Confidence Intervals for Cost-Effectiveness Ratios: A Comparison of Four Methods

Abstract

We evaluated four methods for computing confidence intervals for cost-effectiveness ratios developed from randomized controlled trials: the box method, the Taylor series method, the nonparametric bootstrap method and the Fieller theorem method. We performed a Monte Carlo experiment to compare these methods. We investigated the relative performance of each method and assessed whether or not it was affected by differing distributions of costs (normal and log normal) and effects (10% absolute difference in mortality resulting from mortality rates of 25% versus 15% in the two groups as well as from mortality rates of 55% versus 45%) or by differing levels of correlation between the costs and effects (correlations of -0.50, -0.25, 0.0, 0.25 and 0.50). The principal criterion used to evaluate the performance of the methods was the probability of miscoverage. Symmetrical miscoverage of the intervals was used as a secondary criterion for evaluating the four methods. Overall probabilities of miscoverage for the nonparametric bootstrap method and the Fieller theorem method were more accurate than those for the other the methods. The Taylor series method had confidence intervals that asymmetrically underestimated the upper limit of the interval. Confidence intervals for cost-effectiveness ratios resulting from the nonparametric bootstrap method and the Fieller theorem method were more dependably accurate than those estimated using the Taylor series or box methods. Routine reporting of these intervals will allow individuals using cost-effectiveness ratios to make clinical and policy judgments to better identify when an intervention is a good value for its cost.

Cost -Effectiveness of Invasive Versus Medical Management of Elderly Patients With Chronic Symptomatic Coronary Artery Disease: Findings of the Randomized Trial of Invasive Versus Medical Therapy in Elderly Patients With Chronic Angina (TIME)

“…Bootstrap techniques were used to estimate 95% confidence intervals (95% CI) for differences in average costs and effect measures (i.e. number of MACE, number of days without MACE, utility score) and for the incremental cost-effectiveness ratios presented (each of these simulations using 50 000 replicate 13 ), and also to assess the shape of the joint sampling distribution of the differences in average individual costs and effects between the two treatment groups (with 5000 replicate per simulation). Presenting cost-effectiveness results as ratios with 95% confidence intervals is insufficient, as their interpretation depends on the quadrants of the cost-effectiveness plane (CE plane) into which they fall.…”

Home treatment for mental health problems: a systematic review

“…Various methods have been recommended for dealing with this issue. [72][73][74] Polsky and colleagues (1997) 75 used a Monte Carlo simulation exercise to derive evidence, suggesting that the reliability of confidence intervals around incremental cost-effectiveness ratios generated by different methods may be sensitive to assumptions concerning the distribution of costs and effects data and the extent of correlation between costs and effects. This suggests that the methods used in future evaluations of home treatment should pay closer attention to distributional issues and the correlation between cost and outcomes when deciding on the most statistically reliable methodology to adopt.…”

Design Issues for Conducting Cost-Effectiveness Analyses Alongside Clinical Trials

“…The bootstrap method is perhaps the most robust method for computing sample size estimates for R (26). The bootstrap procedure is the most computationally intensive but also the least dependent on parametric behavior ofR.…”