A. General Cost-Effectiveness Analysis Issues
6. How do I measure quality adjusted life years?
To compare the value of one treatment with another, you need an outcome measure that works across different health states. This rules out disease-specific quality of life measures as they only reflect a specific disease or illness.
In addition, the outcome measure must also work as a measure of preference. Preference measures not only measure health status, but they also measure how the individual values the current health state. The valuation of the current health state usually covers the spectrum from full health to death.
To date, the quality adjusted life year (QALY) model is the preferred metric for estimating the health effects (Gold et al., 1996). QALYs are estimated by multiplying each life year gained with an intervention by a quality-weighting factor that reflects the individual's quality of life in the health state for that year. Utilities, measured on a scale from zero (death or the worst health imaginable) to one (perfect health), can be used as quality-weighting factors. See Gold et al., 1996 for details.
There are different ways to find quality weights. The easiest is to use published reports and league tables. Besides Medline, a great resource for this is the Harvard Center for Risk Analysis.
If existing utility weights do not meet your needs, you may need to collect weights. In doing so, sampling issues should be considered carefully (see Gold et al., 1996). The estimation of the quality weights for a given period (i) and treatment (k) requires the successful completion of two tasks:
- Measuring the impact of an intervention on the distribution of health states. This task requires that the health states influenced by the treatment are completely characterized.
- Assessing the preferences (utilities) for these alternative health states.
This two-step estimation process can be done with different methods.
- Rating Scales
- Standard gamble (SG)
- Time tradeoff (TTO)
A key distinction between the methods is the how they handle risk. The Standard Gamble makes the respondent consider the risk of death. The Time Tradeoff method asks the person to consider a tradeoff with years of life. Some say that the TTO is cognitively easier to understand than SG, although the jury is still out on this. Usually the SG and TTO methods require interview administration, although a lot work is being done to use computer and Internet administration. Given the logisitical complexities, many people turn to the rating scales. The rating scales, however, do not require people to consider risk. There is a controversy over the importance of including risk (See Gold et al, 1996, p 118).
It is also important to note that these methods usually yield different utility weights. This has led some people to use multiple methods.
For more information, see the HERC Cyberseminar by Peter Ubel or visit his center's website.
To ensure the protection of VA information, we have restricted access to certain files (denoted by the"lock" icon). Contact HERC to download these restricted files.
- Introduction to Patient Preferences/Utility Assessment
Date: 8/14/2003 | Size: 606Kb