OBJECTIVES: Age is an important modeling choice in cost effectiveness analysis (CEAs). Often, CEAs follow a single age cohort of individuals, and the results are used to represent an age group. This is sufficient when the single age selected is representative of that group. Otherwise, the results could be incorrect. By contrast, a model could track all ages in the age group of interest, so the results are more accurate for that age group. This study demonstrates how the results vary with the choice of the representative age for single age cohort analysis, and their difference from multi-age cohort analyses.

METHODS: For illustration, we employed a decision-analytic state-transition model for pneumococcal vaccines and compared the cost effectiveness of PCV15+PPSV23 vs. PCV20 for age group 50-64 from a societal perspective. Costs (in $2022 USD) and outcomes were discounted using an annual discount rate of 3%. The incremental cost effectiveness ratios (ICERs) were recorded from each single cohort analysis using 50 to 64 as the representative age, and from multi-age cohort analysis following all ages in 50-64.

RESULTS: The ICERs for the single age cohort analyses fluctuated and ranged from cost saving to $187million. The ICER from the multi-age cohort analysis was $7.2 million. In this example, the multi-age cohort ICER was closest to the ICERs for single age cohort analysis using age 57 (ICER = $6.7 million), and age 58 (ICER = 8.1 million).

CONCLUSIONS: Using a single age to represent an age group when conducting CEAs could be inaccurate as the representative age selected has a considerable impact on outcomes, though single age cohort analysis has its advantages of conceptual simplicity and adequacy in certain circumstances. In some other cases, multi age cohort analyses might be a more accurate approach with the potential cost of longer computation time.