Only Cost Utility Analyses Include Quality of Life as an Outcome

July 1, 1995

LONG BEACH, Calif--Of the different types of economic analyses used in cost studies of medical therapies, only the cost utility analysis takes into account quality of life as an outcome, Jane Weeks, MD, said at the 1995 Quality of Life Symposium, sponsored by St. Mary Medical Center. To understand how this type of analysis works, physicians must add two new measurement units to their vocabulary: "utilities" and "quality-adjusted life years (QALYs)," she said.

LONG BEACH, Calif--Of the different types of economic analysesused in cost studies of medical therapies, only the cost utilityanalysis takes into account quality of life as an outcome, JaneWeeks, MD, said at the 1995 Quality of Life Symposium, sponsoredby St. Mary Medical Center. To understand how this type of analysisworks, physicians must add two new measurement units to theirvocabulary: "utilities" and "quality-adjusted lifeyears (QALYs)," she said.

Rather than using dollars or years of life saved as a measurementunit, cost utility analyses use QALYs, determined by plottingquality of life versus length of survival, said Dr. Weeks, assistantprofessor of medicine, Harvard Medical School, Dana-Farber CancerInstitute.

To use quality of life to adjust survival (so that 1 year of lifeat a "quality" of X is as desirable to the patient as6 months of life at a "quality" of 2X), it must be measuredas a "utility," which Dr. Weeks defined as a quantitativemeasure of the strength of a person's preference for an outcome.By convention, "utilities" are measured on a scale of0 to 1, in which 0 represents death and 1 represents perfect health,she said.

Three major approaches have been used to determine utility scores:a rating scale, such as a visual analogue scale (see "Cost Utility Study Shows Value of Aggressive Antiemetic Use"),completed by the patient in response to questions; the standardgamble, in which the utility of a health state is assessedby asking the patient how much he or she would be willing to riskto improve it; and the time tradeoff, which asks how muchtime one would give up to improve quality of life. The point ofindifference occurs when shortened life expectancy with perfectquality of life is equal to longer life expectancy with lowerquality of life.

Since all three of these methods have problems that make themdifficult to do in a clinical trial situation, Dr. Weeks offeredan alternative method of determining utilities--the "multiattributeutility approach," or health indices.

In this approach, a standard quality of life questionnaire isadministered and a formula is then used to convert the answersinto a utility score. The formula is determined by surveying areference population using standard gamble or time tradeoff techniques.

Several formulas are now available for generating utilities, Dr.Weeks said. At Dana-Farber, she is involved in developing a cancer-specificmultiattribute scale known as the "Q-tility Index,"based on Spitzer's Quality of Life Index (QLI), a five-dimensionscale specific for cancer. "For any possible combinationof answers to the QLI, you can generate a utility using the Q-tilityIndex," she said.

Once utilities are determined, they can be used along with clinicaltrial data to generate estimates of quality-adjusted survival.Q-TWST (quality-adjusted time without symptoms of disease or toxicityof treatment), for example, provides an analytic framework forputting together utility and survival data, she said.