Source of values and preferences
To operationalize this question, we divided values and preferences into following levels based on their sources:
1. Utilities or patient values, which can be further divided into:
a. Utilities or patient values from different measurement techniques including standard gamble, time trade off, rating scales;
b. Multi-attribute utility, e.g. utility based on Health Utility Index (HUI);
c. Mapping results based on Health-Related Quality of Life Measurement.
2. Direct Choice
3. Health States Averseness Measurement.
4. Qualitative preferences.
Utilities or patient values measured based on instruments
Different techniques have been developed to measure patient ...view middle of the document...
Harvie et al reported one study using VAS to estimate the utility for women with urge, stress, and mixed urinary incontinence. The researchers used a 100-point vertically oriented VAS with anchors of 100 for “best imaginable health state” and 0 for “worst imaginable health state”. The utilities for urge, mixed and stress urinary incontinence were 0.78 ± 0.15, 0.78 ± 0.16 and 0.80 ± 0.14, respectively.
Limitation of methodology
The rating scale and VAS gain popularity because of their intuitive nature and simplicity during implementation. However, this methodology also received criticize for inherent drawbacks. Firstly, VAS neither compare different options nor deal with uncertainty, so this methodology does not conform to the utility theory. Secondly, the responses are dependent on the wording of end-point. For example, 0 on the scale may indicate “worst imaginable health state” or “death”, which leads to difference and confusion in comparison and interpretation. In addition, sometimes the VAS produces false precision. Jensen et al. (1994) found that with a 101-point VAS scale for pain (from 0 to 100), almost all of the respondents chose the numbers in multiples of 5 or 10.
The SG is based directly on the notion that people would make rational choice when they deal with uncertainty, that is, axioms of vNM utility theory (or expected utility theory). The standard gamble approach involves a trade-off between two alternatives: a health state that is certain and a scenario needing to gamble with one better (immediate full health with a probability of p) and one worse outcome possible (immediate death with a probability of 1-p). If the probability of reaching a better outcome through gamble makes respondents feel indifferent with staying in the described health state for certain, the probability will be translated into the patient values and preferences based on the equation that utility of certain health state = utility of better health*p + utility of worse health*(1-p). [11,16,17]
Grann et al. measured women's preference for breast cancer using standard gamble. In this survey, respondents were asked to choose between A and B, where B is of 100% probability of occuring, but A involved taking a risk (two alternative outcomes with two different probabilities of occurring). The utilities of prophylactic drug usage were 0.61 for the general population group to 0.74 for the breast cancer group.
Limitation of methodology
The SG is based directly on the expected utility theory. The major criticize on this methodology is the complexity. It is argued that few people are capable to understand or accustomed to deal with probabilities.
Time Trade Off
The difficulty of dealing with uncertainty led to the development of time trade off in 1970s. The TTO approach involves asking subjects to outweigh the amounts of time they would like to trade-off between health improvements and...