Given the recent approval and recommendation of V116, a 21-valent pneumococcal conjugate vaccine (PCV), in the United States(US), we evaluated the cost-effectiveness of using V116 versus the 20-valent PCV(PCV20) or the 15-valent PCV(PCV15) in series with the 23-valent pneumococcal polysaccharide vaccine (PPSV23) among adults aged ≥ 65years in the US who had never received a PCV previously. A static multi-cohort state-transition Markov model was developed to estimate the lifetime incremental clinical and economic impact of V116 vs. PCV20 or PCV15 + PPSV23 from the societal perspective. All model inputs were based on published literature and publicly available databases and/or reports. Model outcomes included undiscounted clinical cases: invasive pneumococcal disease (IPD), inpatient and outpatient non-bacteremic pneumococcal pneumonia (NBPP), post-meningitis sequelae (PMS), deaths from IPD and inpatient NBPP, discounted quality-adjusted life years (QALYs) as well as the discounted total cost (in 2023 USD), which consisted of vaccine acquisition and administration costs, direct and indirect costs associated with the disease, and travel costs for vaccination. The final summary measure was the incremental cost-effectiveness ratio (ICER), reported as $/QALY gained. Three percent was used for the annual discounting rate. Based on the inputs and assumptions used, the results indicated that the V116 strategy prevented 27,766 and 32,387 disease cases/deaths and saved $239 million and $1.8 billion in total costs when compared to the PCV20andPCV15 + PPSV23 strategies, respectively, in vaccine-naïve adults aged ≥ 65years. The estimated ICERs were cost saving in both regimens (i.e., V116 vs. PCV20 or vs. PCV15 + PPSV23). The scenario analysis and deterministic and probabilistic sensitivity analyses also demonstrated the robustness of the qualitative results. These results demonstrated that using V116 in adults aged ≥ 65years in the US can prevent a substantial number of PD cases and deaths while remaining highly favorable economically over a wide range of inputs and scenarios.
Read full abstract