CENTERS FOR DISEASE CONTROL AND PREVENTION
Volume 15 • Number 3 • Fall 2002

Return to index of articles

We Must Identify the Gaps Before We Can Close Them

Health departments must have timely, reliable data if they are to get a clear picture of a community’s health disparities and design programs that work. "If you don’t have data, you’re just assuming there’s a gap," advised CDC Health Educator Alexandria L. Stewart, MS. "Programs must be driven by scientific data and not assumptions." 

But gathering the necessary information and translating it into interventions that target health disparities can be complicated and expensive. "Health departments are betwixt and between because they want to do something about health disparities, but they don’t know what to do," noted Robert G. Robinson, DrPH, Associate Director for Program Development for CDC’s Office on Smoking and Health (OSH). 

Dr. Robinson, Ms. Stewart, and others at CDC are developing guidelines on how to collect, analyze, and use data to eliminate health disparities and how to strengthen these activities by involving communities. Moreover, communities participating in CDC’s Racial and Ethnic Approaches to Community Health (REACH 2010) project are collecting data and developing new community-driven strategies to eliminate health disparities. This 5-year demonstration project is one of the first programs to target health disparities associated with race and ethnicity. 

Gathering Baseline Data 
At 21 of the REACH 2010 sites, surveys are under way to help each community learn more about its risk factors for certain chronic diseases. The National Opinion Research Center (NORC) at the University of Chicago is collecting data on behalf of CDC. 

"We have taken about 60 questions from CDC’s Behavioral Risk Factor Survey to create the REACH 2010 Risk Factor Survey," explained CDC Evaluation Coordinator Pattie J. Tucker, DrPH, RN, who is responsible for the project’s data collection and evaluation activities. "These questions address the health behaviors and practices that are specific to three of our priority health areas — breast and cervical cancer, diabetes, and cardiovascular disease." The surveys ask adults about 

• Their general physical and mental health. 
• Recent visits to health care providers. 
• Exercise or physical activities. 
• Smoking history. 
• Eating habits.
• Weight and height. 
• Income. 
• Screening for cardiovascular disease. 
• Screening for breast and cervical cancer. 

By using the same survey at all sites, REACH 2010 communities can serve as comparison communities for each other. For example, the communities striving to eliminate racial and ethnic disparities in breast and cervical cancer will serve as comparison communities for those communities targeting diabetes and cardiovascular disease. Thus, the communities can see how they compare with other communities that have not implemented the specific strategies, track local changes in risk factors over time, and plan more effective programs. 

Before the surveys are conducted, NORC staff work closely with local REACH 2010 coalitions to identify community boundaries and understand the needs of each community. For example, in communities where many homes have no phones, interviews are conducted in person. Local residents are recruited and trained to conduct the door-to-door interviews. 

Community members will be involved every step of the way, noted Dr. Tucker. "These data will be shared with the coalitions. We must be respectful of community members, and we work hard to gain their trust. This is the first time many of our REACH 2010 grantees have had community-level data collected." 

Getting to the Root of the Problem 
Although surveys can provide a snapshot of what is happening in a community from year to year, they cannot capture the detailed information that face-to-face discussions are so effective in gathering. Therefore, REACH 2010 coalitions are using focus groups and interviews to collect qualitative data.

"Focus groups and in-depth interviews really help you get to the root of the problem," said CDC REACH 2010 Director Imani Ma’at, EdD, EdM, MCP. "For example, in Nashville, the REACH 2010 grantee conducted numerous focus groups to see what is contributing to the high rates of death from cardiovascular disease and diabetes among African American women. Many of the women said that they did not have time to exercise or that they were under a lot of stress because of work and family responsibilities," she noted. The grantee also discovered that "plump is considered good" by many of the African American women in this community. 

Such detailed comments from women in the focus groups allowed the Nashville REACH 2010 Coalition to develop culturally appropriate messages that will reach and motivate women in the target community to take charge of their health. "You’ll only get that level of detail in a focus group," noted Dr. Ma’at. 

Mapping Out Goals 
The REACH 2010 communities are excited about being able to see positive trends at the local level, but they know that 5 years is not long enough to witness any declines in illness and death. "They will be able to see incremental changes in healthy behaviors," said Dr. Tucker. "For example, in Year 4 of the program, people might be more physically active than in Year 1. Or in Year 4, people might be eating more fruits and vegetables than in Year 1." Collecting data to track incremental changes will be essential when the REACH 2010 grantees evaluate their efforts to eliminate health disparities. 

To help communities set intermediate and long-term goals and monitor their progress in meeting those goals, REACH 2010 staff have developed a logic model. "It’s what we think is a logical approach to eliminating disparities," Dr. Tucker noted. "Our logic model asks communities to consider what other partners and service providers are doing as well. So it’s not just about the REACH 2010- funded communities but about all partners. Arrows on the logic model point back and forth to show how these interactions happen continuously in a community." 

For example, a REACH 2010 coalition that is targeting high rates of diabetes in a community might be working with a local clinic to educate health care providers about patients’ need for regular eye and foot exams. "At the same time, there might be a medical association at the state level that wants to educate providers about exams that are important for patients with diabetes," explained Dr. Tucker. The logic model helps communities look beyond their efforts and see how these dual education programs could work together to end disparities. 

CDC is developing a Web-based information system that each coalition can access to enter data about their capacity-building activities, interventions, and changes that have occurred because of their efforts. From these data, communities can create local reports to justify the need for programs targeting health disparities and to promote passage of health laws and policies. 

The REACH 2010 project will conclude in 2004, after rigorous evaluations to identify which approaches are most effective. Details about successful strategies will be disseminated widely. "The goal is to make interventions more community-specific," Dr. Tucker said. "Community X might say, ‘We will use some of the lessons learned from Community Y, but we’ll do it differently.’ So they might develop a hybrid program or just change the application. For example, they might take an intervention originally based in clinics and instead deliver it in the community barbershop because that’s where people in Community X get their health information." 

Helping Health Departments Collect Good Data 
Lessons learned from the REACH 2010 communities will be shared when evaluations are completed. In the meantime, Dr. Robinson and colleagues are publishing a resource guide on eliminating health disparities ("Overcoming the Challenges of Eliminating Disparities in Tobacco Use"). The guide describes a number of strategies that health departments can use to collect good data on populations hard hit by chronic disease — for example, aggregation of data, oversampling, biannual sampling, and community assessments. 

"Oversampling can be used to ensure sufficient numbers," explained Dr. Robinson. "Population groups with substantially low representation in the state or territory can be targeted at sample levels appropriate for statistical analysis. This may be an especially good strategy if the health department wants to accumulate data on ethnic communities, groups, and population strata with low numbers." 

Reaching Diverse Populations 
When planning data collection, health departments need to think about the various groups of people potentially affected by health disparities. Here are some of the groups to consider: 

• African Americans, Asian Americans and Pacific Islanders, Hispanics and Latinos, and Native Americans.
• Women.
• Young people.
• Gays, lesbians, bisexuals, and transgender people.
• People with low incomes and little education.
• Rural residents.
• Blue-collar workers.
• Migrant workers.
• People with mental illness. 
• People with physical disabilities. 
• People in prison.
• Homeless people.

"Moreover," said Dr. Robinson, "attention should be given to groups not identified but nevertheless present who may possess distinct community characteristics and experience disparities, such as Cajuns in Louisiana, cowboys or cowgirls in the West, or African American tobacco farmers." Collecting detailed data from these populations is expensive and complicated, but the data are essential if a health department is to understand the disparities it must target and develop the needed interventions. 

Using the Right Tools 
The tools used to collect data on health disparities must be community competent, Dr. Robinson emphasized ("Community Competence"). Community competence involves thinking about the experiences and circumstances that have shaped the community and then working with community members to incorporate these concepts into the surveys, focus groups, and interviews, he said. 

"Related to community competence is the matter of assessing behavior unique to specific population groups," he stated. "For example, surveys will need to distinguish between the traditional and nontraditional use of tobacco by Native Americans to ensure that interventions reflect their different cultural patterns." 

People in the community must be involved when surveillance strategies are planned, when data are collected, and when evaluations are designed and carried out, Dr. Robinson emphasized. Community members have their own ideas about what health problems are most pressing and how those problems should be addressed. By encouraging community involvement, the health department will have stronger data as well as the community’s trust and support. 

Knowing What to Measure 
Before collecting data on disparities, health departments should consider what they will measure to determine if their efforts are indeed closing the gaps.

For instance, if the disparity is a higher rate of tobacco-related disease and death for African Americans than for whites, here are some of the indicators a community could measure, Dr. Robinson suggested: 

• Tobacco use rates, quit rates, and relapse rates in the community. 
• Rates of various types of tobacco-related disease (not just cancer), 5-year survival rates, death rates, and years of potential life lost. 
• People’s knowledge, attitudes, and practices regarding tobacco use. 
• Prevention services, treatment, and quality of care provided as well as the tobacco industry’s marketing strategies in the community. 
• Capacity and infrastructure data— for instance, research projects and tobacco control programs in the community. 

By assessing a community’s ability to address the problem, one might discover that "the community has not been engaged in tobacco control services because they lack resources to address the problem," said Dr. Robinson. "Baseline assessments will identify community levels of capacity and infrastructure and facilitate strategies to increase these levels. And that will allow the community to participate in the critical decisions regarding setting priorities and allocating resources." 

Looking at Outcomes and Process 
Health departments that focus their surveillance and evaluations on outcomes and ignore the process often end up with too little information to determine why a program worked, what components of a program should be replicated, or what problems associated with implementation ought to be solved, Dr. Robinson warned. 

"For example, merchant education programs in San Diego were effective in lowering sales of tobacco products to youth in Latino and Asian communities but not in African American communities," he recalled. Process evaluation would have enabled program evaluators to describe differences in how the program was implemented in the Latino, Asian, and African American communities, barriers to implementation, and solutions. "Those answers would help us determine how to provide more effective programs targeting African Americans," he said. 

Thus, Dr. Robinson recommended that health departments collect a combination of process and outcome data that 

• Define the population. For example, how prevalent is a risk behavior in the target population? 
• Measure how well the intervention has been developed and conducted. Was the intervention successfully planned and developed? Have good programs been launched in the target community? 
• Measure successful achievement of the goal. Have risk behaviors declined in the target population since the intervention was launched? What about rates of disease and death? 

Communities and states are encouraged and excited when they see declines in risk behaviors just a few years after launching a prevention program. "It gives people hope that their efforts are making a difference," said Ms. Stewart. "For example, data from the 2001 Youth Risk Behavior Survey show that smoking among high school students declined from nearly 35% in 1999 to below 29% in 2001 — that’s a 16% decline. This is great news, and we believe it’s largely the result of price increases from manufacturers of tobacco products, increases in excise taxes, and the fact that more states are conducting tobacco use prevention programs. These data tell us that we need to continue doing what we know works — comprehensive programs, excise tax hikes, and smoke-free environments."

Such efforts will eventually lead to declines in disease and death. "This will require sustained effort over time and a planned commitment," noted Dr. Robinson. "There’s nothing simple about this problem. But eliminating health disparities is the only way we’re going to achieve the America we all dream about."

Return to index of articles

Privacy Policy | Accessibility

Home | Contact Us

CDC Home | Search | Health Topics A-Z

This page last reviewed August 10, 2004

United States Department of Health and Human Services
Centers for Disease Control and Prevention
National Center for Chronic Disease Prevention and Health Promotion