Study to Assess if Internet-Based Tailored Advice Could Modify Behaviour to Improve Health

The Internet-based Cholesterol Assessment Trial (I-CAT) is a masked, randomised, controlled trial that will include at least 3,700 consumers followed for between 8 and 16 weeks. All consumers that participate in the study will ultimately be offered the I-CAT service, which provides individually tailored information about the management of cholesterol-related risk via an interactive website. However, by offering the service immediately to one group (intervention) and delaying it by 8 weeks in the other group (control) it will be possible to precisely and reliably determine the effects of the I-CAT service on consumers and their use of cholesterol lowering interventions.

There will be two types of individuals involved in the study: 'index cases' and 'friends or relatives' of the index cases. The index cases will be randomised to intervention or control and will comprise the study participants on which the main outcome analyses will be based. Friends and relatives of the index case will only be involved insomuch as they will be asked to visit the site by the index case as part of the evaluation of the effectiveness of the intervention.

Index cases will comprise mainly individuals already known to be at high risk of a cholesterol-related disease event - for example individuals with established vascular disease, individuals with multiple risk factors and individuals already using cholesterol-lowering therapy. Index cases will be recruited using advertising materials placed at a range of sites including hospital in patient services, hospital out patient services, general practices, pharmacies, pathology services and the Internet.

All individuals that provide informed consent (index cases and friends or relatives) will be asked to answer as many questions as possible from the XX questions in the baseline questionnaire. The questionnaire will seek information about the individual's cardiovascular disease history, their risk factors, their cholesterol levels, their use of any medications or other strategies to reduce their cholesterol-related cardiovascular risk and any family history of cardiovascular disease or high cholesterol levels.

Once baseline data collection is complete index cases will be randomised to either intervention or control. Randomisation will be done automatically in real time by a central computerised service. The intervention group will receive immediate feedback from the I-CAT service. The control group will receive general information about cholesterol-related disease risk derived from existing materials prepared by the National Heart Foundation of Australia. Hyperlinks to a range of relevant websites will be provided.

In addition, after returning to the website at 8 weeks and providing follow-up information, all control group participants will receive the same tailored feedback from the I-CAT service as was given to the intervention group participants at randomisation. In this way, all study participants will ultimately receive the same advice but by providing one group with immediate advice and one group with delayed advice it will be possible to precisely and reliably determine the value of the I-CAT service to consumers.

All primary analyses will be of comparisons between the index cases in the intervention group and the index cases in the control group conducted according to the principle of intention-to-treat. For each outcome the analyses will compare the proportions of index cases reporting the outcome (or the mean values for each outcome) between randomised groups. Comparisons of proportions will be done using Chi-square tests and comparisons between means will be done using t-tests.