Understanding Let’s Get Healthy California Data
Let’s Get Healthy California aims to make California the healthiest state in the nation by advancing the Triple Aim of better health, better care, and lower cost, and to promote health equity by reducing health disparities. We seek to use data, innovation, and collaboration to impact the overall health of our state.
Some aspects of achieving better health for individuals and populations, improving health care systems, and reducing costs are more complex than others. Because of this, we provide explanations of some of the more complex aspects of these data and indicators. We want to make these concepts more understandable within the context of Let’s Get Healthy California.
Race / Ethnicity Data:
Many Let’s Get Healthy California indicators display data by race and ethnicity. Health disparities between racial and ethnic groups are due to a combination of factors including, social and economic factors, health care and access to care, and health behaviors such as smoking or nutrition. For Let’s Get Healthy California Initiative, we measure differences based on health goals, compare each measure to a target, and aspire to reach the best possible outcome attainable for all groups. “Race” and “ethnicity” are defined by society rather than science. What we see as disparities are often more closely related to social and economic factors than they are to differences in appearance.
For definitions and more information, see below: “Understanding Race/Ethnicity Data in the Context of Health Equity”
While there are many indicators of health, income and wealth play especially important roles in determining health outcomes. Income and wealth are discussed in depth in this section because of their tremendous impact on health, and the inequities in how they are distributed among California’s population.
While America’s constitutional principles emphasize the importance of justice and equity, its policies and practices have historically allowed some population groups disproportionately greater opportunities for building household wealth. As the poet Ralph Waldo Emerson wrote, “The first wealth is health.” That saying has recently been revised to make the point that “wealth equals health,” a point forcefully driven home in the 2006 Handbook for Action: Tackling Health Inequities Through Public Health Practice. This handbook closely examined how U.S. household wealth (meaning the value of all financial and nonfinancial assets, such as real estate owned by a household, minus any debts) serves as the major determinant of health and health inequities, influencing and influenced by virtually all other upstream environmental and socioeconomic factors, including income, education, employment, housing, bank lending policies, child care, recreational opportunities, food supply, health care access, neighborhood safety, and environmental quality.
If health is wealth, it follows that efforts to understand and reverse the drivers of health inequities need to begin by looking at how the policies and actions of private institutions and governments have contributed to the large gaps in wealth that mirror the gaps between the healthy and the unhealthy.
For more information on how wealth affects health, please see the Office of Health Equity’s Portrait of Promise.
Technical Notes for Income Data:
Most of our indicators with income data come from the California Health Interview Survey, which provides data for about a third of our indicators overall. For these, income is stratified by multiples of the Federal Poverty Guideline (FPG) instead of dollars. The FPG adjusts for household size as well as inflation, facilitating comparisons over time. It is also used to determine eligibility for assistance with health insurance, food purchases, utility bills and many other programs. More information about the FPG, including how it translates into dollars, is available at: https://aspe.hhs.gov/poverty-guidelines.
Urban / Rural Categories:
The primary residence (home) of each California Health Interview Survey respondent was classified according to the Claritas Urbanization Model:
- Large City (“urban”)
- Small City (“second city”)
- Small Town or Rural
Definitions of each category can be found at the Claritas website.
Individual values are not shown (i.e. “suppressed”) in visual displays of indicators if the value is determined to be highly unreliable and, therefore, potentially misleading. “Unreliable” means here that even though one “best” value is available, values much smaller and/or much bigger are also quite consistent with the available data. For survey data, this unreliability often is the result of too small a sample size for the specific data being assessed and can be related to complex survey designs. For population data, such unreliability is generally due to a small number of occurrences of the thing being measured, and the natural variability of small numbers in a population.
For details, see below: “Let’s Get Healthy California: Data Suppression Procedures”
Let’s Get Healthy California data comes from variety of sources. These sources are structured differently, operate on varying reporting cycles, and may undergo changes from year to year. There is often a delay of a year or more before results are collected, “cleaned”, analyzed, and reported at the statewide level.
It is challenging to maintain consistent tracking of data sources over time. At times, an identified data source may be discontinued resulting in a data gap. Alternative data sources (proxy data) may be found which address a similar focus, but this data cannot be directly compared to the original baseline and target.
- Most indicators are based on the percentage of a group that reports or is measured to have a certain behavior or characteristic (e.g. percent of third graders….; percent of adults…)
- Other indicators are based on rates in the population of some characteristic (e.g. infant mortality).
- Rates are a special proportion calculation that take into account the size of the population (the “denominator”) and the number (the “numerator”) within that population that have the “event” or characteristic of interest (e.g. number of infant deaths among all live births)
- In some cases, “age-adjusted” rates are presented, rather than “raw” or “crude” rates. Age-adjusted rates use a mathematical adjustment that “removes” the effect of the age distribution on the rate in one population compared to the rate in another population. For example, a city with a very large population of adults who are aging may well have a higher “crude” rate of pneumonia deaths than other cities, solely because of the difference in the “age distribution” of the cities. Age-adjusted rates remove the influence of the age distribution, and make the comparison meaningful.
- Statewide indicators typically change slowly over time. Indicator values may increase or decrease slightly from year to year without representing a “significant change.” For the Let’s Get Healthy California 2015 update, statistical significance of indicator movement is calculated by comparison of confidence intervals (where available) or by variation greater than 10 percent between baseline and current rate.
Considerations for Understanding Indicator Data:
There are many factors that can complicate understanding indicator outcomes. None of these data sources are perfect and there are considerations that data users should be aware of when reading these data.
- For survey data, there can be changes from year-to-year in how exactly questions are asked and even the way the surveys are conducted, such that year-to-year comparisons must be understood with caution, even though they appear to be measuring the same thing.
- Some questions on surveys are asked at one point in time, and not again for several years, so updated results are not available every year.
- Some of the data are based on asking people sensitive questions, and the self-reported responses may not always be completely accurate.
- Baseline rates for Let’s Get Healthy California have been taken from the 2012 Task Force report unless otherwise noted. Due to changes in data source or methodology, sometimes these baselines have been updated to reflect the current indicator data source.
- For some indicators, especially at the county-level, it is necessary to aggregate data over two or more years in order to have sufficient numbers for statistical stability. For example, the “year” variable for some indicators may show “2011-2012”, indicating that data for 2011 and 2012 were combined so that they could be included on the Let’s Get Healthy California website.
- For some indicators, “other”, “unknown”, or “other/unknown” values for a particular stratum (e.g. race/ethnicity, education, etc.) are suppressed in the Let’s Get Healthy California visual displays because they may be misleading or otherwise difficult to understand.
Understanding Race / Ethnicity Data in the Context of Health Equity:
Health Equity refers to efforts to ensure that all people have full and equal access to opportunities that enable them to lead healthy lives.9
Differences in health status among population groups, including racial and ethnic groups, are known as health disparities. For Let’s Get Healthy California, we measure differences based on health goals, compare each measure to a target, and aspire to reach the best possible outcome attainable for all groups.
Health disparities between racial and/or ethnic groups are due to a combination of factors including, social and economic factors, health care and access to care, and health behaviors such as smoking or nutrition.
Health disparities related to race can also result from differences in experiences of discrimination.7 Some racial and ethnic groups are faced with discriminatory practices and policies that can impact socioeconomic status, potentially reducing options for education, housing and nutrition, as well as access to transportation, health care, and other services.
Some disparities in health between populations are considered “inequities” when they are rooted in social disadvantage and they can be considered unfair, preventable, and avoidable. Inequities are persistent, not occasional or random differences, between more and less advantaged social groups.8
Collection Process for Race / Ethnicity Data:
Race is a socially-determined classification based on perceived differences in physical appearance.1 One person’s race can govern how he or she experiences life in a race-conscious society.2 As a social concept, race can change depending on the context. Socially-determined race often does not correspond to genetic ancestry.1, 3-4 Research shows that there is more genetic variation within racial groups than between racial groups.4
Ethnicity is a person’s affiliation with a group of people that have a shared cultural heritage. That heritage can include customs, religion, or language.1 In data collection and presentation, United States federal agencies are required to use a minimum of two ethnicities: “Hispanic or Latino” and “Not Hispanic or Latino”. An Hispanic is a person of Mexican, Puerto Rican, Cuban, Central or South American or other Spanish culture or origin, regardless of race.5-6 This same ethnic classification is used in most Let’s Get Healthy California data.
People of different races can identify with one ethnicity, and people of the same race can identify with different ethnicities.
In the United States, government agencies are required by the Office of Management and Budget to use a minimum of five race categories when collecting racial data;5-6 these same categories are frequently used by the California Health and Human Services Agency and other organizations that provide data for Let’s Get Healthy California indicators:
- Black or African American;
- American Indian or Alaska Native;
- Asian; and,
- Native Hawaiian or Other Pacific Islander.
Other categories include “Other race” and “Multiracial”.
In most instances, a person’s race and ethnicity are obtained by asking the person to select an option from the two recognized ethnicity categories, and then to select one or more races from a list of the recognized racial categories in the United States. When people select more than one race they are identified as “Multiracial” or “Multi-race”. This method of race/ethnicity determination is known as self-identified race and ethnicity.5 There are also times when a person’s racial and ethnic category are chosen by someone else, such as a relative or service provider, for the collection of birth and death data, as well as some clinical and law enforcement data.
Race/ethnicity data involves inherent limitations, due both to the diversity of social interpretation of these classifications, and to the limitations of data collection systems to capture and convey that diversity. It is important to consider race/ethnicity data in the context of other data, particularly socioeconomic factors that may significantly impact health outcomes and contribute to disparities and inequities.
Let’s Get Healthy California – Data Suppression Procedures
Individual values are not shown (i.e. “suppressed”) in visual displays of indicators/measures if the value is determined to be highly unreliable and therefore potentially misleading. “Unreliable” in this context means that even though one “best” value is shown, values much smaller and/or much bigger are also quite consistent with the available data. For survey data, this unreliability often is the result of too small a sample size for the specific data being assessed, which is often related to complex survey designs. For population data (e.g. Vital Records or Hospital Discharge data), unreliability is generally due to a small number of occurrences of the indicator being measured, and the natural variability of small numbers in a population.
Indicator data sets for Let’s Get Healthy California come from a range of systems (e.g. the California Health Interview Survey, Vital Statistics, Hospital Discharge Data, and others). The Let’s Get Healthy California data team examined the types of data generated by each system, and applied suppression rules based on balancing the need to provide as much data as possible with the need to avoid presenting potentially unreliable or misleading data. One common measure of “unreliability”, the Relative Standard Error (RSE), was used as a suppression guidepost for each system if the RSE or the data needed to calculate it were available. The RSE is generally calculated as the “standard error” or statistical spread of an estimate, divided by the estimate itself.
Based on the above needs, the following suppression rules were used:
- Survey data (including the California Health Interview Survey) values are suppressed if the RSE is greater than 20% and, if statewide or non-stratified county-wide data, deemed likely to be misleading based on individual review of the data. The RSE was calculated using the standard methods noted above–this approach differs slightly from some other users of the California Health Interview Survey data, where the divisor for the RSE is 100 minus the percent, if the estimate is > 50%. While 30% or 23% cut points are more standard, we determined that a 20% cut point suppressed many potentially misleading values not suppressed with the standard values. We did not use the “100 minus the percent” approach because we determined that it suppressed many values that were unlikely to be misleading.
- Infant Mortality Rates and other vital statistics measures values are suppressed if the RSE is greater than 30%, and, if the statewide or non-stratified county-wide data, are deemed likely to be misleading based on individual review of the data.
- Hospital Discharge data values are suppressed if the RSE is greater than 30%, and, if statewide or non-stratified county-wide data, are deemed likely to be misleading based on individual review of the data.
- Self-reported race/ethnicity in the age of genomic research: its potential impact on understanding health disparities: http://www.humgenomics.com/content/9/1/1.
- Use of Race and Ethnicity in Public Health Surveillance: http://www.cdc.gov/mmwr/preview/mmwrhtml/00021729.htm.
- RACE, A project of the American Anthropological Association, http://www.understandingrace.org/home.html.
- Overview of Race and Hispanic Origin: http://www.census.gov/prod/cen2010/briefs/c2010br-02.pdf.
- Standards for the Classification of Federal Data on Race and Ethnicity, http://www.whitehouse.gov/omb/fedreg_race-ethnicity/.
- Jones CP 2001. Invited Commentary: “Race,” Racism, and the Practice of Epidemiology, http://depts.washington.edu/ccph/pdf_files/Jones_2001-Invited.pdf.
- Braveman & Gruskin 2003. Defining equity in health: http://jech.bmj.com/content/57/4/254.long.
- California Health and Safety Code Section 131019.5: http://www.cdph.ca.gov/programs/Documents/Health_and_Safety_Code_131019.5.pdf