Interpretation of COVID-19 Epidemiological Trends in Mexico Through Wastewater Surveillance Using Simple Machine...

Area of Science:

• Environmental science
• Epidemiology
• Public health

Background:

• Wastewater-based surveillance (WBS) is an efficient method for tracking infectious diseases.
• Analyzing WBS data for public health decision-making, especially during pandemics like COVID-19, presents challenges.
• Limited data dimensionality can impact the accuracy of predictive models.

Purpose of the Study:

• To develop and evaluate simple clustering and regression models for WBS data analysis.
• To assess the models' ability to inform prevention and control measures in high-affluence settings.
• To determine the effectiveness of WBS-supported models for public health decision-making.

Main Methods:

• Utilized wastewater sampling data from Mexico City and Monterrey (2021-2022).
• Trained clustering models to differentiate between COVID-19 surge weeks and non-surge weeks.
• Employed regression models to predict weekly average daily new COVID-19 cases.

Main Results:

• Clustering models achieved 87.9% accuracy in differentiating COVID-19 surge periods.
• Forecasting accuracy for one and two weeks ahead was 80.4% and 81.8%, respectively.
• Regression model predictions for weekly average new daily cases showed limitations (R² = 0.80, MAPE = 72.6%).

Conclusions:

• Simple WBS-supported models provide valuable insights for public health decision-makers during outbreaks.
• Clustering approaches demonstrate strong accuracy for identifying disease trends.
• Predictive regression models using low-dimensionality WBS data require further refinement for improved accuracy.