Effects Of Internet-Based Tailored Advice on the Use of Cholesterol-Lowering Interventions: A Randomized Controlled Trial

Background: Elevated low-density lipoprotein (LDL) cholesterol is a leading risk factor for cardiovascular disease. Despite the availability of proven interventions to lower LDL cholesterol, their use remains subobtimal. Many websites provide interactive, tailored advice on cardiovascular risk in an attempt to help bridge this evidence-practice gap, yet there is little evidence that provision of such a tool is effective in changing practice. Objectives: The objective was to define the effects on use of cholesterol-lowering interventions of a consumer-targeted tailored advice website. Methods: This was a prospective, double-blind, randomized controlled trial open to any adult Australian with access to the Internet. A total of 2099 participants were randomized. Of these, 45% were male, the mean age of all participants was 56, and 1385 (66%) self-reported hypercholesterolemia. Follow-up information was obtained for 1945 (93%). Participants completed a brief online questionnaire. Individuals assigned to intervention received immediate, fully automated, personally tailored advice (based on current guidelines) regarding the need for commencement of statin therapy, increased statin therapy in those already on treatment, and nondrug intervention strategies. Control group participants were directed to static Web pages providing general information about cholesterol management. Results: The primary outcome was the proportion of participants that commenced or increased use of prescribed cholesterol-lowering therapy. Of the total 2099 randomized participants, 304 (14%) met eligibility criteria for cholesterol-lowering therapy but were not prescribed treatment, and 254 (12%) were prescribed treatment but were not achieving the recommended target level. Treatment was commenced or increased in 64 (6.0%) of the 1062 intervention group participants and 79 (7.6%) of the 1037 control group participants (% difference = -1.6%, 95% confidence interval [CI] -3.75 to 0.57, P = .15). No differences were found between the randomized groups for the secondary outcomes of "discussed treatment with a health professional" (% difference = -3.8%, 95% confidence interval [CI] -8.16 to 0.19, P = .08), "had their cholesterol checked" (% difference = -1.5%, 95% CI -5.79 to 2.71, P = .48), "had their blood pressure checked" (% difference = 1.4%, 95% CI -2.55 to 5.34, P = .49) or made a lifestyle change (P values between .49 and .96). Conclusions: Despite providing specific carefully tailored advice, this website had no detectable effect on cholesterol management strategies. This finding raises considerable uncertainty about the value of Internet-based tools providing tailored advice directly to consumers. Trial Registration: NCT00220974; http://clinicaltrials.gov/ct2/show/NCT00220974 (Archived by WebCite at http://www.webcitation.org/5sdq63rrY)