Presentation (CTI)

OBJECTIVES: Cost-effectiveness analyses (CEAs) for chronic diseases frequently depend on clinical trial data with limited follow-up, requiring extrapolation of clinical and economic outcomes over long time horizons. Such extrapolations introduce uncertainty, which is commonly explored using probabilistic sensitivity analysis (PSA). Increasing availability of real-world evidence (RWE) provides long-term outcomes and reduce extrapolation uncertainty. The objective of this study was to evaluate methodological approaches for incorporating long-term real-world outcome data into the PSA framework of CEAs, and to assess the impact of these approaches on long-term cost-effectiveness results.
METHODS: Short-term trial survival data were simulated with a mean of 36 months (standard deviation [SD] = 5), while long-term RWE survival data had a mean of 60 months (SD = 8). A normal-normal conjugate Bayesian updating framework was applied to combine trial and RWE evidence, yielding a posterior mean survival of 42.74 months (SD = 4.24). Uncertainty around survival parameters was propagated by sampling from posterior distributions within each PSA iteration. Two PSA scenarios were evaluated: a trial-only scenario, and an RWE-integrated scenario with costs and QALYs. Monte Carlo simulation with 1,000 iterations was conducted and cost-effectiveness acceptability curves (CEACs) were generated.
RESULTS: Bayesian updating positioned long-term survival estimates between trial-only and RWE-only projections while reducing parameter uncertainty. The mean ICER decreased from USD 12,608 in the trial-only scenario to USD 11,608 following RWE integration. The 95% uncertainty interval for the ICER narrowed from USD 8,016-18,619 to USD 8,066-15,873. At a willingness-to-pay threshold of USD 20,000 per QALY gained, the probability of cost-effectiveness increased from 98.8% to 100%.
CONCLUSIONS: Incorporating long-term real-world outcome data into PSA represents an important methodological advancement for addressing extrapolation uncertainty. These approaches provide practical options for leveraging RWE to generate more credible, transparent, and decision-relevant cost-effectiveness evidence, supporting health technology assessment decisions in settings requiring long-term projections.