This posting reviews some of the underpinnings of the health level, health behaviors the demand for health care, and health care spending in the economy.
1. What is Health and how it is Measured
WHO defines health as: “Health is a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity”.
This is pretty broad, and certainly encompasses more than simply the presence or absence of acquired diseases. Also included in “health” are aspects of stress associated with poverty, possibly isolated living circumstances, and other situations facing millions of persons in the world. These concerns range from low economic development, side effects of political upheaval and war, natural disasters, and to social circumstances that prevent fairness and cause stress. These are worthy concerns, and contribute to the inability of people to function at optimal levels. Health is certainly a complex concept. Simplifying, it a bit it becomes a definition for Health like:
- the absence of disease or injury, both physical and mental
- full capacity to perform usual daily activities (work, school, child rearing, family caregiving, etc.)
- longevity (remain as free as possible from causes of premature death)
There are a number of ways that “health” or health improvements are measured in health economics work. There is no gold standard. The main measures are:
- surveys of degree of “happiness” in populations (this is an interesting literature)
- surveys self assessed health using questions like: “relative to others like you, would you say your health is excellent, better than average, average, worse than average, or poor”
- surveys of Restricted Activity Days in the last month or so: –days lost from usual activities due to poor health (in the past month or so) — days not able to attend school, or not able to go to work, or do whatever are usual activities
- mortality rates of various kinds (age groups and cohorts, disease specific)
- infant mortality rate (failure to survive to 1 year / births) — generally perceived as a more sensitive indicator of mortality to health system limitations (poor access, inadequate services)
- Disability Adjusted Life Years Lost Due to preventable disease and disability — DALYs— a commonly used metric to measure the burden poor health by country by year (the Institute for Health Metrics and Evaluation, IHME does almost all this work) which sums together (1) an estimate the years of life lost due to preventable causes, and (2) an estimate of years lost with bad quality of life due to preventable causes. Longevity in Japan is a reference population used to estimate maximum years of life possible.
For measuring health as it may relate to particular types of health interventions, or components of the health system other kinds of measures are used:
- survival rates for particular types of persons (with/without the intervention, etc.)
- incidence or prevalence rates for particular diseases, or risks
- Quality Adjusted Life Years— QALYs—- a common measure for capturing the reduction in disease burden (eg health benefits) of an intervention (new vaccine or drug, a diet intervention, improved insurance coverage, etc.) — estimated as the sum of (1) the years of extended life created, and (2) the quality of life improvements and their duration. One QALY is one year in perfect health.
- surveys of changes in self assessed quality of health or restricted activity days
2. How does the U.S. Health System Compare in terms of Health?
As an indicator of U.S. health care perfromance, the chart below shows how countries compare on Infant mortality rates. Three facts are evident: (1) The U.S. doesnt do as well as most developed countries on this measure of health. The reason for this stems from the heterogeneity in health (and many other things) in the U.S., as compared to other developed countries—where incomes, health and other measures end to not vary as much from household to household. But in the U.S. looking more closely (the right side of the chart), there is a huge variation in infant mortality within the U.S. (2) In some places, infant mortality is very comparable to the world leaders or even better. But, in other places in the U.S., (3) we have infant mortality rates are so bad they look like “third world places” (in the Mississippi delta area, and other places in the rural south). This situation stems from lack of access to providers in such places because nobody wants to practice there because of poverty and lack of insurance.
So in the U.S. some persons tend to have very good health because they have access to the world’s best doctors, facilities, technologies, and modern mirtacles of science. But, our average health levels in the U.S. are not so special, since theye are other people here who do not have nearby access to that kind of care, or if they do live in proximity, dont have good enough insurance of other means of financing to be able to use such care. And, as we review below, many people in the U.S. do not live healthy lifestyles and make choices that would make population health better.
- Stepping Back: What Drives the level of Health in Populations
One large study by WHO looking at 120 countries over 30 years showed that variation in the levels of mortality over time and across countries were driven mainly by three things: new drugs and other new diagnostic and treatment technologies (explained 30% of the variation over time and across countrfies in mortality rates), income differences (explained 20% of the variation in health), and level of formal education of females (50%). The latter is probably due to the importance, across cultures, of women in making health choices for families.
A second type of study has looked at the causes of mortality, and assessed the extent to which death might have been preventable. The results of these kinds of studies are typified in the chart below:
Here we see that the drivers of variations in population health are genetics ( about 20% of the variation in preventable mortality can be traced to genetic differences), medical care (10-20%), lifestyle choices like diet, risk behaviors, exercise habits (50%), and environment like water and air quality, noise, etc.(10-20%). Of course, we spend the vast majority of our spending on the least influential factor in this set of health drivers. This isnt to say that health services, hospitals, and technologies don’t produce health, since they do often extend life and improve quality of life—- it only says that other factors are critical drivers in the health of populations. Choices about diet, risky behaviors, and other matters of personal choice are much more crucial in producing health. This fact is generally lost on health professionals, who tend to see “health” as driven by visits, tests, treatment and compliance. But, when the topic of prevention comes up, professionals know they have to “engage” patients, possibly “nudging” them and educating them.
It is well know that Education and literacy are important drivers of health in populations and the related choices people make. Research clearly demonstrates this fact–that investments society makes in education have huge “health ” payoffs too.
“An additional four years of education lowers five-year mortality by 1.8 percentage points; it also reduces the risk of heart disease by 2.16 percentage points, and the risk of diabetes by 1.3 percentage points.” (David Cutler and Lleras-Muney NBER 2006 http://www.nber.org/papers/w12352)
Other simple displays of data by Cultler and Lleras-Muney confirm this pattern:
- Demand for Health Care
Why do persons seek professional services? The “demand” for health care is derivative from the demand for health. Obviously, we value health. It allows us to feel good, and not be restricted in participating in the activities we enjoy. People are, of course different in terms of what they enjoy. And, therefore, are different in their “target” health level. Someone that enjoys vigorous activities may need to be far more healthy to enjoy them than w`ould be someone with more modest requirements for fitness and energy level.
People also differ in terms of the importance and concern they feel for the future. Persons who have strong preferences for their future, and what they’ll be doing, and how they’ll be doing it– will choose to make investments in their future level of health (eg engage in prevention behaviors). On the other hand, persons who only value the moment are not likely to make the same investments in prevention. Here again, people are different, they behavior in accord with their life preferences. Some value health differently, and some value investments in future health differently. We may wish we were all the same— but we are not.
Our demand for health care varies too, in accord with our perceived value of health. Some persons go to their PCP annually, are prompt about checking out changes in our health, and dont procrastinate about emergent problems. Others may go years without a visit, and delay followup against medical advice. Some of these differences in the way people use the health system are the result of different “target level” of current and future health. Other sources of difference are things like income differences, availability of insurance, cultural and peer group health behaviors, differences in geographic convenience levels, and education levels. These are also “drivers” of our demand for health care.
Overall, health care is disproportionately important in our society as an intervention to produce/remedy health situations. We spend about $10,000 per person per year in our society, largely to treat diseases and conditions. This is a huge expenditure (about 18%) of the economy —- and is about 30-50% more than the health spending in other developed countries.
The following chart tells us more about the trends in the American health system and its production of health. It says we don’t get much “Bang for the Buck” (outputs per input) out of the American Health system. This relationship also suggests the ever flattening shape of the U.S. “production function” for Health — something often characterized as “flat of the curve” health care in America. Take a look at this chart. What is making the trajectory of the U.S. health system so different? What might have caused the changes in 1980 that seems to have altered the U.S. trajectory?
This alarming chart provokes the question of the U.S. health care system, and what’s so different about it from the “systems” in other countries. Its not a shortage of spending! Or inadequate investments in scientific drug development, or the quality of training for of doctors, nurses, and other clinicians. Why does our “system” choose to spend so much on health care, but often achieve far less health for the people than is being produced in other countries?
What is a Health System? Well, its the way society organizes, allocates and manages the health care resources of society toward society’s objectives in terms of health status, efficiency, fair access, and satisfaction levels of the people. The slide below on the left panel shows the types of resources (including the scientific backbone) all health systems need to produce health, and how they are linked by the organization or system of organizing, allocating, delivering, managing, monitoring, improving and leading to get it done. This is a generic picture of all health systems, not just ours. It follows the WHO picture of generic health system functions. One important feature of health systems is the “feedback loop” shown on the left panel of the chart. Here, we expect that the health system has a mechanism whereby inadequacies (eg unfairness, or excessive spending) are recognized and somebody takes action to create a remedy.
One way to depict our system is on the right side of this figure. But it is sort of generic too. The blue pyramid suggests that our system is based on the base of HOUSEHOLD CHOICE about how much they know, and how much they care about health in their lives. After all, we are a more of less free society. Then the next key building block of the blue pyramid is Public Health Provision. Basically think about this as more or less universally available and free services (paid by taxes, not based on service consumption) that are aimed at the health of the public. Economists call these “public goods” the health of everyone depends on these things: things like Food and drug safety, water quality and clean air regulations, laws that require government licencing of professionals and hospitals and labs, and maybe free childhood vaccinations at public health departments. The next levels of the blue pyramid are primary care, then secondary (inpatient) care, and then tertiary specialty and medical specialty care. The pyramid shape illustrates the epidemiology of usage of these levels of care: everyone’s health is based on their household values and needs, almost everyone will benefit from whatever public health services are provided, and the majority will certainly need and use primary care, but fewer will need secondary services, and far fewer will need tertiary care. That is a pretty generic functional description of the health care services delivery structure in our health care system. Of course the blue boxes on the top are support services and products in our delivery structure. Note, the green pyramid below describes how the money in our health system is allocated (by the way allocations work in the ‘system’).
Lets get more specific about the structure of the American system—– look at the next slide:
The U.S. Health Care system isnt a organized “system” at all. It is organized like the market for cars, or cell phone plans, or the sale of Christmas trees. Yes, like any of these other markets there are suppliers (who can come and go, and are free to fail) consumers who are free to choose, supply chain firms that provider resources to the sellers, and there are some government rules to follow. Yes, these “market systems” for Christmas trees in December do have buyers, sellers, supply chains, etc, and they sure do deliver successfully in most cases). Markets do work pretty well for delivering products and services of all types in America, and many other countries. To say that ours is a health “system” implies some things that are not true:
1. Almost all providers of services and products are private firms or private practices. This is true of all not for profit organizations (which only means that they are approved for not paying any taxes, and those who donote money to them are allowed a tax break for their donation)—- Consumers & Private organizations can do whatever they want. Here today, gone tomorrow. They are not part of any health system plan for primary care, for secondary care, or drug development—- there are no such plans. Patients are, for the most part, free to choose providers, and self refer to whoever they want. Medicare provides this right explicitly, as do most private insurance— unless it is waived in a “managed care plan”.
2. We have what is known as a mixed health system with minimal government interference–mainly private providers, pretty much free choice for providers, insurers, and patients— a market system with minimal government oversight, and only a few situation where government (state, federal, county) provide services— though nearly 45% of the insurance payments for health care are paid by governments through social insurance plans (Medicare, Medicare, TriCare). (see the chart)
3. there is almost no coordination between providers—as the right side of the slide suggests– there is extreme fragmentation in the marketplace of health care service providers, all types, all sizes, shapes, styles of care, staffing patterns, services offered, everything is fragmented (except within managed care organizations, where they explicitly try to integrate services better to be more economical) —-providers can offer (or not offer) services of many types wherever they want, within their scope of licence. They can go out of business for whatever reason. Resource suppliers (schools of medicine and nursing, EMR vendors, CT /MRI vendors, drug company suppliers, etc.) and private insurors, are also private firms, and can do whatever they want–no coordination, no plans. Sure there are limits and regulations to be followed in order to operate. But no planning, no limits on product and services to sell, nor ANY price regulations. There is No system! Unlimited and Unorganized provider options subject to choice & self referral
4. there is, most astonishingly, nobody in charge of the U.S. health system. No person, no organization, no plan, no explicit goals, and certainly not even the Federal Government (the constitution deligates some “public health functions” authority to States, none to the Feds). This is partly the issue that allows courts to reject parts of the ACA and other progressive (tax funded) health legislation. The limited role of government, the role of others, and the situation in other health systems are shown in the next three slides.
In summary, see the key words to describe the health System here:
5. Household Health Production
The next few sections here offer some background of the concept of “health production” how households make choices about their health, as if it is some sort of investment in future happiness.
The economic view of the importance of the “choices” people make about their preferred level of health is emphasized in the “theory of household health production.
This theory suggests that people (households) combine purchased goods and services with their own time and know-how to produce activities that are valued. (Games of tennis, vacations, restaurant meals, raising children, etc). Households “produce” these activities to reflect their preferences, and their economic constraint like time, money and know how.
The chart shows the “production function” for producing health. More inputs (horizontal axis) generate more health (vertical axis). For a given type of person, (where ‘know how’, age, genetics, and education are constant) additional health can be produced by expending more time or money on inputs—- causing a movement up the curve. The level of health is also subject to shifts in the production function. If, for example, persons purchase more education, their production function will shift up (they have more information, or are more efficient at searching for relevant information). As persons age, then the production function shifts down (as illustrated). Economists study the shape of production functions, the sensitivity of health to variations inputs, and shifts.
But, the more health we try to make by purchasing services and expending personal time, the more are going to be required to a specific increment of health (this is called ‘diminishing returns’). The lower and lower productivity we see as we try to produce more and more, is the result of a “fixed supply” of know how. If we invested more in know how (eg education or experience) the entire production function would shift up. This means that it is possible to make more health with the same level of inputs.
The theory predicts that in choosing to produce health (games of tennis, a vacation, etc) people will differ in terms of “how” they produce. People with lots of money and scarce time will choose to use more good and less time to produce their activities. People with more time than money will choose a different combination of inputs.
Marketers take advantage of these differences. Consierge medicine is “convenient” (less time) and costs more money. Some insurance plans have more options for access (convenience). All market segmentation is based on differential demands among groups of customers Often this boils down to different scarcities of time and money.
When “know how” is higher, it means that the full price of producing a particular level of health is lower. What this will mean, in turn, is that a lower full price will tend to increase the demand for the activity of producing health. (or tennis, or restaurant meals, etc.)—- relative to the demand for other activities we might have spent our time and money to produce. So, persons with more “know how to produce health in the household” will demand higher target levels of health.
Will this means that persons with more know how will demand more medical care? not necessarily. For some persons, yes, for other persons not so much–depending on the relative prices of time and the scarcity of money. Insurance comes in here.
- More on Health Production Functions
Health production functions are economic tools used to estimate the influence of population characteristics on health. How effective are various inputs to produce health? How much does the production function shift as a result of more education, or age? These multivariate relationships are estimated from individual or country data using regression analysis. Regression analysis estimates a quantitative relationship between two variables (eg Health and Education). It standardizes that two-variable relationship for the influence of other factors (age, race). A statistical test is used to test if the association between the 2 variables is larger than would be likely given the variation seen in the data. Regression is used to understand the size and significance of the drivers of health status, using data on thousands of individuals . The following chart describes the data as a scatter diagram and the “best fitting straight regression line” that is estimated from the data—- and the slope of the line being the relationship between the dependent variable (health) and a one unit change in the independent variable (eg a, or b).
The results of two research projects to estimate production functions for health are shown below. The left side shows production functions based on a number of health measures, where the data is taken from countries around the world. The study on the right measures health by a Health Interview Survey of self reported health by Americans (on a scale of low to high):
The cross country study on the left of the chart reports on several measures of health. It shows that life expectancy gets longer when literacy rates in the country are higher, when health spending is higher, and when calorie intact is lower. The results are consistent with the three models that use mortality rates as health measures.
The second panel here (from my dissertation) focuses on the several drivers of health (as measured by persons own self assessment vis other people they know. The ideas is look at the “health effects of things like education level, age, # cigarettes smoked per day, whether you’re married. It separates the study sample into three groups according to their education levels (<8 years, 9-12, 12+). For each groups three different statistical techniques are tried— but for this purpose just look at the Model 1 results for each segment. Focus on the last row— which shows the coefficient associate with age (actually it is the logarithm of age). That coefficient shows what happens to health for every extra year of life. It is negative–indicating the every extra year subtracts from the self assessed health level, all other things the same. But compare the model 1 for the three panels—- the size of the age coefficient gets smaller as the years of education increases.This can be interpreted as more education makes people healthier, observed here as slowing the rate at which their health seems to be deteriorating with age. Note also, that higher education is also reducing the negative impacts on health associated with how many cigarettes the person smokes. Education is almost like the fountain of youth!
- Flat of the Curve Medicine
This concept is often used to describe the U.S. health system—– it means that when looking at the production function for health in the U.S. it looks “flat” , which means that addind more inputs (money, resources) into the system will not proudce that much additional health.
A multi country health production function is shown below in graphical form. The Health measure is life expectancy at birth. The horizontal axis is per capita spending on health care (a summative measure of all the purchased resources consumed in the production of health). What we see here is the extraordinary performance of the U.S. health economy— very very high levels of spending relative to other countries, with not proportionately high Health levels. This combination is the central defining feature of the U.S. health economy. Variously interpreted: pretty low output given the high resource inputs; little bang for the buck, excessive spending to achieve modest health levels. Waste, inefficiency, ineffectiveness, and slack are all possible descriptors of this flattening phenomina—- health is going up at a declining rate as we spend additional increments of money on the health system.
The production function here (the line) is upsloping, meaning that adding resources will increase health. But the data across countries shows that as more and more resources are added, the increases in health status is not growing proportionately— the increment in health status per increment in resources used gets smaller and smaller as we add more resources. This is called the principle of “diminishing returns” in production — as we add more resources to a production process we fail to achieve proportional returns in output.
For the U.S., the production functio for health that gies throughh the U.S. data point is far to the right–and far flatter, than the health production function facing the other countries. It says that as we do more (more tests, procedures, visits, etc.) we can augment health, but the productivity of the added resource use gets smaller and smaller (the production function gets flatter and flatter). So, we speak of “flat of the curve” medicine in the U.S. because we have dedicated so many resources to our health care system, that additional resources yield very little.
Why is the American health system have this ‘Flat of the curve’ problem? Why is our spending level so high that the contribution of incremental spending to health is so low? Why are we so different in this unattractive way?
Possible root causes are many:
- fragmented financing system — lots of insurers, high transaction costs, lots of overhead – administrative spending (by insurers and by provider organizations to bill and collect payment)
- dedication of the vast majority of health spending on treating serious illness, and development of a significant flow of innovative approaches to support treatment for these problems (eg spending huge amounts on curing disease, not preventing it)
- Fee for service payment — incentives for providers to seek to provide more services for their patients in order to increase their income. Payment incentives that might shift risk of “poor health results” to providers, such as capitation, remain small fractions of the the provider payments to medical care providers.
- insurance — lowering the “price” the patient must pay to consult or be treated by the provider. Recent attempts to employ Pay for performance payment incentives by Medicare, the AQC in MassBCBS, and others remain a very tiny fraction of provider payments
- But, much of the blame for “flatness” must rest on the patient’s behavior in our culture — where people do not want to take responsibility for their own health—-they want the health system and technology (drugs devices and medical procedures) to do it for them. Medical professionals are viewed like a mechanic, who takes care of your car, and keeps in running. In general, Americans are not accountable for primary prevention drivers of chronic disease. Health is largely controlled by household lifestyle choices: the risks we create by how we eat, what we smoke, what risks we choose to take, etc.
6. There is also an important contribution to “flatness” that is evident in the underusing persons in our society; the millions without insurance– and the millions who dont live in proximity to providers. This is a problem of
This chart makes this point– that we are spending more and more on fewer and fewer people in the health care system.
professorgaumer 