Reproducibility Status of Economic Evaluations of Cancer Drugs: An Assessment of 150 Cost-Effectiveness Models

Purpose Model-based cost-effectiveness analyses (CEAs) published in scientific journals have increased in recent years. The studies involve many discretionary judgements that, if not properly reported and justified, may reduce the credibility of cost-effectiveness evidence. In this study, we aimed to evaluate the reproducibility of published model-based CEAs and to explore associations between reproducibility and study characteristics. Methods We conducted a systematic search for model-based CEAs of cancer drugs. Data were extracted on the journal type, impact factor, study design, model structure, and model inputs. Reproducibility was assessed based on access to model code (Computational Reproducibility) and whether reporting was transparent and sufficiently detailed across four criteria (model structure, transitions, costs, and health outcomes) for an external researcher to recreate the model (Recreate Reproducibility). Logistic regression was used to investigate associations between reproducibility and study characteristics. Results Among 266 eligible CEAs, 150 were randomly selected for full-text review, of which 2.7% (95% CI, 1.0 to 6.7) were assessed as Computationally Reproducible and 16.7% (95% CI, 11.6 to 23.4) as Recreate Reproducible. While most studies reported the overall model structure (85% [95% CI, 77 to 91]), only 39% (95% CI, 28 to 51) fully reported all cost inputs. Regression analyses showed no association between reproducibility and study characteristics such as funding source, country setting, or journal impact factor. Conclusion Only a small fraction of oncology CEAs was assessed as reproducible, irrespective of journal or study characteristics. Critical information necessary for external review and reproducibility remains largely unavailable, even in high-impact oncology journals. Standardized reporting and open modeling practices are needed to strengthen the scientific reliability of cost-effectiveness evidence in oncology.