II. METHODS
The following section details how the data for the community health assessment was compiled and analyzed, as well as the broader lens used to guide this process. Specifically, the community health assessment defines health in the broadest sense and recognizes numerous factors at multiple levels— from lifestyle behaviors (e.g., diet and exercise) to clinical care (e.g., access to medical services) to social and economic factors (e.g., employment opportunities) to the physical environment (e.g., air quality)—all have an impact on the community’s health. The beginning discussion of this section describes the larger social determinants of health framework which helped guide this overarching process.
Social Determinants of Health Framework
It is important to recognize that a multiple of factors have an impact on health, and there is a dynamic relationship between real people and their lived environments. Where we are born, grow, live, work, and age—from the environment in the womb to our community environment later in life—and the interconnections among these factors are critical to consider. That is, not only do people’s genes and lifestyle behaviors affect their health, but health is also influenced by more upstream factors such as employment status and quality of housing stock. The social determinants of health framework addresses the distribution of wellness and illness among a population—its contours, its origins, and its implications. While the data to which we have access is often a snapshot of a population in time, the people represented by that data have lived their lives in ways that are constrained and enabled by economic circumstances, social context, and government policies. Building on this framework, this assessment approaches data in a manner designed to discuss who is healthiest and least healthy in the community as well as examines the larger social and economic factors associated with good and ill health.
The following diagram provides a visual representation of this relationship, demonstrating how individual lifestyle factors, which are closest to health outcomes, are influenced by more upstream factors such as employment status and educational opportunities. This report provides information on many of these factors, as well as reviews key health outcomes among the people of Mercer County.
Quantitative Data: Reviewing Existing Secondary Data
To develop a social, economic, and health portrait of Mercer County, through a social determinants of health framework, existing data were drawn from state, county, and local sources. Sources of data included, but were not limited to, the U.S. Census, U.S. Bureau of Labor Statistics, Federal Bureau of Investigation Uniform Crime Reports, State of New Jersey Department of Health and Senior Services and New Jersey Council on Teaching Hospitals. Types of data included self-report of health behaviors from large, population-based surveys such as the Behavioral Risk Factor Surveillance System (BRFSS) and the New Jersey High School Survey County Rankings, as well as vital statistics based on birth and death records. It should be noted that other than population counts and racial/ethnic distribution, other data from the U.S. Census derive from the American Community Survey which includes data from a sample of a geographic area. Per Census recommendations, aggregated data from the past five years was used for these indicators to yield a large enough sample size to look at results by municipality.
Raw hospitalization discharge data for 2010 (most current year available) for the primary hospitals that are part of the GMPHP were obtained from the New Jersey Department of Health and Senior Services. Data were analyzed for primary diagnosis for inpatient and emergency room admissions and adjusted for age and population size per the 2010 U.S. Census. As categorized on the datasets provided, hospitalization data were re-coded using pre-determined categories from the ICD-9 codes (International Statistical Classification of Diseases and Related Health Problems).
Qualitative Data: Focus Groups and Interviews
From February – May 2012, focus groups and interviews were conducted with leaders from wide range of organizations in different sectors, community stakeholders, and residents to gauge their perceptions of the community, their health concerns, and what programming, services, or initiatives are most needed to address these concerns. To this end, a total of 29 focus groups, 17 interviews with community stakeholders, and 1 Forces of Change session consisting of 6 core discussion groups were conducted. Ultimately, the qualitative research amounted to participation of over 400 individuals.
Focus Groups and Interviews
In total, 29 focus groups and 17 interviews were conducted with individuals from across the thirteen municipalities that comprise Mercer County. Focus groups were with the general public, leaders and providers in specific communities, and special interest or vulnerable populations. For example, four groups were conducted with youth, one group with people living with a disability and their families, two groups with senior citizens, and one group with participants in a drug addiction recovery program. A total of 343 individuals participated in the focus groups. Interviews were conducted with 17 individuals representing a range of sectors. These included government officials, educational leaders, social service providers, and health care providers. A full list of the different sectors engaged during the focus group and interview process can be found in Appendix A.
Focus group and interview discussions explored participants’ perceptions of their communities, priority health concerns, perceptions of public health, prevention, and health care services, and suggestions for future programming and services to address these issues. A semi-structured moderator’s guide was used across all discussions to ensure consistency in the topics covered. Each focus group and interview was facilitated by a trained moderator, and detailed notes were taken during conversations. On average, focus groups lasted 90 minutes and included 6-12 participants, while interviews lasted approximately 30-60 minutes. Participants for the focus groups were recruited by community and social service organizations located throughout Mercer County. The goal was to talk to a cross-section of residents and leaders.
Forces of Change Assessment: Mixed Group of Community Leadership and Residents
To understand the larger context in which health occurs, a forces of change session was conducted with key stakeholders and community leaders specifically to explore the larger external factors in Mercer County. Approximately 60 members from the Community Advisory Board and other community residents joined together for an event in late March 2012 to discuss these issues. Breaking into six smaller discussion groups, conversations focused on generating a list of external factors (e.g., emerging legislation, the political context, environmental issues, infrastructure, physical geography) that are most critical to the region and identified opportunities and threats for each force. The focus groups for this component served as a brainstorming session for leaders from community-based organizations, health care institutions and hospitals, and health and social service agencies to identify these external factors, how they might impact—for better or worse—the population’s health, and ways to capitalize on opportunities they provide for future initiative planning.
Analyses
The collected qualitative information was manually coded and then analyzed thematically for main categories and sub-themes. Data analysts identified key themes that emerged across all groups and interviews as well as the unique issues that were noted for specific populations. Frequency and intensity of discussions on a specific topic were key indicators used for extracting main themes. While municipality differences are noted where appropriate, analyses emphasized findings common across Mercer County. Selected paraphrased quotes – without personal identifying information – are presented in the narrative of this report to further illustrate points within topic areas.
Limitations
As with all research efforts, there are several limitations related to the assessment’s research methods that should be acknowledged. It should be noted that for the secondary data analyses, in several instances, county-level data could not be disaggregated into municipalities. While the intent of the assessment was to define municipalities by zip code, this could not necessarily be carried out since various data sources used different delineations for community boundaries. Therefore, due to the challenges of working with secondary datasets, zip codes were not generally used as the delineation for geographic boundaries. Additionally, several sources did not provide current data stratified by race/ethnicity, gender, or age –thus these data could only be analyzed by total population. Finally, youth-specific data were largely not available, and in cases where such data were available, sample sizes were often small and must be interpreted with caution.
Likewise, data based on self-reports should be interpreted with particular caution. In some instances, respondents may over- or underreport behaviors and illnesses based on fear of social stigma or misunderstanding the question being asked. In addition, respondents may be prone to recall bias—that is, they may attempt to answer accurately but remember incorrectly. In some surveys, reporting and recall bias may differ according to a risk factor or health outcome of interest. Despite these limitations, most of the self-report surveys here benefit from large sample sizes and repeated administrations, enabling comparison over time.
While the focus groups and interviews conducted for this study provide valuable insights, results are not statistically representative of a larger population due to non-random recruiting techniques and a small sample size. Recruitment for focus groups was conducted by community organizations, and participants were those individuals already involved in community programming. Because of this, it is possible that the responses received only provide one perspective of the issues discussed. While efforts were made to talk to a diverse cross-section of individuals, demographic characteristics were not collected of the focus group and interview participants from the assessment, so it is not possible to confirm whether they reflect the composition of the region. In addition, organizations did not exclude participants if they did not live in one of Mercer County’s municipalities, so participants in a specific community’s focus group might not necessarily live in that area, although they did spend time there through the organization. Lastly, it is important to note that data were collected at one point in time, so findings, while directional and descriptive, should not be interpreted as definitive.
