We understand that "where one lives makes a difference to health in addition to who you are", and that the effects of the place of residence or context unit (CU) on public health are due to the heterogeneity of the physical and social environment characteristics, in addition to the individual and aggregate attributes of the population nested in the CU. Those attributes, although intrinsically dependent on the individuals, are typically external to them and susceptible to intervention. Also, the UC's are understood as ecological units nested within successively larger communities. Depending on the study hypothesis, census-defined areas may be used as proxy for the CU. Alternatively, the CU may be defined by the individual's perception of his/her neighborhood. The renewed interest on the health effects of the CU are associated with new trends in public health, namely: new hierarchical levels of exposure beyond individual level characteristics, inequalities and social determinants of health, urbanization and the need to evaluate interventions not traditionally associated to public health. Our objective was, first, to review options while choosing the relevant CU and second, to review strategies to determine and quantify the characteristics of the CU using social systematic observation (SSO). The combination of census-defined data, information on the neighborhood defined by the local population, surveys and SSO still needs conceptual, methodological and analytical development. However, the distribution of the physical and social attributes of the CU will permit to incorporate other hierarchical level of complexity to better understand the incidence and prevalence of health related events in populations.
urban health; neighborhood; systematic social observation