Natural history of diseases: Statistical designs and issues.

Understanding the natural history of a disease is an important prerequisite for designing studies that assess the impact of interventions, both chemotherapeutic and environmental, on the initiation and expression of the condition. Identification of biomarkers that mark disease progression may provide important indicators for drug targets and surrogate outcomes for clinical trials. However, collecting and visualizing data on natural history is challenging, in part, because disease processes are complex and evolve in different chronological periods for different subjects. Various epidemiological designs are used to elucidate components of the natural history process. We briefly discuss statistical issues, limitations, and challenges associated with various epidemiological designs.