Measuring sensitivity in pharmacoeconomic studies. Refining point sensitivity and range sensitivity by incorporating...

Area of Science:

• Health economics
• Decision analysis

Background:

• Economic models are crucial for healthcare decision-making.
• Quantifying uncertainty is vital for robust economic evaluations.
• Existing sensitivity analysis methods have limitations in reflecting true uncertainty.

Purpose of the Study:

• To refine a point sensitivity analysis method for economic studies.
• To improve the assessment of uncertainty in economic models.
• To enhance the accuracy of sensitivity analysis in healthcare economics.

Main Methods:

• Incorporation of probability distributions into the sensitivity analysis.
• Utilizing a bootstrap method for fixed input variables with limited data.
• Application of the refined method to a Markov model of depression.

Main Results:

• The refined method significantly altered the ranking of sensitive variables.
• Response rate emerged as the most sensitive variable, surpassing 'per diem' for hospitalization.
• Identified key drivers of uncertainty in the depression Markov model.

Conclusions:

• The refined method provides a more realistic reflection of uncertainty in economic studies.
• Enhanced sensitivity analysis improves the validity of economic evaluations.
• The updated approach offers a more accurate understanding of model uncertainty.