Clinical & Policy

The AHRI with its focus on health disparities narrowing and elimination through spatial and racial/race variability in risk factors and health outcomes recognizes the fundamental roles of clinic-based research and the impact of policy in advancing health agenda in this nation. With these notions, the Associates at AHRI under the leadership of Dr. Franklin Opara have dedicated effort in examining policies that are based on epidemiologic and clinical findings in shaping the direction of public health in our nation. In addition, this division of the AHRI is charged with correlating the findings from the large datasets (primary data sources for AHRI research) with the clinical research findings in several settings (chronic diseases, cancer, diabetes mellitus, hypertension, etc).

This division overlaps with the major arm of AHRI in these aspects of disease conceptualization at the population level:


The association between geography and disease is as old as the history of modern epidemiology itself. John Snow (1855) mapped the cholera epidemic around the Broad Street water supply source in London in order to determine the probable cause or causation. The health effect of environment or region can be quantified through disease mapping. Such mappings are characterized by points or regions, and the scale of measurement of the mapping variable. Mapping whether it involves continuous or discrete data can illustrate disease or events rates, and mean in a specified region. Whatever scale of measurement used in constructing the map of disease incidence or mortality, analysis or assessment involves the division of the geographic region of concerns into discrete areas, obtaining standardized rates or means in each areas (zip codes, census tracts), categorizing the rates or means, and associating a color, image or shading scheme to represent the rates, cumulative incidence, proportion or mean in a chloropleth map.

The mapping of disease requires data on both the events and the population at risk (size, sex, age, race/ethnicity), and in this sense differs from time-trend or temporal analysis. In conducting a spatial or time-spatial analysis study, conceptualization involves: (a) choice of geographic area and the isolation of such regions into areas such as zip codes, (b) selection of standardization method, (c) Decision on number of categories of rates, and (d) Determining the cutoff point or reference.

While mapping primarily generates crude rates of disease, adjusted rates can be generated with standardization such as when the race/ethnicity, sex and age are controlled for before the rates computation upon which the mapping is based. This produces an unconfounded disease rate, a more desirable measure of disease occurrence and effect. Disease mapping can serve as the starting point for etiologic research or causation. However in rare disease clustering, interpretation of rates requires the basic understanding of the influence of population size on the magnitude of risk.


Racial/ethnic disparities or variability in health and health outcomes have been reported in specific populations namely US, Canada and UK. Multiple studies in this perspective continue to demonstrate elevated incidence and mortality as well as survival disadvantage of African Americans relative to Caucasians in cancer, cardiovascular and metabolic diseases regardless of geographic area in the United States. With this volume of data, and the need to narrow these disparities, there is an urgent need to build on what we know as race surrogates and develop etio-pathogenetic model of health disparities. Whereas several studies continue to observe these disparities, interventions to address the risk and predisposing factors remain to be fully implemented. But what is most important is the understanding of health disparities etio-pathogenesis.

Racial/ethnic variability in hypertension incidence and prevalence in the US has not been fully explained. The notion of race/ethnicity and its association with disease and health outcomes is not very well understood despite litanies of studies perpetually demonstrating this association. The questions remain: Does race cause diseases? Is race/ethnicity associated with disease? What are the surrogates for race if race is not a direct predisposing factor in disease occurrence? Does race imply biology, gene or gene-environment interaction? A critical attempt to address these questions is a viable alternative to mere association observed in epidemiologic studies. Race/ethnicity may facilitate disease progression or its onset in many ways, namely in the socioeconomic, environmental and demographic distributions of these variables among these racial/ethnic groups. Spatial correlation with race is essential to explanation of the role of race in diseases. For example, the association between race/ethnicity may be removed by controlling for geographic area such as zip codes.


At the AHRI, Clinical and Policy Research Division (CPRD), we provide the researchers with the tools needed to understand the conduct and interpretation of disparities in clinical research, and how to utilize such data in the formulation and implementation of policies to improve clinical guidelines in the understanding of disease at the population level, based on individual patients studied at the clinical settings. An important aspect of CPRD is to emphasize the effects of treatments or risk/predisposition/exposure, rather than focus on statistical stability or random variability in the application of research findings in improving individual and population health. Therefore by de-emphasizing the p value and focusing on point estimate or effect of treatment or exposure, one is able to appraise treatments and exposure while at the same time considering the possibilities of such results being influenced by random variability or error in sampling.