Disease affects populations in two major and complementary ways: at the individual level, the infected may die or suffer other non-lethal but still undesirable consequences from the disease; at the population level, there may be devastation caused be it by diseases that decimate large numbers of individuals, or by creating endemic situations that are very costly in terms of treatment, prevention or isolation, for example. For many decades mathematical models have been proposed and analyzed to describe the evolution of epidemics through populations. Aside from the simplest ODE-based models that contain only epidemiologically relevant variables, several structural variables have been introduced through the years, such as size, age of disease, chronological age, gender, etc. Separately, in last few decades, several models have been proposed and analyzed for the immunological response at the individual level that is the main determinant of recovery, chronic disease, or death. Even though these types of models naturally beckon modelers to combine them, only very recently has this idea taken strength. We propose a general framework for immuno-epidemiological models that use variables of immunological nature (e.g. viral load and some measure of immune response such as T-cell density) as structure variables of epidemiological models. This results in coupled systems of ordinary and partial differential equations, possibly with nonlocal terms and/or boundary conditions. Under this general framework, we describe a simple such combination in an SIR case, and present a partial analysis of the resulting model, as well as generalizations.