The aim of this research is to develop a barrier model (or models) for participation of women 60 years and older in prevention clinical trials studies for breast cancer and heart disease, two major causes of death in these women. The model will have two components: one reflecting participation decisions made by women; the second reflecting recommendation decisions by primary care physicians. The purpose of this model is to inform recruitment strategies and prevention clinical trials planning so as to ensure greater participation by this under- represented group of women in future breast cancer and heart disease prevention clinical trials. The investigation of a barrier model for two diseases permits an investigation of the generality of either model to the other, and thus to other diseases. The experimental design for women blocks on ethnicity (Latino, African- American, and white) and age (60-70 and 71-80. The experimental design for physicians blocks on gender, ethnicity and age (male and female; African-American, Latino and white: aged under 40, 40-55, over 55), 35 respondents within each stratum provide 210 and 630 physicians for conducting multi variate analyses. Controlled decision making experiments for barrier model tests will be administered within stratum, making it possible to systematically evaluate differences in participation and recommendation barrier models for heart disease and breast cancer among these groups. Although the rigor of controlled experimental trials has become the gold standard for assessing effects of medical interventions, this same scientific rigor has not been extended to judgement research in the health-care area. This research will apply advanced psychological measurement methods (advancements to traditional Conjoint and Functional Measurement) where factors hypothesized to affect decisions are manipulated in judgment experiments, making it possible to draw causal conclusions about how people value and tradeoff barrier information in judging their decisions. Respondents' barrier factor values and an understanding about how they are traded off in their participation decisions are obtained from the algebraic barrier model that passes its tests in accounting for respondents' data where alternative models fail. This falsifiability feature, new to health care research where """"""""measures"""""""" come from assumed models or operational definitions) provides an internal validity base for conclusions about values, value tradeoffs and models. The validated barrier model(s) will explain the conditions under which elderly women would likely participate and primary care physicians would likely recommend participation in prevention clinical trials. Model parameters will be included in multi variate analyses with data from background and opinion questionnaires (the Spanish forms developed from independent forward/back translations) for a better understanding of who values what with regard to breast cancer and coronary heart disease clinical trials participation. The barrier model(s) will be converted to a computer program for ready use by future planners of clinical trials to identify promising approaches for affecting positive participating decisions under a range of policy options.
