Practical unidentifiability of a simple vector-borne disease model: Implications for parameter estimation and...

Area of Science:

• Epidemiology
• Mathematical Biology
• Disease Modeling

Background:

• Mathematical modeling is crucial for understanding and predicting vector-borne diseases.
• Parameter estimation links models to real-world data but is often unevaluated.
• Identifiability analysis is essential to ensure model parameters can be uniquely estimated.

Purpose of the Study:

• To evaluate the structural and practical identifiability of a common compartmental model for mosquito-borne diseases.
• To assess parameter estimation accuracy using a dengue epidemic case study (Taiwan, 2010).
• To determine the impact of identifiability on prediction and intervention strategy evaluation.

Main Methods:

• Conducted structural and practical identifiability analyses on a compartmental mosquito-borne disease model.
• Utilized time series data from a human and mosquito population during a dengue epidemic.
• Examined parameter estimation under various measurement scenarios.

Main Results:

• The model was found to be structurally identifiable but practically unidentifiable with typical data.
• Transmission parameters could not be estimated separately, forming an identifiable combination.
• Despite parameter unidentifiability, the basic reproduction number was reliably estimated.

Conclusions:

• Common vector-borne disease models may suffer from practical unidentifiability, challenging parameter estimation and intervention predictions.
• Combining experimental, field, and case data is vital for resolving identifiability issues.
• Identifiability analysis is critical for reliable model-based predictions and decision-making in epidemiology.