction with decision between the intervention (mean 4.37, SD 0.49) and control (mean 4.37, SD 0.51) groups or in decisional conflict between the intervention (mean 4.37, SD 0.49) and control (mean 4.37, SD 0.51) groups, p=0.99. Among those who used HT at baseline, lower decisional conflict in the factors of uncertainty sub-scores were noted, (p=0.047).

Conclusions: The negative results of this RCT suggest that a technically complex decision-aid may not improve the decision process compared to simpler interventions. More work is required to understand the elements of a decision-aid that improve the decision-making process among primary care patients.