SUMMARY BEHAVIOURAL DECISION
THEORY IN HEALTH
CONTENT
Lecture 1: Introduction ................................................................................................................................ 3
Weinstein et al. (2009) ................................................................................................................................ 3
Lecture 2: theoretical proporties of QALYs ............................................................................................... 4
Economic Evaluations (EE) ........................................................................................................................ 4
Value of life................................................................................................................................................. 4
Ryen & Svensson (2014) Meta-Analysis .................................................................................................... 4
Deriving the QALY model .......................................................................................................................... 4
Lecture 3: Empirical evidence on QALYs .................................................................................................. 7
How to test risk neutrality ........................................................................................................................... 7
How to test utility independence/standard gamble invariance .................................................................... 7
Bleichrodt & Pinto (2005)....................................................................................................................... 7
How to test constant proportional trade-offs (CPT) .................................................................................... 8
Maximal endurable time (MET).................................................................................................................. 8
Variable health ............................................................................................................................................ 8
Preference-, additive- and utility independence ...................................................................................... 9
Lecture 4: discounting the future............................................................................................................... 11
Relevance and examples ........................................................................................................................... 11
Reasons for discounting ............................................................................................................................ 11
Definition discounting ............................................................................................................................... 11
Theory ....................................................................................................................................................... 12
Issues in discounting literature .................................................................................................................. 12
Constant discounting ................................................................................................................................. 13
Hyperbolic discounting ............................................................................................................................. 13
Violations of classical theory .................................................................................................................... 14
Lecture 5: Applications of discounting ...................................................................................................... 16
Discounting of costs vs. benefits ............................................................................................................... 16
Role of discounting in utility of life duration ............................................................................................ 16
Intertemporal Choice and examples of quasi-hyperbolic discounting in every-day decision tasks .......... 17
,Lecture 6: Empirical evidence on QALYs, experienced utility ............................................................... 20
Different forms of utility ........................................................................................................................... 20
Examples in the literature .......................................................................................................................... 21
Lecture 7: Preference Reversals ................................................................................................................ 23
Definition and empirical evidence ............................................................................................................ 23
Neumann-Böhme, Lipman, Brouwer & Attema, (2021) ........................................................................... 24
Lecture 8 & 9: Biases in health utility measurement ............................................................................... 25
Recap ......................................................................................................................................................... 25
Measuring utility L(T)............................................................................................................................... 25
Empirical evidence on violations .............................................................................................................. 26
Reactions to the inconsistencies ................................................................................................................ 27
Debiasing – Prospect Theory .................................................................................................................... 27
Lecture 10 & 11: Which method is best for health state valuation? ....................................................... 29
Reasons for violations of EU..................................................................................................................... 30
Lecture 12: Equity weighting & QALYs ................................................................................................... 32
Aggregating QALYs ................................................................................................................................. 32
Combining equity and efficiency concerns ............................................................................................... 33
Equity-efficiency trade-off ..................................................................................................................... 33
Rank-Dependent QALY model ................................................................................................................. 34
Lecture 13: From theory to practice – correct(ing) methods for health state valuation ....................... 36
New developments in health utility measurement..................................................................................... 37
Discrete choice experiments (Bansback et al., 2021) ............................................................................ 37
Individual corrections for bias (Lipman et al., 2019) ........................................................................... 37
,LECTURE 1: INTRODUCTION
No notes needed.
Weinstein et al. (2009)
The conventional QALY is a valuation of health benefit. Health is defined as the value-weighted time—life-
years weighted by their quality—accumulated over the relevant time horizon to yield QALYs. To permit
aggregation of QALY changes, the value scale should have interval scale properties. Value is equated with
preference or desirability. You can measure individual preferences, preferences for a small set of health
domains or attributes, or in terms of people’s preferences about the health of the community.
Matrix characterizing uses and definitions of QALYs:
Underlying assumptions of the conventional QALY approach:
1. A resource-allocation decision must be made.
2. The outcomes of the alternatives can be specified in terms of health states, changes, and durations.
3. Resources are limited, and each alternative has resource implications (costs).
4. A major objective of the decision-maker is to maximize health of the population, subject to resource
constraints.
5. Health is defined as value-weighted time (QALYs) over the relevant time horizon.
6. Value is measured in terms of preference (desirability).
7. Each individual is risk neutral with respect to longevity and has utility that is additive across time.
8. Value scores (preferences) measured across individuals can be aggregated and used for the group.
9. QALYs can be aggregated across individuals; i.e., a QALY is a QALY regardless of who gains/loses
it.
There’s a discussion about whether QALYs should be discounted. Argument against: if the risk neutrality
assumption holds, QALYs are valued equally over time. Argument pro: QALYs are driven largely by
opportunity costs.
Equity and fairness, need to be considered separately in the conventional QALY approach because
conventional QALYs are not intended to incorporate all concerns of decision-makers.
, LECTURE 2: THEORETICAL PROPORTIES OF QALYS
Economic Evaluations (EE)
An EE compares costs and effects. The main question is: “How can we express the benefits of health care
numerically?”. Different approaches:
• Cost minimization: only look at costs and ignore effects.
• Express benefits as life-years gained, if you then adjust for quality, you get QALYs.
Value of life
When calculating the value of life, you encounter the following problems:
1. Sensitivity to irrelevant information.
− Starting point bias: the number that is first shown to a respondent determines their answer.
− Range effect: the answering range given can influence the respondent’s answer.
− Disparity between WTP/WTA: how much you are willing to pay somebody else will be
lower than how much you are willing to accept for something you have and are giving away.
2. Insensitivity to relevant information.
− Scope effects: the value of life depends on the amount (of reduction) considered whereas it
should be the same.
Ryen & Svensson (2014) Meta-Analysis
Mean WTP-Q (= willingness to pay for a QALY) is 118,839 EUR while the median is 24,226 EUR → so
it’s heavily skewed. They conclude that stated preference studies give higher estimates that VSL (value of
a statistical life) studies. They also found higher values when the risk of death is included than when pure
quality of life changes are being valued. Finally, the found that WTP-Q is not consistent across QALY
changes. Larger changes give lower WTP-Q estimates (same as the insensitivity to scope).
Deriving the QALY model
The most common used model is the Quality-Adjusted-Life-Years (QALYs). It’s an additive model where
a health profile is expressed as (q1 … qt) = quality of life in each time period. According to the QALY model,
the utility of this health profile U(q1 … qt) = H (qt) = the summation of the utility H of health state q in each
period. For a chronic health state, q is the same in all health periods, H(Q)*T = the utility is multiplied by
this chronic health state times the duration.
In a decision tree, you can multiply the years in the health state by utility H.
Advantages of the QALY model are that multiply the number of years with the quality of life as a fraction
between 0 and 1, is intuitively appealing and easy to use in practice. As a disadvantage, it may be too simple.
The main purpose of the model is to represent preferences. For example if X > Y, then V(X) > V(Y).
THEORY IN HEALTH
CONTENT
Lecture 1: Introduction ................................................................................................................................ 3
Weinstein et al. (2009) ................................................................................................................................ 3
Lecture 2: theoretical proporties of QALYs ............................................................................................... 4
Economic Evaluations (EE) ........................................................................................................................ 4
Value of life................................................................................................................................................. 4
Ryen & Svensson (2014) Meta-Analysis .................................................................................................... 4
Deriving the QALY model .......................................................................................................................... 4
Lecture 3: Empirical evidence on QALYs .................................................................................................. 7
How to test risk neutrality ........................................................................................................................... 7
How to test utility independence/standard gamble invariance .................................................................... 7
Bleichrodt & Pinto (2005)....................................................................................................................... 7
How to test constant proportional trade-offs (CPT) .................................................................................... 8
Maximal endurable time (MET).................................................................................................................. 8
Variable health ............................................................................................................................................ 8
Preference-, additive- and utility independence ...................................................................................... 9
Lecture 4: discounting the future............................................................................................................... 11
Relevance and examples ........................................................................................................................... 11
Reasons for discounting ............................................................................................................................ 11
Definition discounting ............................................................................................................................... 11
Theory ....................................................................................................................................................... 12
Issues in discounting literature .................................................................................................................. 12
Constant discounting ................................................................................................................................. 13
Hyperbolic discounting ............................................................................................................................. 13
Violations of classical theory .................................................................................................................... 14
Lecture 5: Applications of discounting ...................................................................................................... 16
Discounting of costs vs. benefits ............................................................................................................... 16
Role of discounting in utility of life duration ............................................................................................ 16
Intertemporal Choice and examples of quasi-hyperbolic discounting in every-day decision tasks .......... 17
,Lecture 6: Empirical evidence on QALYs, experienced utility ............................................................... 20
Different forms of utility ........................................................................................................................... 20
Examples in the literature .......................................................................................................................... 21
Lecture 7: Preference Reversals ................................................................................................................ 23
Definition and empirical evidence ............................................................................................................ 23
Neumann-Böhme, Lipman, Brouwer & Attema, (2021) ........................................................................... 24
Lecture 8 & 9: Biases in health utility measurement ............................................................................... 25
Recap ......................................................................................................................................................... 25
Measuring utility L(T)............................................................................................................................... 25
Empirical evidence on violations .............................................................................................................. 26
Reactions to the inconsistencies ................................................................................................................ 27
Debiasing – Prospect Theory .................................................................................................................... 27
Lecture 10 & 11: Which method is best for health state valuation? ....................................................... 29
Reasons for violations of EU..................................................................................................................... 30
Lecture 12: Equity weighting & QALYs ................................................................................................... 32
Aggregating QALYs ................................................................................................................................. 32
Combining equity and efficiency concerns ............................................................................................... 33
Equity-efficiency trade-off ..................................................................................................................... 33
Rank-Dependent QALY model ................................................................................................................. 34
Lecture 13: From theory to practice – correct(ing) methods for health state valuation ....................... 36
New developments in health utility measurement..................................................................................... 37
Discrete choice experiments (Bansback et al., 2021) ............................................................................ 37
Individual corrections for bias (Lipman et al., 2019) ........................................................................... 37
,LECTURE 1: INTRODUCTION
No notes needed.
Weinstein et al. (2009)
The conventional QALY is a valuation of health benefit. Health is defined as the value-weighted time—life-
years weighted by their quality—accumulated over the relevant time horizon to yield QALYs. To permit
aggregation of QALY changes, the value scale should have interval scale properties. Value is equated with
preference or desirability. You can measure individual preferences, preferences for a small set of health
domains or attributes, or in terms of people’s preferences about the health of the community.
Matrix characterizing uses and definitions of QALYs:
Underlying assumptions of the conventional QALY approach:
1. A resource-allocation decision must be made.
2. The outcomes of the alternatives can be specified in terms of health states, changes, and durations.
3. Resources are limited, and each alternative has resource implications (costs).
4. A major objective of the decision-maker is to maximize health of the population, subject to resource
constraints.
5. Health is defined as value-weighted time (QALYs) over the relevant time horizon.
6. Value is measured in terms of preference (desirability).
7. Each individual is risk neutral with respect to longevity and has utility that is additive across time.
8. Value scores (preferences) measured across individuals can be aggregated and used for the group.
9. QALYs can be aggregated across individuals; i.e., a QALY is a QALY regardless of who gains/loses
it.
There’s a discussion about whether QALYs should be discounted. Argument against: if the risk neutrality
assumption holds, QALYs are valued equally over time. Argument pro: QALYs are driven largely by
opportunity costs.
Equity and fairness, need to be considered separately in the conventional QALY approach because
conventional QALYs are not intended to incorporate all concerns of decision-makers.
, LECTURE 2: THEORETICAL PROPORTIES OF QALYS
Economic Evaluations (EE)
An EE compares costs and effects. The main question is: “How can we express the benefits of health care
numerically?”. Different approaches:
• Cost minimization: only look at costs and ignore effects.
• Express benefits as life-years gained, if you then adjust for quality, you get QALYs.
Value of life
When calculating the value of life, you encounter the following problems:
1. Sensitivity to irrelevant information.
− Starting point bias: the number that is first shown to a respondent determines their answer.
− Range effect: the answering range given can influence the respondent’s answer.
− Disparity between WTP/WTA: how much you are willing to pay somebody else will be
lower than how much you are willing to accept for something you have and are giving away.
2. Insensitivity to relevant information.
− Scope effects: the value of life depends on the amount (of reduction) considered whereas it
should be the same.
Ryen & Svensson (2014) Meta-Analysis
Mean WTP-Q (= willingness to pay for a QALY) is 118,839 EUR while the median is 24,226 EUR → so
it’s heavily skewed. They conclude that stated preference studies give higher estimates that VSL (value of
a statistical life) studies. They also found higher values when the risk of death is included than when pure
quality of life changes are being valued. Finally, the found that WTP-Q is not consistent across QALY
changes. Larger changes give lower WTP-Q estimates (same as the insensitivity to scope).
Deriving the QALY model
The most common used model is the Quality-Adjusted-Life-Years (QALYs). It’s an additive model where
a health profile is expressed as (q1 … qt) = quality of life in each time period. According to the QALY model,
the utility of this health profile U(q1 … qt) = H (qt) = the summation of the utility H of health state q in each
period. For a chronic health state, q is the same in all health periods, H(Q)*T = the utility is multiplied by
this chronic health state times the duration.
In a decision tree, you can multiply the years in the health state by utility H.
Advantages of the QALY model are that multiply the number of years with the quality of life as a fraction
between 0 and 1, is intuitively appealing and easy to use in practice. As a disadvantage, it may be too simple.
The main purpose of the model is to represent preferences. For example if X > Y, then V(X) > V(Y).
