The research agenda for applying QALYs to studies of adolescents is challenging. There are several central issues. First, a central tenet of outcomes research is that it takes the respondent's perspective. Many studies of children and some studies of adolescents use proxy reports by parents. We know very little about the validity of these proxy reports. Further, little has been done to compare the results of analyses from the patient and the parent perspectives.
A second important research area concerns the way adolescents trade current versus future benefits concerning their health. Although some analyses discount future benefits, most make the assumption that people make rational trade-offs. The well-documented present orientation of adolescents may suggest that their health-related preferences are slanted toward maximization of present benefits without concern about future consequences on their health. However, at present there are very few preference studies involving children or adolescents (for exceptions see Feeny et al., this volume;Saigal et al., this volume).
Finally, many of the policy analyses make assumptions about future benefits of interventions on health. For example, prevention of cigarette smoking may result in the prevention of cancer and lung and heart diseases later in life. However, these effects are typically assumed rather than observed. Often the policy conclusions are heavily influenced by the values assumed for these analyses. Consequently, more research is necessary to build accurate models for policy analyses that are relevant to adolescent health.
Over the last 20 years there has been remarkable development of HRQOL measures. This chapter argues that measurement of health-related outcomes for children and adolescents requires the consideration of difficult conceptual issues, such as estimation of the area under the treatment curve, the value of generic measures, and the need to explicitly measure the quality in QOL. The most popular measures in use today typically fail to address these issues. In my view, decision-theory-based measures more adequately address these conceptual issues. However, adapting these measures for studies of children and adolescents raises some important and challenging methodological issues that should be addressed in future research.