Tuesday, October 20, 2009
Grand Ballroom, Salons 1 & 2 (Renaissance Hollywood Hotel)
Lyndal Trevena, PhD, The University of Sydney, New South Wales, Australia, Siranda Torvaldsen, PhD, The University of Sydney, University of Sydney, Australia and Jack Dowie, PhD, London School of Hygiene and Tropical Medicine, London, United Kingdom
Purpose: Given an increasingly overwhelming list of options facing patients who contemplate a ‘check-up’, we aimed to develop a web-based tool that would prioritize and personalize preventive health care options, including individual patient preferences Method: We developed a multi-criteria decision analytic (MCDA) model that combined key utilities or preferences for maximising health and well-being along with probabilities appropriate to each preventive health option. The consumer-focussed website ‘My Health Check’, is designed for 30-69 year olds who complete a questionnaire and are shown four ‘attributes’ or core utility/preference constructs. These are four shaded rectangles adjustable in length to reflect ‘real-time’ personal utilities. Focus groups determined these to be 1) Living a long life 2) Living a life without disability 3) Avoiding the difficulties preventive health activities might involve and 4) Reducing financial impact. A minimum set of ten preventive healthcare options were selected from evidence-based guidelines including lifestyle changes and cancer screening. A ‘do nothing’ option is a baseline for the model. The MCDA application is called AnnaLisa (c). Years of life lost (YLL) and Years Lived with disability (YLD) data from the Australian Burden of Disease report provided age-gender-specific population-based estimates for each preventive health option in the first two utility/preference constructs of our model. However, data is absent for the disutility of most preventive health behaviours and is likely to vary from person-to-person as is financial cost/savings depending on personal habits. A unique feature of our model is that a pre-questionnaire feeds individual option-specific probability estimates and allows for person-to-person variation in the third and fourth utility sections. Result: The model generates ranked lists of personalised preventive health options for men and women aged 30-44, 45-59 and 60-69 years. Options not relevant to the individual are dropped from the model (eg cease smoking is not relevant to a non-smoker). Rankings for maximal health gain vary with age and gender. Reduced alcohol consumption is the most beneficial for younger males and losing weight for older women (if relevant). By dragging the utility/preference bars to indicate the personal importance of each ‘attribute’ these rankings vary in real-time for the consumer.
Conclusion: Personalized evidence and individual patient preferences can be combined using MCDA to prioritize preventive health care options for consumers. This is being evaluated by randomized controlled trial.
Candidate for the Lee B. Lusted Student Prize Competition
