Data Resources

Slider

“From Data to Action” provides resources for data literacy and interpreting the data in the neighborhood health reports in context.

“Making Sense of the Data” provides more details about the data included in the reports as well as frequently asked questions.

From Data to Action

Data from Neighborhood Health Partnerships is an important tool, but it is meant to complement your experience and knowledge of your local community. It can be a challenge to translate data into action. To help you get started, here are some resources to explain what data is and how to use it.

Getting to Know Data: Resources for Data Literacy

There are many resources that can help you use data and create a common understanding around data among your partners. We recommend the Data Literacy modules from the County Health Rankings and Roadmaps Action Center.

Interpreting Data in Context: What Numbers Do Not Tell You

Datasets—a collection of related items of data—are always incomplete. Any single dataset can only tell us part of the story of the health of a community and the factors that shape health. The links between neighborhood and health are complex, and there are many reasons why some communities have better health outcomes than others.

Health inequities are preventable and unjust differences between who does and does not have access to resources for living healthy and safe lives. Health inequities are often caused by social and institutional practices, both historical and ongoing. These root causes of health inequities are not reflected in the Neighborhood Health Partnership reports. Root causes are also not measured or emphasized as much as health outcomes are. Without the knowledge of root causes, it is hard to see the full picture of the health of a community.

An example of one health inequity is the burden of diabetes among people of color. The data in your report may show that a ZIP code with a larger population of people of color has lower rates of blood sugar control amongst people with diabetes. However, that dataset does not speak to neighborhood factors like the lack of a grocery store, safe outdoor spaces to move and relax in, or access to good jobs and a living wage. This is just one example of many. The health inequity (differences in blood sugar control rates) is visible in the data, while the root causes (lack of nutritious food, green space, and jobs) remain invisible. This is why local knowledge and expertise is essential, to explain what data sets alone cannot.

Without recognizing the role of the root causes of health inequities, we risk drawing conclusions about communities that could be untrue and harmful. It is important for all partners to recognize what is and is not represented in a dataset. Take caution not to draw conclusions from datasets that might reinforce harmful and limiting stereotypes about communities, and appreciate that a dataset can only ever tell us part of the full story. Using an equity frame is important when using data to advance efforts to create conditions in which all residents of a community can be healthy. For more information on the social determinants of health, visit our Resources page.

Getting to Know WCHQ Data: Understanding Your Report

The primary data source used for the Neighborhood Health Partnerships reports is electronic health records from health systems participating in the Wisconsin Collaborative for Healthcare Quality (WCHQ). This data comes from information that healthcare providers record in a patient’s electronic health record during primary care visits. They record information about vaccinations, screenings for cancer and other health risk factors, and management of diabetes and heart disease. The health systems then provide that data, while protecting patient confidentiality, for the reports. For all WCHQ measures, higher performance is considered better (e.g. higher vaccination and screening rates, higher percentages of people with their blood pressure under control).

The reports include any individual that received primary care from a participating WCHQ member health system. Because the data only includes WCHQ health care systems, the estimates based on this data do not reflect the entire population of the ZIP code. People living in the ZIP code who do not go to a WCHQ member health system are not included in this data. This includes people who get their health care elsewhere, those who do not have access to health care, and those who simply have not had a primary care visit in the last one to three years.

It is possible people sought care for the same health issue from more than one WCHQ health care system. For example, if a person who is covered by one WCHQ member health system visits another WCHQ member, that person’s health issue could be counted twice. Also, because some of the measures used here rely on patients who have received a formal diagnosis, people who have conditions such as diabetes or heart disease but do not yet have a formal diagnosis will not be counted in these numbers.

The report also shows the proportion of people from the entire population of a ZIP code that are included in your report. The population of the ZIP code is estimated using data from the American Community Survey (ACS). When a measurement is related to an underlying condition like diabetes, hypertension or heart disease, the prevalence of that condition in the ZIP code is estimated by using data from the Wisconsin Behavioral Risk Factor Survey (BRFS).

Some of the population sizes represented in the reports are small. This means that small fluctuations in health behaviors, care, and outcomes could have a large impact on the measure results. For example, if the population of people included in a report is 50 people, 15 of whom are smokers, just a few people quitting smoking may have a dramatic-looking impact on your reports.

To summarize, every dataset has its strengths and limitations. It is important to be able to understand the data in the context in which they exist. For more information on specific details about the data in the reports, read “Making Sense of the Data.”

Making Sense of the Data

This page provides more detailed information about the data in the reports. For a more general overview of the data and data literacy resources, see “From Data to Action.”

Where do the data come from?

The data used in the reports comes from multiple sources including the Wisconsin Collaborative for Healthcare Quality (WCHQ), the U.S. Census Bureau’s American Community Survey (ACS), and the Wisconsin Behavioral Risk Factor Survey (BRFS).

Data from the U.S. Census Bureau’s 2014-2018 American Community Survey 5-year estimates are used in these reports for population size and demographic information.

The American Community Survey (ACS) is an ongoing survey that provides information about people living in the United States and Puerto Rico. Survey questions cover many topics including demographics, jobs, education, housing, and more. More information about this data source is available at the links below:

Data from the Wisconsin Behavioral Risk Factor Survey (BRFS) from 2011-2017 was used in the reports to calculate prevalence estimates of heart disease, hypertension and diabetes in Wisconsin ZIP codes. See our FAQ on heart disease and diabetes prevalence for more information.

The Wisconsin BRFS is part of a national system of telephone surveys that collect data from U.S. residents about their health-related risk behaviors, chronic health conditions, and use of preventive services. BRFS randomly selects cell phone and landline users to participate in the survey. More information about this data source is available here: Wisconsin Behavioral Risk Factor Survey - Methodology in Brief.

Frequently Asked Questions

These reports include all WCHQ diabetes measures. Health systems that do not submit data for all diabetes measures were excluded from these reports to ensure a consistent reporting population. As a result, the number of people included in the diabetes profile reports may differ from that of other NHP ZIP code reports for the same corresponding diabetes measures.

Two of the diabetes measures (statin use and daily aspirin for patients with heart disease) have more specific eligibility criteria than the other diabetes measures. As a result, fewer people are eligible for these measures than the other diabetes measures.

The diabetes profile reports also compare performance of each diabetes measure in the ZIP code. Performance of each diabetes measure is benchmarked against all WCHQ member health systems that submit data (both patient level and in aggregate) in Wisconsin and bordering states from lowest performing (bottom quartile) to highest performing (top quartile) ZIP codes. These quartiles may differ slightly from that of other NHP ZIP code reports and maps due to the consistent reporting population used in these reports.

The health systems that are currently included in the reports are shown below. These may change as data from additional health systems may become available.

Note: Some of the systems shown may not be included in reports for all measures, and some additional health systems not pictured below publicly reported data in prior timeframes and are included in the trend line of the reports, but not the rest of the report content.

Health systems in Wisconsin that do not submit and publicly report patient-level data to WCHQ are not included in these reports. This currently includes most Federally Qualified Health Centers (FQHCs), Indian Health Service clinics, and some other clinics and health systems in Wisconsin. We hope that additional health systems will be able to submit data to WCHQ in the future, which would make these reports more representative of Wisconsin neighborhoods. If you represent a clinic or health system and are interested in submitting data for this project, contact reports@hip.wisc.edu for more details.

The WCHQ data includes ZIP codes from patient addresses. These ZIP codes were aggregated to the ZIP Code Tabulation Area (ZCTA) level so that we could link to comparable ACS data for condition prevalence and coverage estimates.

Some ZIP codes that mapped to non-residential areas (e.g. P.O. Boxes) were excluded. To protect confidentiality, reports are not distributed for ZIP codes with sample sizes of less than 50 people and/or ZIP codes in which estimated coverage is less than 10% of the ACS population based on the selected measure.

Coverage refers to the number of individuals who receive primary care from a WCHQ member health system, compared to the ACS 5-Year Estimates. We have calculated estimated coverage in each ZIP code by taking the total number of individuals in the WCHQ data (numerator: eligible patients who receive primary care from a WCHQ member health system), divided by the ACS estimated number of eligible individuals (denominator).

None of the measure definitions have changed since the beginning of our reporting period for these reports in December 2016.

Estimated coverage refers to the number of individuals who receive primary care from a WCHQ member health system that are eligible for the selected measure compared to the number of individuals estimated to live in the selected ZIP code that are eligible for the selected measure. These comparison estimates are calculated as follows -

For diabetes, heart disease, and hypertension measures: WCHQ measures related to diabetes, heart disease, and hypertension require that a person has a diagnosis of the condition in their electronic health record. Weighted sampling is used to estimate the prevalence of these conditions. The Wisconsin Behavioral Risk Factor Survey (BRFS) asks a sample of people if they have chronic conditions and calculates disease prevalence estimates at the county level, which is then used to estimate the prevalence of diabetes, heart disease, and hypertension for each ZIP code. This estimated prevalence is then combined with data from the American Community Survey (ACS) to estimate the number of individuals with the condition in each ZIP code.

For all other measures: Data on age and sex from ACS are used to estimate the number of patients eligible for a certain measure that are living in a ZIP code.

In some cases, the eligible population in the report is larger than the ACS/BRFS estimated population for a measure in the ZIP code. In these cases, the eligible population is used instead of the ACS/BRFS estimated population. While the ACS and BRFS are the best sources for detailed population information, they are still just an estimate for one point in time. Portions of the population may be missed and/or the population may fluctuate over time as people move in and out of a particular community.

The highest performing ZIP code for a measure is determined from among residential ZIP codes in the selected county with over 50 people eligible for the measure and an estimated coverage of at least 10% of the ACS population for the measure.

In rare cases, the county average may appear higher than the highest performing ZIP code in the county. This is because all residential ZIP codes are included in the county average, but only residential ZIP codes meeting the measure eligibility and estimated coverage thresholds are included for the highest performing ZIP code in the county. This is more likely to occur in rural counties with fewer residents.

Time points on the trend line will only be displayed if there were 50 or more patients in that reporting period.

Demographic data is included to provide information on some of the characteristics of individuals included in the report and how it compares to data from the ACS. We included measures that we had quality data for (sex, age, insurance status, and race/ethnicity) in the WCHQ data so that it could be compared to the ACS data. The reports provide information about people seeking care at WCHQ member health systems, and therefore may not be representative of the total population in a neighborhood.

It is normal to expect that some demographic measures for the people included in the report will look different from those of the overall ACS ZIP code population. For example:

  • Breast cancer screening, cervical cancer screening, and chlamydia screening measures only include females, so the population in the report will be 100% female.
  • Childhood vaccinations and well child visit in the first 15 months of life measures are limited to children, so the population in the report over age 65 will be 0%.
  • The pneumonia vaccination measure is only measuring people over age 65, so the population in the report will be 100% over age 65.
  • The percentage of uninsured individuals in the reports is often lower than the overall ZIP code population because people who do not have health insurance are less likely to go to healthcare providers due to cost.
  • Many of our reports include a larger percentage of people over age 65 in the ZIP code than what is reported by the ACS. This is likely because older adults tend to go to the doctor more than younger people. Additionally, females tend to go to the doctor more than males.

Providing this contextual information helps users recognize strengths, limitations, and representativeness of data in the reports.

Data about sex in the report (Report % female) comes from the electronic health records submitted by WCHQ member health systems and refers to biological sex. In the WCHQ data, options for sex are: male, female, or missing. Data collection methods for sex may vary by health system. Any patient records with unknown sex are excluded when calculating the "Report % Female."

Data about sex from the ACS (ACS % female) is self-reported by individuals that complete their surveys. The ACS provides two options for sex: male and female. More information can be found here: Why We Ask About...Sex | American Community Survey | US Census Bureau.

It is important to note that data on sex is limited by current data collection methods using binary categories of male and female, and therefore may not accurately encompass all individuals, such as those that are intersex or trans.

The age of persons in the report is calculated using their date of birth in the electronic health records submitted by WCHQ member health systems.

The age of persons in the ACS data is collected by asking for age and date of birth on surveys. More information can be found here: Why We Ask About... Age and Date of Birth | American Community Survey.

The insurance status for persons in the report comes from the primary payer field in the electronic health record. The payer categories mapped include Commercial, Medicaid, Medicare, and Uninsured. Any patient records with unknown insurance status are excluded when calculating the "Report % Uninsured."

The insurance status for persons from the ACS is self-reported. The survey asks if the respondent is currently covered by one or more of a list of health insurance types. More information can be found here: Why We Ask About... Health Insurance Coverage | American Community Survey.

Data on race and ethnicity in the report comes from the primary race and ethnicity fields in the electronic health records submitted by WCHQ member health systems. WCHQ provides member health systems with resources containing guidance on collecting race and ethnicity data. These include the Office of Management and Budget Standards, the Health Research & Education Trust Disparities Toolkit, and the Minnesota Community Measurement Handbook on the Collection of Race/Ethnicity/Language Data in Medical Groups. Actual data collection methods for race/ethnicity may vary by health system. Any patient records with unknown race and ethnicity are excluded when calculating report race/ethnicity percentages.

Data on race and ethnicity from the ACS is self-reported by those who fill out the survey. More information about collecting data on race and the choices in the survey can be found here: Why We Ask About...Race | American Community Survey | US Census Bureau.

There is also a question about Hispanic or Latino origin which is to be filled out in addition to the question about race. For this survey, Hispanic origins are not races. More information can be found here: Why We Ask About...Hispanic or Latino Origin. The answer to these questions results in combined race/ethnicity categories in which ethnicity trumps race.

These reports provide data about people who have sought care in 2016-2018 at health systems that report data to WCHQ. People who do not receive primary care from a WCHQ member health system that publicly reports their data are not included. This includes people who get their health care elsewhere, those who do not have adequate health care, and those who simply have not gone in for a primary care visit during this period of time. As a result, these reports may not accurately reflect the entire population of the ZIP code.

The numbers provided in the reports include only one visit per patient, per health care system, but it is possible people sought care for the same health issue from more than WCHQ member health care system. For example, if a person who is covered by one WCHQ member health system visits another WCHQ member health system, that person’s health issue could be counted twice for a given WCHQ measure. Also, because some of the measures used here rely on a person receiving a formal diagnosis, people who have conditions such as diabetes or heart disease, but do not have a formal diagnosis would not be counted in these numbers. More detailed measure specifications are available here: WCHQ Measures at a Glance – Fall 2020.

WCHQ is currently in the process of transferring to a new data vendor. This will lead to delays in processing the 2019 and 2020 data. We expect new data to be available at regular intervals (every 6-9 months) beginning in fall 2021.

 

For additional questions about the data and interpreting these reports, please contact the NHP team at nhp@hip.wisc.edu.