New statistical model for misreported data with application to current public health challenges

Area of Science:

• Epidemiology
• Biostatistics
• Time Series Analysis

Background:

• Accurate time series data is crucial for epidemiological surveillance.
• Continuous time series data can suffer from underreporting or overreporting.
• Existing models may not adequately address autocorrelation in misreported data.

Purpose of the Study:

• To develop and validate a novel statistical model for continuous time series data with potential reporting inaccuracies.
• To incorporate the autocorrelation structure of time series data into the model.
• To assess the model's performance using simulations and real-world epidemiological data.

Main Methods:

• Development of a new statistical model designed for continuous time series.
• Simulation studies to evaluate model performance under various autocorrelation structures.
• Application of the model to human papillomavirus (HPV) incidence data from Girona, Spain.
• Application to Covid-19 incidence data from Heilongjiang, China, and Catalonia, Spain.

Main Results:

• The proposed model effectively handles autocorrelation in continuous time series data.
• Demonstrated ability to manage partially or totally underreported/overreported data.
• Successful application to real-world epidemiological datasets, including HPV and Covid-19 incidence.

Conclusions:

• The new model provides a robust approach for analyzing time series data with reporting errors.
• It improves the reliability of epidemiological surveillance by accounting for data inaccuracies and autocorrelation.
• The model shows promise for public health applications in monitoring disease incidence.