Of Lyme disease and machine learning in a One Health world.

Area of Science:

• Public Health
• Epidemiology
• One Health

Background:

• Lyme disease is an emerging vector-borne zoonosis in Ontario, influenced by population growth and climate change.
• It exemplifies the One Health concept, linking human, animal, and environmental health.
• Health communication and disease surveillance are crucial for managing Lyme disease in the absence of a vaccine.

Purpose of the Study:

• To forecast the future burden of Lyme disease in Ontario.
• To evaluate the effectiveness of automated machine learning and statistical learning approaches for Lyme disease surveillance.

Main Methods:

• Utilized Lyme disease surveillance data from Ontario's integrated Public Health Information System (2005-2023).
• Employed a feed-forward single-layer neural network and a seasonal autoregressive integrated moving-average (SARIMA) model.
• Trained models on data from 2005-2021 and validated using data from 2022-2023.

Main Results:

• Lyme disease burden in Ontario is projected to increase dramatically.
• Both the neural network and SARIMA models demonstrated comparable accuracy in their forecasts.
• Forecasts from both models served as benchmarks for each other.

Conclusions:

• The escalating burden of Lyme disease poses significant public health concerns.
• This trend underscores ongoing ecosystem changes and presents challenges for canine health.
• Human Lyme disease surveillance data offers valuable insights for veterinary professionals.