To Register      SMDM Homepage

Tuesday, 19 October 2004 - 2:45 PM

This presentation is part of: Oral Concurrent Session A - Cost Effective Analysis: Methods

$50,000 PER QALY: INERTIA, INDIFFERENCE, OR IRRATIONALITY?

R. Scott Braithwaite, MD and Mark S. Roberts, MD. University of Pittsburgh, Section of Decision Sciences and Clinical Systems Modeling, Pittsburgh, PA

Purpose: $50,000 per Quality-Life Year (QALY) is a threshold commonly used to delineate cost-effectiveness, but it is controversial because it lacks theoretical basis, is not inflation-adjusted, and does not consider important societal factors (e.g. resources, technology growth). Therefore, our objective was to determine what cost-effectiveness thresholds are compatible with varying assumptions regarding health care resources and technology growth.

Methods: We developed a computer simulation that represents how individuals in the US interact with health care services. Use of a service produces an incremental gain in duration of life commensurate with its expenditure and cost-effectiveness. The simulation permits a cost-effectiveness threshold to be designated, which then constrains which available services are offered. If needs for offered services exceed allocated resources, services are excluded randomly without further regard to cost-effectiveness, and allocated resources may not be spent efficiently. If allocated resources exceed the needs for offered services, some resources may be unused. Data sources were based on published cost-effectiveness analyses of 1276 interventions, and US government surveys. We examined thresholds of $20,000, $50,000, $100,000, and unlimited $/QALY. All costs were in 2002 US dollars.

Results: 44% of health care services cost less than $20,000/QALY, and 78% cost less than $100,000/QALY. In the base case analysis (allocated resources = 14.9% of current US GDP), mean life expectancy (LE) was highest (74.9 years) without use of any cost-effectiveness threshold because each threshold resulted in some allocated resources being unused. When allocated resources were decreased to 10% of GNP, LE was highest using a threshold of $100,000/QALY (73.0 years) because resources were partially unused with lower thresholds (e.g., 33% of resources unused with $20,000/QALY, LE 70.5 years) and were used inefficiently with higher thresholds (unlimited $/QALY, LE 71.2 years). When allocated resources were decreased to 5%, LE was highest (70.2 years) with the lowest examined threshold ($20,000/QALY) because resources were used completely and more efficiently than with higher thresholds. When technology growth doubled the volume of available services but resources stayed constant, LE was highest (76.8 years) with a threshold of $50,000/QALY.

Conclusion: Optimal cost-effectiveness thresholds decrease with fewer resources or greater technology growth. Thresholds established without regard to these factors may lower life expectancy substantially compared to optimal thresholds, and sometimes compared to no threshold at all.


See more of Oral Concurrent Session A - Cost Effective Analysis: Methods
See more of The 26th Annual Meeting of the Society for Medical Decision Making (October 17-20, 2004)