THE MARGINAL COST-EFFECTIVENESS OF CHILDHOOD VACCINATION COVERAGE

Purpose:

   To explore the change in cost-effectiveness for marginal changes to population-level vaccination coverage and determine the net-cost-minimizing level of coverage. 

 

Method:

   A susceptible-infected-recovered (SIR) model representing a theoretical infectious disease was created to simulate disease spread. SIR model inputs included parameters intended to reflect childhood vaccine-preventable diseases. Costs included vaccination costs, disease costs, and monetized QALYs lost from disease. Analyses explored changes in population level vaccination coverage from 0 – 100% and evaluated vaccine efficacy between 80 – 100%. Different theoretical reproductive rates (R₀) from 2.0 to 18.0 were examined.  Primary model outcomes were the number of infected people and the total costs of vaccination and disease per person.

  

   Results were analyzed to find the level of population coverage that would have the minimum net cost.   We examined how much additional cost would be incurred if the optimal level of vaccination were not achieved.  We explored how the R₀, cost of vaccination, and cost of disease affected the optimal level of vaccination coverage and the costs of suboptimal vaccination coverage.

 

Result:

   Results show that under a wide range of assumptions, over vaccination is less expensive compared to under vaccination. For example, the figure below shows a disease with an R₀ of 3.0, vaccine efficacy of 90%, and a 1:10 vaccination to disease cost ratio.  It has a minimum net cost of $73.77 per person at 72.9% population vaccination coverage.  In this example under-vaccination by 2 percentage points of population coverage is just as costly as over-vaccination by 11 percentage points. This observation is violated when the cost of vaccination becomes approximately equal to the cost of disease.

 

Figure

 

Conclusion:

   These results suggest that under vaccination is more costly than over vaccination. All of the models from this study were simple and assume homogeneous population mixing. Despite the simplicity of these models, this work helps provide information for assessing the impact of changes in vaccination coverage at a societal level.  Further research should examine more complex models and specific vaccines.