Methods

Data Acquisition and Processing

The Wisconsin Health Atlas Obesity Prevalence Estimates use data on body mass index (BMI), age, sex, billing ZIP Code, and payer from individuals who visited a participating health care system in 2015 and 2016. BMI is recorded in health systems’ electronic health records as part of routine care. Health care systems securely transfer de-identified data to the Health Innovation Program (HIP), where the data are combined and stored on a secure server. Obesity Prevention Initiative staff prepare and aggregate data to create ZIP Code level estimates of obesity prevalence. These aggregated estimates are then removed from the HIP secure server for use on the Wisconsin Health Atlas website.

Data Quality

Participating health care systems review data for quality assurance before it is submitted to HIP for inclusion in the Wisconsin Health Atlas Obesity Prevalence Estimates. Obesity Prevention Initiative (OPI) staff further review submitted data for compatibility across systems. One record per patient from each health care system is used to create the prevalence estimates.

Definitions

Body Mass Index (BMI): Equal to weight in kilograms divided by height in meters squared.

Obesity: Obesity is defined differently for adults and children. For adults (ages 18 and over), obesity is defined as having a Body Mass Index of 30 or higher. For children (ages 5 to 17), obesity is defined based on the 2000 CDC Growth Charts. A child is considered to have obesity if their BMI is at or above the 95th percentile for their age in months and gender. For more information about obesity definitions, visit: http://www.cdc.gov/obesity/

Numerator: The total number of individuals with a valid BMI value that classifies them as having obesity, for a given geographic location (ZIP Code or statewide), age group, and (if applicable), sex and/or payer.

Denominator: The total number of individuals with a valid BMI value for a given geographic location (ZIP Code or statewide), age group, and (if applicable), sex and/or payer.

Coverage: Estimates the percentage of a population that is included in the available, valid patient data. Coverage is calculated by dividing the total number of patients in a ZIP Code by the estimated population size of the ZIP Code. ZIP Code population sizes are approximated using the 2015 American Community Survey (ACS) 5-year population estimates for ZIP Code Tabulation Areas. Coverage is calculated separately for adults and children for each ZIP Code. For more information on the American Community Survey, visit: https://www.census.gov/programs-surveys/acs/

Insufficient Data: ZIP Codes estimates are suppressed when there is not enough data available to create an estimate. Insufficient data is defined in five ways:

  1. Coverage less than 10%: The number of included individuals is less than 10% of the ACS estimated population size for the age group, gender and ZIP Code
  2. Numerator less than 10: Estimates based on a numerator of less than 10 individuals are suppressed to ensure patient anonymity
  3. Denominator less than 30: Estimates based on a denominator of less than 30 individuals are suppressed to ensure patient anonymity and for estimate stability.
  4. Population size less than 30: The ACS estimated population size for the age group, gender and ZIP Code is less than 30 individuals.
  5. Relative Standard Error (RSE) greater than 30%: The standard error of the prevalence estimate is equal to or greater than 30% of the prevalence estimate.

Confidence Interval: 95% Wilson (score) confidence intervals are calculated for each prevalence estimate.

Unadjusted Obesity Prevalence Estimates: Gives the approximate percent of the population in the ZIP Code, age, and sex group with obesity.

Limitations

These prevalence estimates are based on patient populations and reflect those who sought care from participating health care systems during 2015 and 2016. Only individuals who received care during the submitted period were included in these data, and estimates may not accurately reflect the prevalence of obesity in the Wisconsin population.

BMI values are based on heights and weights recorded during patient encounters. There are no standard practices for height and weight measurement across health systems and clinics. Some variation may occur in height and weight measurements based on the clothing and/or shoes that a patient wore during their height and weight measurement.

While estimates include only one visit per patient per health care system, it is possible that patients sought care from more than one reporting system. Such visits would be included as duplicate measurements.

Not all health care organizations that provide care to Wisconsin residents currently participate in this project. This results in limited or missing data for some ZIP Codes.

The geographies used to define coverage for ZIP Codes do not perfectly coincide. The ZIP Code Tabulation Area (ZCTA) boundaries are defined by the U.S. Census Bureau and are composed for Census blocks in which a majority of residents fall into a certain ZIP Code. For more information, see: https://www.census.gov/geo/reference/zctas.html

Disclaimers

Because of limitations in the available data, unadjusted obesity prevalence estimates presented by the Wisconsin Health Atlas may not be representative of the true obesity prevalence in the population.

System Notes

These data represent individuals who visited a participating health care system during the included period and had a valid BMI measurement recorded. They may not be representative of the entire population of a ZIP Code.

Valid BMI measurements are based on visits during which a valid height and weight were recorded.

All data sharing and preparation was conducted in accordance with Health Insurance Portability and Accountability Act (HIPAA) regulations to ensure privacy and confidentiality.

Data Sources

We collaborate with healthcare systems throughout the state as part of the Wisconsin Obesity Surveillance Partnership. Currently contributing health care organizations include:

  • Aspirus Clinics, Inc.
  • Columbia St. Mary's Community Physicians
  • Dean Clinic
  • Froedtert & The Medical College of Wisconsin
  • Gundersen Health System
  • Marshfield Clinic
  • Mayo Clinic Health System - Franciscan Healthcare
  • Mercy Health
  • Meriter Medical Group
  • Monroe Clinic
  • Prevea Health
  • ProHealth Medical Group
  • ThedaCare
  • UW Health Physicians
  • Wheaton Franciscan Medical Group

Future contributing partners:

  • Holy Family Memorial
  • Wildwood Family Clinic

The Wisconsin Obesity Surveillance Partnership welcomes new data partners.  For more information, please contact us at wihealthatlas@hslc.wisc.edu.

Percent coverage is calculated using population estimates from the American Community Survey 2011-2015 5-Year Estimates (U.S. Census).

Funding

Funding for this project was provided by the University of Wisconsin School of Medicine and Public Health through the Wisconsin Partnership Program.