Did You Know That Confounding Could Bias Results Obtained within an Economic Evaluation?

Abstract

Background: The increasing use of real-world data in economic evaluation raises concerns about the potential for confounding bias. Methods to control for this bias have been developed, but there is still much to be understood about how confounding influences economic evaluations. Objectives: To illustrate the impact of unadjusted confounding variables in economic evaluations of observational studies. Methods: We simulated the costs and effectiveness of 2 treatments across 9 possible confounding effect scenarios. We considered these scenarios in the context in which one treatment is more costly and more effective with mild correlation between confounders and outcomes. All scenarios examined the incremental cost and effectiveness of 400 randomly generated individuals, reflecting sample sizes commonly seen within observational economic evaluations. Results were illustrated with the use of cost-effectiveness planes and cost-effectiveness acceptability curves (CEAC). Results: Our simulations illustrate that confounding bias can have a significant effect on incremental costs and incremental effectiveness estimates. These simulations also illustrate that confounders can affect the evaluation of uncertainty by causing a shift in the CEACs. Such results hint that inadequate consideration of confounding bias can potentially lead to flawed judgments about the cost-effectiveness of a treatment. Discussion: Results of economic evaluations can be influenced by confounding when they are conducted in an observational setting. Efforts must be made to limit their impact to ensure an accurate assessment of the economic value of treatments and to prevent potential losses of population health.