Dissertation Title:  "Decision Making for Medical Innovations"

The objective of this dissertation is to evaluate decision making related to medical innovations. Paper 1 uses micro-simulation modeling techniques to evaluate the potential benefits of the forthcoming 9-valent Human Papillomavirus (HPV) vaccine, taking into account several key sources of uncertainty. The results suggest that the 9-valent vaccine yields greater reductions in cervical cancer, compared to currently available vaccines, and that the prevalence of co-infection with multiple HPV types and unidentifiable HPV types in cancer cases can influence estimates of vaccine effectiveness, but that the magnitude of effects may be moderated by vaccine cross-protection against HPV types not targeted by the vaccine. The results provide ranges of potential benefits and cost thresholds by which the 9-valent vaccine would be cost-effective compared to current vaccines in Kenya and Uganda.

Paper 2 uses surveys to investigate whether and how patients and physicians may differ in their perceptions of future ambiguous benefits of cord-blood stem cell banking. The results suggest that patients and physicians have different perceptions of the value of cord-blood stem cell banking, primarily due to their different perceptions of the probability of realizing its ambiguous future benefits. Physicians underestimate the amount of money that families, particularly those with low incomes, are willing to pay for the service. The results also show that physicians think it is significantly less important to discuss decisions with patients and families when only ambiguous benefits are possible. These differences suggest the value of improved communication between patients and physicians when the benefits of a medical service are ambiguous.

Paper 3 builds on Paper 2 and investigates the effects of a physician's conflicts of interest (COI) on the perceived value of a service that provides ambiguous benefits. The results show that physicians overestimate the effects of COI in low-income families when the service involves unambiguous benefits, but not when the service offers only ambiguous benefits. The physicians' overestimation of the COI effects could be a barrier to COI disclosure. The results also show that COI reduce patients' trust and willingness to share the decision with their physician, regardless of the degree of ambiguity about future benefits.