REGARDS and Disparity in Stroke Mortality: A Model Study

From the article from the UAB Center for Clinical and Translational Science:

Young investigators interested in emulating success would do well to take note of a recent REGARDS study, which found blacks are four times more likely to die of stroke at age 45 than their white counterparts because blacks have more strokes. The study, “Where to Focus Efforts to Reduce the Black-White Disparity in Stroke Mortality: Incidence versus Case Fatality,” appeared in the July 2016 issue of Stroke.

“We’ve known about the difference in black and white stroke mortality for over 50 years,” said lead PI Dr. George Howard, professor of Biostatistics, UAB School of Public Health. “But there have been virtually no studies asking why?” Instead, research has been largely focused on treating stroke patients and preventing recurrent strokes.

“The implications of our finding—that the driving force of racial disparities in stroke mortality is the higher incidence of strokes in blacks rather than case fatality—are profound, highlighting where we need to make changes that will effectively reduce stroke mortality in the black community,” Howard noted. High blood pressure, diabetes, and other traditional and nontraditional risk factors that lead to stroke are more common in the black community. However, just the higher prevalence of these risk factors explains less than 50% of the excess risk in blacks.

While more blacks die from stroke, it Is not clear whether this is because blacks are at higher risk of having a stroke, or blacks are more likely to die from the stroke once it occurs. “While it is important to ensure blacks and whites receive the same care once a stroke happens, any differences in care do not appear to be the reason more blacks die from stroke. Rather, we need to focus on prevention and better control of risk factors for stroke,” he concluded.

The study’s premise underscores the powerful insights that can result when one asks the right questions of the right big data set.

The REGARDS data are the result of a study that used a novel, non-clinic based methodology to recruit a cohort of more than 30,000. The Clinical and Operations Center (COC) method is based on a sampling frame using the same list as the national Behavioral Risk Factor Surveillance System (BRFSS) to identify a national sample of the general population aged 45+ with overrepresentation of African Americans and residents of the “Stroke Belt” (NC, SC, GA, TN, AL, MS, AR, LA) and “Stroke Buckle” regions (coastal plain region of NC, SC, and GA). Potential participants were randomly selected from this list and contacted first by letter and brochure, followed by telephone interview of those who agreed to participate. The novel aspect of the REGARDS design is that the telephone assessment was complemented by an in-person physical exam conducted by health professionals (often nurses) who frequently perform assessments for life insurance companies. Specimens collected in this “in-home” visit were analyzed by a central lab at the University of Vermont, while ECGs were evaluated at a central reading center at Wake Forest University. Self-administered questionnaires were also used to gather additional information. Participants have been contacted at 6-month intervals for more than 8 years (the effort is ongoing). The annual retention rate of the cohort has been 97.4%. During this time, investigators identified more than 4,000 potential strokes. Medical records for 92% of these were received and adjudicated by a physician panel, documenting over 1,400 strokes.

Advantages of the COC approach are (1) reduced costs associated with a clinical setting and the substantial study-wide staff require and extrapolated across the sites involved and (2) providing a geographically representative sample, since traditional clinical study recruitment is often tied to regions around tertiary medical facilities.

Read the full article.