proaches into the delivery of health care and social services (Anderson et al., 2003a; Andrulis, 2005).
Address multiple health issues pertinent to a range of stakeholders in order to build momentum and community support for interventions (e.g., the Harlem Fitness Zone and the Cherokee Choices projects) and the Active Living by Design interventions (Sallis et al., 2006).
Recognize that the use of nonintervention control groups may not be acceptable to ethnic minority communities; rather, delayed intervention designs may be more tenable (Boon and Clydesdale, 2005).
Involve researchers who are knowledgeable about the racial/ethnic groups or cultures being studied or evaluated (American College of Epidemiology, 1995; Anderson et al., 2003a; Gil and Bob, 1999; Hopson, 2003; IOM, 1994; Kumanyika et al., 2005).
Focus attention on the process of intervening in addition to achieving outcomes. This may include particular attention to delivery channels, messengers, materials and messages, and other cultural adaptations or targeting. Subgroup analyses are critical, and, when differences are found, further examination is needed to explore why the interventions are not effective for certain subgroups (e.g., characteristics of the intervention itself or how it was implemented) (Kreuter and McClure, 2004; Yancey et al., 2006b).
Data collection and analyses for surveillance and monitoring are core functions of governmental public health practices. However, methodological limitations to the identification and documentation of health disparities must be addressed. The public health infrastructure has the capacity to monitor aggregate racial/ethnic groups (e.g., categories defined by the U.S. Bureau of the Census). However, some individuals in racially/ethnically diverse communities may not participate in formal national data-gathering efforts because of logistical issues and concerns about how the data might be used. These challenges may result in the underrepresentation of racial/ ethnic minority populations in national surveillance systems. Representative surveillance and monitoring systems must be established to allow the monitoring of minority populations at potential health risk. The REACH 2010 Risk Factor Survey, for example, is conducted annually in African-American, Hispanic/Latino, Asian/Pacific Islander, and American Indian communities throughout the United States (CDC, 2004). Data from this survey demonstrate that residents in racial/ethnic minority communities experience greater disease risk and burden than individuals in the general population living in similar areas or states.