Technical notes

Confidence Interval (also known as error bar) is the range of values that includes the true value 95% of the time. If the confidence intervals of two groups do not overlap, the difference between groups is considered statistically significant (meaning that chance or random variation is unlikely to explain the difference). The wider or larger the confidence interval, the less precise the estimate. For example, if an indicator had a rate of 42% (CI: 18%, 66%), the 18% is the "low end" of the confidence interval, 66% is the "high end", and the estimate is 42%. That means that it is highly likely the "true value" would fall within the 18 to 66 percent range. Another sample or survey year might have a value of 32%, but because of the wide intervals, is unlikely to be a true difference. Confidence intervals are placed in a text box over each point or bar, and are presented visually as an error bar in a chart. Confidence intervals are necessary since observed health statistics (counts or rates computed from surveys, registries, or other systems), may not accurately capture the underlying risk in the population. Observed rates can vary from sample to sample or year to year. Surveys, which are based on a population sample, have "sampling error," or random variation due to having only a subset of the population. Sampling error and random variation increase when the population or number of events is small.

• Counts give the estimated number of health events, which can be helpful for understanding the magnitude of the problem, or how many people are affected. In areas where there are more people, counts are often higher, so we calculate rates to allow for comparison.
• Rates are usually expressed as the number of events per population per year. When this applies to the total population (all ages), and is not adjusted, the rate is called the crude rate.
• When the measure is calculated for a specific age group (e.g., age 15-24), it is called the age-specific rate.
• Crude and age-specific rates present the actual magnitude of an event within a population or age group.

Since diseases vary by age, a rate that is not affected by differences in the age composition of the populations allows for fair comparisons between geographies. This is the age-adjusted rate. For example, if a neighborhood with a high proportion of older people also has a higher-than-average death rate, it will be difficult to determine if that neighborhood's death rate is higher than average for residents of all ages or if it simply reflects the higher death rate that naturally occurs among older people. The age-adjusted rate mathematically removes the effect of the population's age distribution on the indicator. By convention, the rate is adjusted to the age distribution of the estimated 2000 US population. In this report, age-adjusted rates per 100,000 are presented unless otherwise noted.

Infant mortality, maternal smoking, and other maternal/child health measures are calculated with live births as the denominator and presented as a rate per 1,000 live births (infant mortality) or percent of births (preterm, low birth weight, etc.).

Prevalence rates (the number who are living with a condition) from the Behavioral Risk Factor Surveillance Survey are expressed as percent of the adult population, usually ages 18+. Exceptions to the age range are noted in the indicator title. These rates are not age-adjusted. Prevalence rates from the Healthy Youth Survey (HYS) are for public school students in the specified grades. The HYS is only asked of students in grades 6 (abbreviated version), 8, 10, and 12.

Incidence rates (the number of new cases diagnosed in a calendar year) are presented for cancer and HIV/AIDS.

Titles show the years that are included in each analysis. A dashed line indicates that all years were used (e.g. 2011-2015 includes 2011, 2012, 2013, 2014, and 2015), where a comma indicates a break in years (e.g. 2011, 2013, 2015 is only those specific years).

Time trend graphs are shown as rolling averages (averaged rates that overlap) to visually minimize variation of unstable estimates by smoothing the trend line. For relatively small populations, small changes in the number of events will cause the rate to fluctuate substantially and create jagged lines of statistical "noise" with peaks and valleys. To help stabilize the rate and observe the time trend of an event, rates are sometimes aggregated into 3-year "rolling" averages across the total observed period. For example, for events for years 2000 to 2013, the rates are reported as three-year rolling averages: 2000-2002, 2001-2003...2011-2013. Adjacent data points will contain overlapping years of data. The overlap between time intervals makes statistical tests for trend invalid.

Geographies: Whenever possible, indicators are reported for King County as a whole and for four regions within the county. If enough data are available for a valid analysis, they may also be reported by smaller geographic areas (cities, neighborhoods within large cities, and groups of smaller cities and unincorporated areas). Education data are reported by school district. For more detail, plus maps, see our CHI geographical definitions page.

Cities and neighborhoods (also known as Health Reporting Areas (HRAs)): Health Reporting Areas are sub-county geographies based on Census 2020 results for reporting health statistics.
The 61 Health Reporting Areas (HRAs) are aggregates of Census 2020 blocks created to facilitate the analysis and presentation of sub-county health statistics. Where possible, HRAs are defined as neighborhoods within large cities, smaller cities, unincorporated areas in King County, or a combination of these geographies, with a Census 2020 population over 30,000.

Federal Poverty Guidelines, issued by the Department of Health and Human Services are a simplified version of the federal poverty thresholds. The guidelines are used to determine financial eligibility for various federal, state, and local assistance programs. For a family of 4, the federal poverty guideline was $27,750 in 2022; in 2023 it was $30,000.

Neighborhood poverty levels are based on the proportion of people in a census tract in with an annual household income (as reported in the US Census Bureau’s American Community Survey) that falls below 200% of the federal poverty level.

• Low poverty: less than 10% of the population is below 200% of the federal poverty level. Using this criterion, 22.9% of the King County population lives in low-poverty neighborhoods.
• Medium poverty: 10% to 14% of the population is below 200% of the federal poverty level. Using this criterion, 28.0% of the King County population lives in medium-poverty neighborhoods.
• High poverty: 15% to 24% of the population is below 200% of the federal poverty level. Using this criterion, 26.1% of the King County population lives in high-poverty neighborhoods.
• Very high poverty: 25% or more of the population is below 200% of the federal poverty level. Using this criterion, 22.9 % of the King County population lives in very high-poverty neighborhoods.

This neighborhood-level characteristic is used where individual measures of income or poverty level are not available.
For area-based measures of poverty, a census tract is considered a neighborhood. Data sources where census tract information is not available use ZIP Code Tabulation Areas (ZCTAs) to designate the neighborhood.

Race/ethnicity and discrimination: Race and ethnicity are markers for complex social, economic, and political factors that can influence community and individual health in important ways. Many communities of color have experienced social and economic discrimination and other forms of racism that can negatively affect the health and well-being of these communities. We continue to analyze and present data by race/ethnicity because we believe it is important to be aware of racial and ethnic group disparities in these indicators. We hope this will lead to strategies that address these issues, as well as the social and economic inequities which underlie them.

Race/ethnicity terms: Federal standards mandate that race and ethnicity (Hispanic origin) are distinct concepts requiring two separate questions when collecting data from an individual. "Hispanic origin" is meant to capture the heritage, nationality group, lineage, or country of birth of an individual (or his/her parents) before arriving in the United States. Persons of Hispanic ethnicity can be of any race. 2010 Census terms: Hispanic or Latino, Not Hispanic or Latino, White alone (Not Hispanic or Latino), Black or African American, American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, White, Some Other Race, and Two or More Races. Persons of Hispanic ethnicity can be of any race and are included in other racial categories. Racial/ethnic groups are sometimes combined when sample sizes are too small for valid statistical comparisons of more discrete groups.

Some surveys collect race/ethnicity information using only one question on race. These terms are Hispanic, Non-Hispanic, white Non-Hispanic, Black, American Indian/Alaska Native (AIAN), Asian, Native Hawaiian/Pacific Islander (NHPI), white, and Multiple Race (Multiple).

Sexual orientation:

Indicators include Lesbian, bi-sexual, and gay (LGB) demographics where the data are available, in the Behavioral Risk Factor Surveillance System and the Health Youth Survey data. HYS sexual orientation data are only available in 2016.