|Home | About | Journals | Submit | Contact Us | Français|
This supplement to Public Health Reports (PHR) focuses on data systems and their use in addressing social determinants of health (SDH). This particular topic requires attention now given the evidence of increasing burden and worsening inequities in some health outcomes, in spite of decades of work to change individual behaviors, as well as the need to be efficient in our use of existing data. A holistic approach to disease prevention is urgently needed to reduce the inequities that have been perpetuated in our society for so long.
Despite concerted, targeted, and coordinated efforts to reduce inequities in health outcomes, gross inequities still exist,1–4 and some evidence indicates that the gap between the best health outcomes and the worst health outcomes is growing.1,3–5 Well-meaning efforts have substantially focused on individual-related behavior changes, with less focus on wider social and structural determinants of health, which can be defined as follows:6,7
Structural factors include those physical, social, cultural, organizational, community, economic, legal, or policy aspects of the environment that impede or facilitate efforts to avoid disease transmission. Social factors include the economic and social conditions that influence the health of people and communities as a whole, and include the conditions for early childhood development, education, employment, income and job security, food security, health services, and access to services, housing, social exclusion, and stigma.8
In addition to addressing individual factors, there is an urgent need to address social and structural factors and to better understand their relationship to each other as we develop effective programs and policies to reduce inequities.
A holistic approach to disease prevention involves not only addressing individual, social, structural, and environmental determinants, but also working with a wide array of sectors, such as health, education, justice, environment, and labor. Additionally, it means working with diverse kinds of data, including disease surveillance, legal, land use, marketing, workforce, education, and financial. Making the best use of a wide variety of data at the individual, neighborhood, community, and county levels, for example, can provide a more complete description of the underlying factors that may influence health outcomes than using disease surveillance data alone. As a matter of fact, using disease surveillance data alone, which often are limited to variables such as disease of interest, age, sex or gender, and race/ethnicity, can be stigmatizing and only tells part of the story. Public health professionals have an obligation to fairly and accurately describe disease occurrence in populations. As a result, we should be compelled to use data from available sources to provide a complete picture of the environment in which the disease occurs and any underlying factors contributing to its occurrence.
Addressing underlying factors of health has been advocated by many health practitioners for decades.1,9–12 The Institute of Medicine Committee on Public Health Strategies to Improve Health released a report in 2010 that recommended gathering, analyzing, and communicating health information that includes not only disease-outcome data, but also data on underlying factors contributing to poor health.13 In many cases, national disease surveillance systems do not include information on underlying determinants of disease, necessitating linking to existing sources of social, structural, legal, environmental, and financial data to provide a more comprehensive description of the affected population.14
This special issue of PHRaims to reflect on the types of data we routinely gather, analyze, report, and communicate, and it calls us to take a holistic approach to data use both in the sources (e.g., United Nations, Centers for Disease Control and Prevention [CDC], Census Bureau, Department of Transportation, and Department of Justice) and kinds (e.g., disease outcome, policy, financial, land use, service usage, achievement, and segregation) of data used in public health. It calls us to be good public health stewards by challenging us to move beyond our routine analyses based mostly on individual-level data and include data from other sectors and levels in the work we do. This supplement provides examples of innovative uses and analyses of data for local, state, and national governments and organizations to consider.
Promoting health equity through a holistic approach is a major strategic priority of CDC's National Center for HIV/AIDS, Viral Hepatitis, STD, and TB -Prevention (NCHHSTP).15 NCHHSTP's recent white paper entitled “Establishing a Holistic Framework to Reduce Inequities in HIV, Viral Hepatitis, STDs, and Tuberculosis in the United States” calls for a systematic approach to monitoring disease by simultaneously reporting on disease outcomes and underlying factors of poor health.16 NCHHSTP is also placing more emphasis on addressing structural determinants of health, including health policy, economic and social interventions, and cross-sectoral collaborations. The articles in this supplement clearly expand the knowledge base on social determinants and data use and are examples of the holistic approach to public health suggested in the CDC white paper.
Three articles suggest novel ways to use surveillance data. Beltran et al. inventoried SDH variables in U.S. national infectious disease surveillance systems in NCHHSTP.17 Very few SDH variables are collected in NCHHSTP surveillance systems. The authors recommend routine, simultaneous monitoring of a key set of SDH variables and disease outcomes to better describe the contribution of underlying factors and improve monitoring of health inequities. They also underscore the need to use comparable measures of inequity (both relative and absolute) when reporting. Song et al. used acquired immunodeficiency syndrome (AIDS) surveillance data linked with American Community Survey data from the Census Bureau to develop a statistical tool that brings together standard, accessible, and well-understood analytic approaches, and that uses publicly available SDH data.18 The authors demonstrate that area-based measures of socioeconomic variables should be considered by surveillance systems for routine monitoring of underlying determinants of health. Finally, Soto et al., through use of Connecticut invasive pneumococcal disease data linked to U.S. Census data, found that differences in neighborhood poverty levels describe disparities in rates of invasive pneumococcal disease as large as those described by race/ethnicity.19 These articles all underscore the importance of routine monitoring of area-based SDH simultaneously with demographic and disease-outcome data.
Comer et al. describe the utility of linking electronic health record systems and community information systems to address inequities in SDH in Indianapolis, Indiana.20 Marrying these two types of systems could be a model for other communities to consider as social, structural, and physical environment information can directly inform policy development. As part of their work, the authors identified the need for both a multidisciplinary approach and tools that can be used by multiple sectors.
One article addressed mental health issues and human immunodeficiency virus (HIV) prevalence. Walkup et al. used Medicaid claims data and looked at the prevalence of diagnosed HIV/AIDS among individuals with severe mental illness (SMI) at the community level through metropolitan statistical areas (MSAs).21 Medicaid data were linked with a range of social and contextual variables at the MSA level, including, economic measures, crime rates, and substance abuse treatment resources, to name a few. The authors suggest a need for a new generation of research on HIV among people with SMI, looking at geographic variation. The authors also clearly demonstrate the utility of Medicaid claims data in monitoring small area variations in physical health among people with SMI.
Housing and labor are important SDH, as they determine the neighborhoods we inhabit and reflect the quality of employment choices we have, which, in turn, affects income and quality of life. Mendez et al. used Home Mortgage Disclosure Act (HMDA) data to develop an index for mortgage discrimination at the community level.22 The authors linked redlining indices from HMDA data to data from a pregnancy cohort study and to U.S. Census data. They then spatially mapped levels of redlining for each census tract for the pregnancy cohort from Philadelphia County, Pennsylvania. This work provides an excellent example of linking public health data with housing data to assess discrimination.
Lòpez-De Fede and colleagues used sexually transmitted disease and HIV surveillance data linked to county-level unemployment data from the Bureau of Labor Statistics and county poverty data from the Census Bureau to assess relative socioeconomic disadvantage using three indicators—an unemployment index, a poverty index, and the Townsend index, which is a composite of unemployment, household tenure, household crowding, and vehicle access.23 The authors used ring maps to graphically display the coincidence of syphilis, gonorrhea, Chlamydia, and HIV infection across South Carolina counties. The incorporation of the geospatial component in their analysis adds a much-needed dimension to the routine use of disease surveillance data.
An excellent example of community-based participatory research aimed at promoting healthier workplaces is shared by Gaydos and colleagues.24 The San Francisco Department of Public Health collaborated with university and Chinatown community partners to develop and pilot, in 106 Chinatown restaurants, an observational checklist with the goal to assess preventable occupational injury hazards and compliance with employee notification regulations. The authors aimed to promote healthier workplaces by identifying ways to monitor compliance with local safety ordinances.
Gender equity is an often-overlooked determinant of health. In her commentary, Phillips discusses how to measure, and why one should measure, gender equity in public health research, surveillance, and programs.25 Phillips further encourages public health professionals to use gender and gender equity in research to enhance our understanding of how this social determinant affects health outcomes.
Policy data can be very useful to assess the impact of laws, regulations, and rules that affect SDH and health outcomes. Heymann and colleagues provide an example of how national public policy is often one of our best approaches to have the widest impact on SDH in the shortest amount of time.26 The authors looked at duration of paid maternal leave and neonatal, infant, and child mortality rates in 141 countries, controlling for a number of country characteristics. They found that an increase in 10 full-time-equivalent weeks of paid maternal leave was associated with a 10% lower neonatal and infant mortality rate and a 9% lower mortality rate for children younger than 5 years of age. While causation cannot be claimed, the authors leave us with ideas for advancing the use of policy data in public health analyses.
Another article in this supplement promotes the use of policy data—specifically, legal variables—in public health research. Burris provides a public health law researcher's perspective on how to address SDH and how to incorporate law and research on law into the effort.27 Burris distinguishes between how law operates in creating and maintaining pathological social conditions, and the role of law as a pathway along which social structures are transformed into levels and distributions of health. He also highlights the value of integrating law more frequently into public health research, even when law is not the topic of study. Finally, in very compelling fashion, he argues for putting legal variables and hypotheses on an equal footing with other social and attitudinal factors influencing health outcomes.
Recent advancements in knowledge related to “psychosocial pathways” that may help explain the socioeconomic gradient in health are highlighted by Braveman.28 In her article, she discusses stress, high external demands combined with low control at work, and one's perceived position in a social hierarchy as psychosocial factors that impact the socioeconomic gradient. She further explores these and other psychosocial factors while discussing their implications for public health research and policy, and notes that it is important to adequately consider social factors in surveillance and one-time studies, even when they are not the focus of the study. Finally, Braveman calls on health leaders to be vocal advocates for policies to reduce social disadvantage.
Allen et al. provide an essential state perspective on the NCHHSTP Strategic Plan to holistically address SDH.29 The authors challenge leaders at CDC to avoid the fate of other strategic plans by (1) defining health equity-related concepts in concrete terms, (2) moving beyond program silos to embrace a syndemic approach, (3) using SDH data to allocate resources, and (4) overcoming the public's general discomfort with the topic to effectively address racism.
Senior leaders from CDC's NCHHSTP gathered with Burris, Braveman, and Allen to explore additional options for addressing SDH at the National Center. Colbert and McDavid Harrison capture a summary of that discussion, which focused on identifying efficient ways to address SDH.30 Discussions included suggestions to (1) routinely and simultaneously monitor SDH and disease outcomes as a way to provide a more complete, nonstigmatizing description of affected populations; (2) engage the Procurement and Grants Office at CDC to become a full partner in addressing SDH through, for example, inclusion of language to address SDH in all funding documents as the National Center fully incorporates an SDH approach into all of its work; and (3) put how we communicate about SDH at the center of our efforts to raise awareness and gain buy-in across CDC.
Sadana and Harper challenge the public health community to work to describe compelling theories at a program level that can explain complex situations and take into account biological, social, and political processes.31 They further point out the need to identify institutional mechanisms, technical norms, and appropriate incentives to share data at all levels and across all sectors. Finally, they contend that while policy advances are certainly also important at this stage, what may be of greater importance is wider availability of analysis tools to measure and monitor health outcomes and training to avoid mistakes of inappropriately analyzing or interpreting data.
In his invited commentary, Dr. Koh reminds us that surveillance systems are at the root of our work in public health.32 If we do not have data on the underlying determinants of health in our surveillance systems, or have them easily accessible, then we likely will not routinely report on them. This incomplete, often simplistic, picture of affected populations inevitably leads to “deficient and misleading” messages, which then shape our programs and policies. He calls us to have variables on SDH available for analysis, and reminds us that to embrace a more holistic approach, current systems should incorporate SDH into routine reporting, either by collecting SDH-related variables or by linking to other databases that do.
Recognizing the need to incorporate a more holistic framework to eliminate health disparities, several wide-reaching public health efforts include incorporation of data on SDH in health-monitoring systems. In the U.S., Healthy People 2020 contains an overarching topical area on SDH,33 and, for the first time, includes indicators and objectives addressing SDH.34 The World Health Organization's Commission on Social Determinants of Health recommended that data on SDH be collected and analyzed in conjunction with health data.12 The goal of both efforts is to provide comprehensive data to develop holistic programs to reduce health inequities.
We hope the information presented in this supplement serves as a catalyst for change in two ways: (1) to encourage public health professionals to embrace the use of data on SDH in the collection, analysis, and presentation of health data; and (2) to demonstrate the added value of utilizing data systems to describe the interplay of SDH with other determinants of health when routinely analyzing existing data systems or developing new public health data systems. In summary, by utilizing data on SDH in conjunction with health data, the added value to public health will be a healthier populace.