February 2002, (Vol. 3, No. 2)
Why Did Wisconsin Fall in State Health Rankings?
By: Paul Peppard, Ph.D., David A. Kindig, MD, Ph.D., and Patrick Remington, MD., MPH

Introduction

Beginning in 1990, in a series of highly publicized annual reports, UnitedHealth Foundation (formerly Northwest National Life and ReliaStar) has ranked each of the 50 states by a composite measure of population health. This composite measure was developed to reflect multiple facets of states' population health, including health behaviors, access to health care, and morbidity and mortality. Generally, Wisconsin has done well, ranking 6th "healthiest" in 1990, peaking at 4th in 1997, but falling to 11th in the latest 2001 ranking. To what can this recent decline in rankings be attributed? This issue brief contrasts the first UnitedHealth report with the 2001 report in an effort to explain the net decline.

Methods

Data used to create the UnitedHealth rankings for years 1990-2001 were obtained courtesy of UnitedHealth Foundation. UnitedHealth assembled data from a variety of publicly-accessible sources including, for example, vital statistics and public health surveillance systems. In all, 17 different health-related data sources ("components") were used to calculate rankings. Each component was assigned a contributing "weight" to the overall rankings, ranging from 2.5% to 10%. Specific weights were assigned by UnitedHealth, in consultation with an expert panel, to reflect the degree to which the component reliably and independently measured overall health. For example, annual total mortality rate was given a weight of 10%. A complete description of the rankings and methodology is detailed in America's Health: UnitedHealth Foundation State Health Rankings 2001 Edition, available online at: http://www.americashealthrankings.org.

Findings

The table (backside) presents the 17 health components of UnitedHealth rankings, along with each component's weight; value for 1990 and 2001; and rank for 1990 and 2001. The health components are divided into 5 categories, intended to span important domains of population health. Browsing the table, one can see, for instance, that Wisconsin ranked 50th (i.e., worst among all states) in "heart disease risk" in 1990, but rose to 16th by 2001. The value for heart disease risk in 1990 was +18%, indicating that Wisconsin was 18% worse than the US average in a combination of lifestyle-related heart disease risk factors, including overweight and obesity, physical inactivity and hypertension. By 2001, Wisconsin had 4% lower "risk" than US average.

Along with the UnitedHealth rankings for individual components, we calculated ranks for each health category. In each health category, Wisconsin's rank worsened from 1990 to 2001. Note that, in some instances, the values used by UnitedHealth do not precisely correspond to actual state values in a given year. For instance, "total mortality" is a three-year race- and age-adjusted average mortality rate. The final column of the table presents the effect of the Wisconsin's relative change in health-component score on Wisconsin's overall UnitedHealth 2001 rank. This column is best described by example: Wisconsin's actual 2001 overall rank was 11th healthiest out of the 50 states. However, had Wisconsin maintained its status as having the 6th lowest smoking prevalence in 2001, as it was in 1990, Wisconsin would have been ranked 8th overall in 2001 (thus the "8" in the "smoking prevalence" row of the final column). The same is true of infant mortality: had Wisconsin maintained its 1990 infant mortality ranking, the state would have ranked 8th overall in 2000. Conversely, had Wisconsin not improved its "heart disease risk" rank from 50th to 16th, Wisconsin's overall 2001 rank would have been even lower-13th. Even though other health components' values and rankings increased or decreased from 1990 to 2001, none of those changes had an appreciable impact on Wisconsin's overall decline in rank. This is because the other factors did not have as large a magnitude in change in rank and/or the factor was weighted less in the overall rankings.
chart

Figures 1 and 2 depict, respectively, trends in smoking prevalence and infant mortality in Wisconsin and the US. For both components, there has been a net improvement in the actual degree of tobacco use and infant mortality in Wisconsin and the entire US over the last decade. However, the declines in smoking prevalence and infant mortality have been more pronounced in the US as a whole. In 1990, US smoking prevalence and infant mortality rate were both more than 10% higher than Wisconsin’s.
Figure 1:  Wisconsin and US smoking prevalence Wisconsin and US infant mortality
By 2001, the US had declined faster in both, nearly matching the Wisconsin infant mortality rate, and reducing tobacco use even below Wisconsin’s prevalence. Thus, there has been a net worsening of Wisconsin’s rank in both casesfrom 6th to 34th for smoking and 8th to 21st for infant mortality.

Conclusions

Wisconsin’s decline from the 6th healthiest state in 1990 to 11th in 2001 is largely a result of Wisconsin not keeping pace with the declines in tobacco use and infant mortality seen in the US as a whole. If Wisconsin had maintained both its 1990 smoking and infant mortality ranks (6th and 8th, respectively) the state would have ranked 5th, rather than 11th overall in 2001.

A further decline in overall ranking was avoided by a strong improvement in heart disease risk. Over the 1990’s, based on several measures of health, Wisconsin’s health has been consistently in the upper tier of states. However, that position is threatened if Wisconsin continues its current trend, relative to other states. Wisconsin has experienced at least small decrements in ranks among all UnitedHealth health categories including lifestyle, access to health care, disability, morbidity and mortality. The relative declines have been most pronounced in the lifestyle and mortality categories.

Despite its usefulness, the UnitedHealth health rankings depend, to varying degrees, on the particular choices of health components, the weights that the components are assigned, and the accuracy with which each health component is measured. The Wisconsin Public Health & Health Policy Institute is currently working to extend the UnitedHealth approach to Wisconsin counties and communities. In doing so, we will address these issues in an effort to develop measures that can usefully summarize, assess and track progress in Wisconsin’s populations’ health.