An article from University of Oslo
Health economic priorities are made on an unsound basis
Methods used in health economic evaluations are not always sufficiently valid.
Denne artikkelen er over ti år gammel og kan inneholde utdatert informasjon.
What would you prefer to avoid - high cholesterol or migraine? How serious do you think it is to get arthritis?
When health economists make priorities for the health service, their evaluations are in part based on the average population's attitudes to illness. However, researchers in the Institute of health and society at the University of Oslo have now shown significant weaknesses in this method.
"The available methods that are used on a large scale to measure health-related quality of life are quite weak, and they are reported with a sense of confidence that has no basis", says MD and researcher Liv Ariane Augestad.
Together with researcher and psychologist Kim Rand-Hendriksen, she has studied the methods that are most commonly used today to create the normative data to measure patients' health-related quality of life.
Rand-Hendriksen and Augestad's message is clear: Spend more on health economic evaluations. Health economic research is important and we need to have methods to quantify benefits in this area but treat the answers with a healthy pinch of salt.
And the two researchers warn against pretending that current methods provide a complete and true picture of reality.
"The analysis is not neutral. If an analysis is commissioned by a pharmaceutical company, they are likely to have selected the set of norms that provides the best outcome for their medicine. All pharmaceutical companies want their medicines to be cost-effective", Augestad says.
Making tough priorities
Like many other areas in life, the health sector has also been hit by tough priorities. Who should we prioritise in the treatment queue? Which medicines should be given precedence?
"To know how to best use the available funds, we must compare patients across different illnesses. All ailments and illnesses must be compared to each other, because this is the only way we can find the gradations", says Augestad.
However, measuring the aspect of quality of life that is health-related is complicated. The measurements must also be cheap. The effects of new medicines must be tested on several thousand people. If this cannot be done cheaply, it is not done at all.
The common solution is to use simple questionnaires in which patients are asked to describe some specific aspects of their own health, such as pain, anxiety, or mobility. This is fast, practical and cheap. However, the questionnaires can be completed in different ways. Setting comparable values of how good or bad these different ways are is challenging, and this is where the normative sets of data come in.
The price tag on health
The economists believe that to find out how bad an illness is, we should ask the healthy segment of the population whether they can imagine the illness.
This is similar to the methods used in market economics when trying to measure people’s willingness to pay for goods and services. How much is a product or service worth?
"Trying to assume and quantify a possible reduced health-related quality of life is a completely new exercise for people. You are used to keeping an eye on the petrol price, and you know how much a cinema ticket costs."
"However, you do not routinely consider how much it costs to get sick in the form of reduced quality of life and that you can buy your way out of it", says Augestad.
Changing the method in ways that in theory are insignificant turns out to greatly influence the responses - and thus also the priorities in the health service. Respondents are obviously affected by issues that are not really supposed to be significant. How the questions are formulated, the order in which they are asked and the respondents' attitudes to death all affect the answers.
Easily influenced
Additionally, respondents' limited knowledge of the topics that the questions focus on has an effect. As most people have little knowledge about what it would be like to get an illness they have never had, they are uncertain of how to answer.
"The more uncertain you are, the easier you are to influence", explains Rand-Hendriksen.
When you are asked to assess health, you have a poor basis for the assessment and you are therefore easily influenced by factors that you strictly speaking should not be influenced by.
This is where the problem lies.
"This means that the values we use today do not necessarily represent a true picture of how good or bad the patients' health situations are", says Rand-Hendriksen.
There are also significant differences between the illnesses or afflictions that the general population fear, and those that actually reduce quality of life. There is a gap between fear and reality. Thus, resources are spent on conditions that people fear instead of on what patients actually experience. This is unfortunate, both for the patients and for the economy, according to the researchers.
Reference:
Liv Ariane Augestad and Kim Rand-Hendriksen, Influence of construct-irrelevant factors and effects of methodological choices on EQ-5D health state valuation. PhD thesis, University of Oslo 2012-
