Personalized medicine is still in its infancy, with costly genetic tests providing little actionable information in terms of efficient prevention decisions. As a consequence, few people undertake these tests currently, and health insurance contracts pool all agents irrespective of their genetic background. Cheaper and especially more informative tests will induce more people to undertake these tests, potentially impacting not only the pricing but also the type of health insurance contracts. We develop a setting with endogenous observable prevention and adverse selection and we study which contract type (pooling or separating) emerges at equilibrium as a function of the proportion of agents undertaking the genetic test as well as of the informativeness of this test. Starting from the current low take-up rate generating at equilibrium a pooling contract with no prevention effort, we show that an increase in the take-up rate may decrease welfare as long as the equilibrium remains pooling and is especially detrimental when the equilibrium becomes separating. Similarly, decreasing the prevention effort cost (a proxy for more informative tests) is detrimental to welfare when it changes the type of equilibrium from pooling to separating. These results imply that the desirability of public policies encouraging genetic test taking or decreasing the cost of prevention effort varies according to the type of contracts observed in health insurance markets. Especially, such policies may not be advisable in the short run, as long as the equilibrium is pooling.