Cost-utility analysis of bevacizumab versus ranibizumab in neovascular age-related macular degeneration using a Markov model

Objective To evaluate the cost‐effectiveness of intravitreal bevacizumab to ranibizumab in patients with neovascular age‐related macular degeneration (AMD). Methods A cost‐utility analysis using a Markov model was performed to evaluate incremental cost‐effectiveness ratio [ICER, $US per quality‐adjusted life year (QALY) gained] between bevacizumab and ranibizumab from a US payer perspective. Transition probabilities for ranibizumab and bevacizumab were extrapolated from published studies and local institutional data. Utility values, likewise, were obtained from another published study. Mortality rates were determined from the Centers for Disease Control 2003 Life Tables. Resource utilization and total direct costs were estimated using the Centers for Medicare and Medicaid Services and the Veterans Affairs Decision Support System. A hypothetical cohort of 1000 patients was simulated through the model for 20 years. Sensitivity analyses were performed using univariate and probabilistic sensitivity analysis (PSA) on all costs, transition probabilities and utility values. An acceptability curve was generated to illustrate the cost‐effectiveness probability of bevacizumab to ranibizumab with increasing willingness‐to‐pay (WTP). Results The cost‐effectiveness ratios (CER) for bevacizumab and ranibizumab were $1405 per QALY and $12 177 per QALY, respectively. The ICER for bevacizumab was dominant compared to ranibizumab. The base‐case CER was sensitive to drug costs of the study medications with a breakeven point of $44 for ranibizumab and $2666 for bevacizumab. PSA revealed a 95% probability of bevacizumab being more cost‐effective than ranibizumab at a WTP of $50 000 per QALY gained. Conclusion Based on a WTP defined at $50 000 per QALY gained, bevacizumab was cost‐effective versus ranibizumab 95% of the time because of lower acquisition costs and increased efficacy.