Examining Methods Used to Evaluate the Cost-Effectiveness of Childhood Obesity Interventions

Abstract

This dissertation examines methods used to evaluate the cost-effectiveness of childhood obesity interventions in order to help decision-makers prioritize among competing health programs using standardized outcomes. Chapter 1 generates inputs for use in cost-effectiveness analyses (CEAs) of childhood obesity interventions. In Chapter 1.1, I use data from the Medical Expenditure Panel Survey to predict expenditures associated with obesity in childhood and adolescence. I found that obese children and adolescents have significantly different expenditures than their normal weight counterparts. I conclude that exclusion of obesity-related medical expenditures can potentially undervalue the cost-effectiveness of interventions. In Chapter 1.2, I use data from the Study of Early Child Care and Youth Development to examine the longitudinal trajectory of child weight. I derived probabilities of transitioning between weight classes that can be used in a decision-analytic model to extrapolate the effectiveness of childhood obesity interventions beyond childhood. I found that deviating from CDC BMI reference categories can more accurately capture the risk of future obesity. In Chapter 2, I evaluate the cost-effectiveness of a primary care-based obesity prevention program, High Five for Kids. Over two years, High Five for Kids was low-cost, but only marginally effective in reducing BMI. I used a decision analytic simulation model to extrapolate trial outcomes over a 10-year horizon, and found that in the long-term, primary care based obesity prevention was likely to be cost-effective relative to usual care. I also found that key methodological considerations can meaningfully influence the cost-effectiveness of childhood obesity interventions. In Chapter 3, I develop an agent-based model to explore the dynamics of the potential spread of obesity within families. I found that the “contagion” of obesity could result in significant collateral weight loss in family members not targeted in an intervention. As a result, CEAs may underestimate the benefits of obesity interventions. Moreover, I found that unless interventions are targeted toward all obese children in a family, the contagion of obesity can hinder weight loss in intervention targets. This model can be leveraged as a tool to optimize family-based obesity intervention strategies and inform randomized controlled obesity prevention trials.