A Markov model for sample size calculation and inference in vaccine cost-effectiveness studies

Area of Science:

• Health Economics
• Vaccinology
• Biostatistics

Background:

• Sample size calculations in vaccine cost-effectiveness trials have historically focused on vaccine effectiveness.
• Existing methods may not adequately address the complexities of economic evaluations, particularly regarding non-independent events like work absences.

Purpose of the Study:

• To derive novel sample size and power formulae for designing robust vaccine cost-effectiveness trials in working adults.
• To incorporate key economic factors, including vaccine costs and the impact on workdays lost.

Main Methods:

• Development of sample size and power formulae tailored for vaccine cost-effectiveness analysis.
• Introduction of a Markov model to handle the non-independence of working days lost due to illness.
• Specification of two critical effect sizes: reduction in absence episodes and reduction in mean episode duration.

Main Results:

• Application of the derived formulae to published influenza vaccine trials suggests that most were underpowered.
• Underpowered studies may fail to detect statistically significant financial benefits of vaccination.
• Biased variance estimates in previous studies likely led to incorrect conclusions regarding cost-effectiveness.

Conclusions:

• The developed formulae provide a more accurate framework for sample size determination in vaccine cost-effectiveness research.
• Accurate sample size and power calculations are crucial for reliable economic evaluations of vaccines.
• Future trials should utilize these advanced methods to ensure adequate power for detecting economic benefits.