DOSE-RESPONSE MODELING

In general terms dose-response models describe how the probability or frequency of a specified response changes with the dose level.  For example, if the specified response is the occurrence of a lung tumor by 70 years of age and the dose level is the average amount of a chemical taken into the body per day, then a dose-response model would indicate how the probability of a lung tumor occurring by 70 years of age changes as the dose level changes.  Thus, in its simplest form, the dose-response model is a curve which indicates the probability of a specified response for each possible value of the dose

Dose-response models are inferred from either animal or human data on the observed frequency of a specific response at different dose levels.  For example, a set of laboratory animal data might consist of three or four dose levels (including controls) and for each dose level the total number of animals at risk of developing a specific response as well as the number of animals who responded (i.e., the number of animals that actually developed the specific response).  Thus, at each dose level in the study there are individuals who responded (i.e., developed the specified response) and individuals who were non-responders (i.e., did not develop the specified response).  The dose-response curve is inferred from the impact of the dose level on the proportions of responders and non-responders.