Health Indicators Overview Transcript

Event Started: 1/20/2010

Welcome to the first of the four part health indicators webinar series brought to you by the National Information Center on Health Services Research and Health Care Technology, a component of the National Library of Medicine. We are delighted to introduce Cheryl Wold, our presenter today, who will provide an overview of health indicators. Cheryl has an MPH in epidemiology from Boston University School of Public Health, extensive experience in public health evaluation in Los Angeles County, and is a recognized expert in health indicators including developing a compilation of health indicator projects for the Institute of Medicine.

Hi, thanks very much. I am thrilled to be with you today to provide an overview of health indicators. As Kate mentioned, this is the 1st module in a 4-part webinar series.

Module 2 will explore the community health status indicators, an online indicator resource; Module 3 will provide a case study of librarian involvement in indicators; and Module 4 will provide an in-depth look at a few health indicator reports and on-line tools. At the end of today's presentation, I hope you will understand the variety of health indicators, their data sources, key attributes and uses along with contextual concerns. I will first describe health indicators, identify common health indicators and data sources, and then describe how health indicators are used and presented along with some examples. As for my background, I have worked several years in public health developing health indicators and data sources both in the Boston area and here in LA County and more recently in my consulting role I conducted a review for the State of the USA which is creating a comprehensive web-based set of indicators for the U.S. ; this effort is set to launch in March. This review was provided to an Institute of Medicine committee that was charged with selecting 20 health indicators for the nation. This committee had to assure that the recommended indicators were highly credible and representative of the best thinking and data sources on health indicators. The review summarized the range of health indicators and the variety of ways in which they were being packaged and used in indicator reports as background for the IOM committee in commencing its work.

It is helpful to understand the broad definition of health and this is put forth by the World Health Organization. It says that health is the state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity. It goes on to state that it is the ability to change or cope with the environment - to define health as a resource for everyday life. This is a positive concept. And with the goal of good health in mind think of an indicator as a measure that helps quantify the achievement of a goal. Mark Friedman is well known for his work in writing about indicators and his leadership and results-based accountability work. And the assumption behind this statement is that the indicator quantifies a meaningful result when it comes to health. The real power of indicators is that they are designed to provide standardized comparable measures of health over time between groups of people and across geographic areas in the most essential domains of health. The attributes of a good indicator are also described by Mark Friedman, who boils it down to three essential criteria: data power, proxy power, and communication power. Data power is that the data for the indicator were created from high-quality data collection methods, complete, unbiased and that the data are available and updated on a regular basis. Second is proxy power and that is that the indicator says something important about health - that it's meaningful to health. In addition it pulls other indicators along with it. One example is the immunization rate; it is not only an important indicator for child health and the prevention of communicable diseases but it also indicates that kids have access to care. Third, communication power is whether the indicator speaks to a broad and diverse audience. If you had to explain the result you want for your community, for example, healthier behaviors, what two or three indicators would you choose so that your friends and neighbors would understand what is meant by healthier behaviors? You might cite decreases in smoking rates or increases in physical activity, for example.

So that health indicators are meaningful to health they are based in conceptual models for what influences health which in turn are rooted in scientific research. Such models have evolved over time as have definitions of health but trace historical improvements in health, for example, increases in life expectancy and declines in mortality due to infectious diseases; context is really important. For example, in 1900 the leading causes of death were pneumonia, tuberculosis, diarrhea and enteritis as compared to heart disease, cancer and stroke in more recent times.

Just as it's important for health indicators to be rooted in science, it is also important that they are relevant to the pressing concerns of today. An example of a current concern is the unsustainable rate of health care spending and furthermore that our level of spending is not consistent with our level of health, quality of life, and the performance of our health care system - nor is health always equitably distributed. What this means is that there are many opportunities for prevention and increasing the rate of return on investments in health. So current frameworks for health indicators address the burden of chronic diseases, for example, premature mortality and health care costs, they address complex health problems such as violence, obesity, HIV transmission, as well as health related behaviors such as diet, physical activity and sexual practices. In addition current frameworks address the growing literature on the powerful influences of social and environmental conditions on health outcomes and these are either direct influences, for example, children in poor neighborhoods are more likely to live in poorly maintained buildings and thus be exposed to lead or these can be indirect influences, for example, counties with lower socioeconomic levels have reduced average life expectancies compared to more affluent counties. Some of the most familiar frameworks are shown here and they often appear in indicator reports either alone or merged together and without going into too much detail let me quickly run through them. The first one - the broad determinants of health model - is widely acknowledged and holds that the physical environment, the social environment, health behaviors and access to medical care all profoundly influence health, and individual factors such as human biology and genetics and individual behaviors are important but it is how these factors interact with environmental factors in a dynamic way that produce health or illness. The second is the life course model. This emphasizes conditions and experiences early in life in establishing our health trajectory. That trajectory can be altered through our interactions not only with our environments but with systems and policies that affect us at different developmental stages throughout the life course. For example, children who are screened and receive services for developmental problems early in life will do much better than those children with the same problems detected later in life when the problem will be more compounded. And the third, the social determinants, is very similar to the broad determinants but it emphasizes those educational, economic, and social conditions and interactions that shape multiple health outcomes.

Indicators are powerful tools for monitoring and communicating critical information about population health. They are useful to support planning, for example, identifying priorities or targeting resources, and for tracking progress towards broad community objectives. They help to engage partners into civic and collaborative action by helping to build awareness of problems and trends. In addition they help inform policy and policy makers and can be used to promote accountability among governmental and non-governmental agencies.

Thousands of diverse groups all over the country and internationally use health indicators and their compilation is typically the result of public-private collaboration. Often health indicators are compiled at the state and local level by Public Health Departments with other agencies such as nonprofits, foundations, philanthropic organizations, businesses, civic organizations, and universities. Internationally, the Organization for Economic Cooperation and Development and the World Health Organization compile health indicators.

The most common indicators are measures related to birth and death and these include life expectancy, premature mortality, such as years of potential life lost, age specific death rates, costs specific death rates, birth to teens, the proportion of babies born at very low or low birth weight, the proportion of births to mothers who received adequate prenatal care, as well as the social characteristics of the parents such as maternal educational attainment at the time of birth, . These indicators are robust--they're complete, widely available, and comparable across jurisdictions, but they are also limited for several reasons. For example, as mortality rates overall decline and people are living longer with chronic conditions, there's more of a focus on disease prevention and reducing disparities. Also these indicators do not tell us about the risk factors that are driving some of these poor health outcomes.

Other invaluable indicators are measures of morbidity health status. These include health related quality of life indicators such as poor health days and self-rated health, and these are useful measures of overall health and functioning, and predictive of the use of health-care services as well as premature mortality. Other indicators include obesity, diabetes and other chronic diseases and are important both for their impact on premature mortality and the use of services but also they are amenable to prevention strategies.

Health-related behaviors are critical indicators and these are ultimately what drive an estimated 50% of all mortality and are the object of many policies and other interventions. Stated in a positive here, these include not smoking, regular physical activity, diet and nutrition and moderate or no drinking and not using drugs. Access to health care and interactions with the health care system are also important. These include having health insurance, having a regular source of care, avoidable hospitalizations--these are hospital admissions that could have likely been prevented with timely and appropriate primary care, and the receipt of preventive services.

Some of the data sources for the common indicators include the national vital statistics system and lots of surveys including the Behavioral Risk Factor survey and the National Health Interview Survey and many others. There are disease surveillance systems - at local health departments - as well as administrative data such as Medicare and Medicaid claims and hospital billing data. What is important to understand is that there is a whole web of data collection and dissemination that supports indicators and most of them are from public sources and require a lot of secondary data analysis for dissemination. And because it's important to get this information out, these activities are increasingly supported by public private collaborations.

Some examples of some commonly used social indicators include area-based measures such as income, poverty, and population density, access to public transportation and housing, environmental pollution such as air and water quality. And for the social environment individual or family income, educational attainment such as a high-school graduation rates or reading/doing math at grade level, and measures of social support and connectedness.

Additional social indicators include those that are Child and Family focused for example parent educational attainment, parenting practice such as breast feeding, reading to children and other family routines. Access to service such as early developmental screening and the family's social environment such as parent stress or depression or other factors that impact children.

Some data sources for the indicators include the Census, the American Community Survey and the Current Population Survey, the National Survey of Child Health, data from the Education Department such as the National Assessment of Education progress, data from the Environmental Protection Agency, such as Air Quality Monitoring data and the Toxics Release Inventory. These federal sources are supplemented on the state and local level with administrative data from any number of systems such as human services, educational, and justice systems.

Of course, there are limitations to available data and there are some notable gaps. Real quickly these include data about early life and conditions throughout childhood. Kindergarten entry is important with so much emphasis on school readiness and yet we do not have a lot of good data about that time period. Another example is childhood obesity data which is really important and is currently only really consistently available on the national level. In addition data for older adults including community care and support, long-term care needs and the availability of quality of services is incomplete. At the same time, data sources are constantly evolving based on science and do respond to health trends. At this point it might be useful to point out that data - the development of data -- for indicators is a very dynamic process and while some measures have settled others are still evolving. For example, refinements in how data is collected in surveys can occur and this can result in discrepancies of indicators over time. One notable example is all of the different ways there are to assess levels of physical activity; I'm not sure the science is yet settled around that.

Another big area for health indicators are those for help system performance. Briefly, these fall into three major categories including access and those factors affecting it such as trends in health insurance enrollment, and the primary care provider supply in rural areas. They include costs and economic measures, such as total health expenditures and prescription drug costs. And they include measures related to the quality of care - quality and effectiveness including patient safety.

The data sources for health systems performance indicators include surveys of patient experiences, the Medical Expenditure Panel Survey which is an interesting survey of households medical providers and insurance companies for a sample of families annually. They include billing data for hospitalizations that are generally reported to the state agency that regulates the industry. They include federal claims data such as Medicare as well as data that are voluntarily reported to the Agency for Healthcare Quality and Research for example or mandatory measures such as HEDIS measures - the Healthcare Effectiveness Data and Information Set - and also include a variety of public and private studies.

Now I want to switch gears and discuss some uses of health indicators and come back to the report that I mentioned earlier conducted for the State of the USA. The review that I mentioned was based on examples from reports currently in circulation. The criteria I used for selecting the reports was that they were created from high-quality and currently available data, that they used expert input as well as broad participatory processes, that they were designed for dissemination to diverse audiences, and that they represented a variety of geographic regions including international examples.

The review identified the various types of health indicators reports they can be grouped into three categories. These are the general health indicator reports, produced on the national, state, local and international level, the quality of life or comprehensive indicator systems. And these are indicator systems where health is one component among eight or ten other components that gauge the well being of a particular geographic area or jurisdiction. They also include health system performance reports. The review also identified the universe of health indicators and their attributes, data sources and their ability to be disaggregated down to specific geographic levels, and the range of practices related to accessing and presenting the indicators as well as gaps in available data.

These reports have several features and focus areas. I want to show you some examples of reports of the three different types of health indicator reports that I mentioned as well as some notable features.

First are the reports on disparities and social indicators and the first example is the Georgia Health Disparities Reports reported by the state of Georgia Public Health Department. This provides a county by county assessment and each county is graded on the extent of disparity or equity and they offer a clear criteria for grading as well as improvement. The other example is the Health of Wisconsin report card and they use just a few measures but they do a really nice job of grading all of these measures by various social criteria, for example educational attainment; I will show you an example in a second. The other resource is from the Commission on a Healthier U.S. that is supported by the Robert Wood Johnson Foundation; they offer educational material about the importance of social factors and disparities in health outcomes. They compare health indicators in the U.S. with other countries. This site also has slides and graphics and other useful resources and is very accessible.

Here is an example of the Health of Wisconsin report card. This shows working age adults from ages 25 to 64. This is measured from survey data - a health related quality of life measure - the average number of days that a person was in poor health in the past 30 days. Shown here for example by educational attainment people with a high-school education or less reporting an average of 6.2 days their health was poor in the past 30 days, receiving a letter grade of C. If you look among college graduates reported an average of 4.1 unhealthy days receiving letter grade A. This is a nice way of boiling down some complicated information to tables that can be easily shared with media and policymakers as well as communicated to the general public.

Many reports feature rankings, benchmarks and trend data, and make effective use of interactive graphics such as maps. The first example is America's Health Rankings produced by the United Health Foundation, the American Public Health Association and the Partnership for Prevention. It was static when I conducted my review and is now available on line; I will show you a bit more later. The next is the Big Cities Health Inventory produced by the National Association of City and County Health Officials; this ranks the 54 largest urban areas in the U.S. on a variety of indicators and includes Healthy People 2010 objectives as benchmarks. The third example is the State Scorecard on Health System Performance from the Commonwealth Fund which provides rankings and multiple comparisons that I will show you in a minute.

This is the first example from America's Health Rankings, which has trend data and rates the nation's progress overall on approximately 15 indicators. It provides rankings by state and information about significant changes in those indicators. It highlights obesity related health problems and costs and other special health topics and also discusses these overall trends in terms of disparities and other important contextual information.

This is the State Scorecard on Health System Performance; it was static when I did the review but is now highly interactive. This ranks states into quartiles for aggregate measures of access, prevention and treatment, avoidable hospital use and costs, equity and healthy lives -- there are about 12 indicators in each of these categories and this is an example of those measures rolled up into summary measures and all of the states ranked. This report site also features some PowerPoint slides that can be downloaded and that is the reason why this slide looks so nice.

Another example from that site is here - not only do they have the summary measures but some more detailed data as well. This shows the range of performance on various measures related to health care quality. I really like it because it highlights there is room for improvement in reducing these variations in quality.

Two other reports related to health system performance also offer numerous detailed summary measures and these include the Dartmouth Atlas of Health Care as well as a federal effort, the Hospital Compare produced by the [sic Centers for Medicare and Medicaid]. These are also highly interactive and allow comparisons, in a range of comparisons for a very large geographic area down to individual hospitals.

These next examples are reports that have an age group or life stage focus; there are many health indicator reports that focus on children and families. Kids Count is a dynamic online tool designed to search for information related to children and families and to offer a range of ranking tools ; it offers flexibility and ease of use. The next example is America's Children which is compiled by the Federal Interagency Forum on Child and Family Statistics - this is a static report but the forum in this publication represents an important leap forward in data sharing and compilation of federal statistics in children and youth. They started producing this report in 1997. Modeled after that report is Older Americans that is produced by the Interagency Forum on Aging Related Statistics.

Another major group of general health reports are those produced at the local level, they are used for planning and policy at the local level. The first example is the Seattle-King County Communities Count report ; I think it's one of the leading examples of using indicators in a very innovative way. There is an active collaboration on the part of local agencies (public and private) and it is part of a very integrated planning process that is responsive to local needs which are assessed via a periodic community survey.

Another good example is New York City. They produce indicators for 42 neighborhoods within five boroughs and these results are linked to local health assessment activities--surveys and health screenings--and are geared toward education. They use the indicator reports as an educational tool to encourage people to adopt healthier behaviors.

I mentioned the comprehensive systems with a focus on overall quality of life and that health is one of the many components that are featured. The State of the USA is based on this comprehensive model. The first example is the Boston Indicator Project led by the Boston Foundation also with the City of Boston and the Metro Area Planning Council. The second is the Jacksonville Community Council Inc. a long running quality of life indicator project. The Organization of Economic Cooperation and Development also has quality of life indicators.

Shown here is an example of the Boston Indicators Project. I want to show you the other components of quality of life -- and this is typical of the kinds of categories that you would see in these comprehensive indicator systems. The health indicators themselves are similar to those you would find in the general health indicators reports but presented in a different context of overall well-being.

One of the important things about indicators is that they can get overwhelming especially if you have too many and one way to combat this is the use of the index and this is a composite measure involving the use of several weighted measures. These are just a few examples of index efforts that measure overall well-being. The first is the Canadian Index - I will come back to this. The second is the Child Well-being Index which is similar and tracks the well-being of children relative to baseline years. The third example is an index based on a survey conducted with about 1,000 people every day - on approximately 350 days a year - and the focus is on perceptions of well-being and is also an interesting approach.

This example is from the Institute of Well-Being in Canada - the social indicators used in this effort was an attempt to bring together those many efforts on social indicators into one national index. The index is still in its developmental phase but they published this report on line. You can tell how new it is, it even has a typo -- and that was not our typo. They identify eight domains of well-being that contribute to the index. The intent is to have a single number that moves up and down like the Dow Jones Industrial index, but this one measures well being. You can see that the median index is the lowest darker line and the average is in the middle and this is compared to the gross domestic product shown on top. And indexes can be very useful however, they need to be carefully constructed. They can be insensitive to changes or meaningless unless compared with well understood benchmarks - and this is critical - this index is compared to GDP (gross domestic product) because that is something that we all understand. The second problem with indexes is that they can sometimes mask trends in smaller groups. In these examples I have shown are examples of better efforts - they actually explore what is going on in different subgroups and the trends behind the overall trends, a good practice. And although these are extremely valuable they are also limited in what they can tell you about how to improve health and where to target your energies.

In conclusion I want to leave you with the framework for the 20 or so indicators that were recommended by the Institute of Medicine Committee for the State of the USA indicators. I hope after today you'll have a good sense of what these might look like just by looking at these categories. I hope you understand why health indicators are so broadly used as well as useful, and I hope that you will be eager to participate in our forthcoming modules and hope you find them useful to your work.

If you are interested in the review, that report is available on my website at the link shown and you just need to click on the resources link to find that. If you are interested in learning more about the Institute of Medicine committee's recommendations to The State of the USA you can find the text of their report at the second link shown.

The National Library of Medicine (NLM) would like you to know about two valuable resources. The HSRR (Health Systems and Services Research Resources) is a searchable database of health datasets, survey instruments, and software. This is published by the National Information Center on Health Services Research and Healthcare Technology at NLM. The second is a tutorial on health statistics which covers many surveys that are the basis of health indicators.

I have also included some resources and further reading -- one resource is Mark Friedman and this is useful for understanding how indicators are used at the local level for planning and collaboration. The Government Accountability Office reference is a great history and compendium of comprehensive indicator systems used across the U.S. and internationally. It is a really important historical document regarding the creation of The State of the USA. The third link is the Robert Wood Johnson link that I discussed, called What Drives Health.

With that I'd like to thank you for your attention and I hope that you can attend future modules.