(Applicant?s abstract): Escalating health care costs for treatments and medical technologies has led decision and policy makers to focus on preventive interventions that are both efficacious and cost-effective. Longitudinal prevention programs in mental health should include a cost-effectiveness or cost-benefit component that use the best methodology available to assess the long-term benefits of the intervention. The economic consequences of mental illness prevention programs often span multiple years into another stage of life beyond the period of intervention. This is particularly true of prevention programs targeting children. The interventions are aimed at proximal targets (mediators) thereby altering the developmental trajectory of mental health outcomes. This statistical design issue alone makes economic analysis for prevention programs more complex. Prevention scientists need alternate methods to permit both efficacy and economic evaluations in the design and analyses of studies of preventive interventions. The methods for long-term benefits need to include developmental trajectories. Developing such methods requires extensive knowledge of statistics, prevention science, and economics. The K25 mechanism will help integrate formal basic training in economic analysis with recent training in prevention science by building on a solid foundation in statistical methodology. The specific goals that will be achieved through training and mentoring are: (a) a basic theoretical understanding of economic analysis and of the use of cost and benefit measures in prevention research, (b) ability to effectively collaborate with mental health economists. The research plan integrates training and mentoring by addressing the question: What are the short and long-term benefits of mental health interventions and how do they vary across clusters of individuals with distinct developmental trajectories of long-term outcomes? Our approach is to extend a new statistical method called general growth mixture models (GGMM) to identify distinct clusters of individuals using time invariant and time varying covariates using mixture models and developmental trajectories of distal outcomes. A cost-benefit analysis will be conducted for the clusters of individuals using data from an ongoing federally funded longitudinal childhood mental health intervention study at the Johns Hopkins Prevention Center. Since missing data (planned or due to attrition) occur naturally in longitudinal studies we will also investigate the effect of missing data patterns on cluster membership of individuals using GGMM.
| Indurkhya, Alka; Mitra, Nandita; Schrag, Deborah (2006) Using propensity scores to estimate the cost-effectiveness of medical therapies. Stat Med 25:1561-76 |
| Mitra, Nandita; Indurkhya, Alka (2005) A propensity score approach to estimating the cost-effectiveness of medical therapies from observational data. Health Econ 14:805-15 |
| Gray, Ronald F; Indurkhya, Alka; McCormick, Marie C (2004) Prevalence, stability, and predictors of clinically significant behavior problems in low birth weight children at 3, 5, and 8 years of age. Pediatrics 114:736-43 |
