Climatic risk factors and predictive modelling of canine parvoviral enteritis outbreaks: An eight-year study from Southern India.

Abstract

Canine Parvovirus (CPV) poses a significant threat to young puppies globally, leading to high morbidity and mortality rates. Despite its widespread impact, research on the influence of climate on CPV outbreaks is limited. This study aimed to analyze risk factors, investigate climatic associations and develop forecasting models for Canine Parvoviral Enteritis (CPVE) using eight years of data from the Teaching Veterinary Hospital at Madras Veterinary College, Chennai, Southern India. Among 6105 suspected cases, 4258 dogs were diagnosed with CPVE, resulting in a positivity rate of 69.75 %. The study identified winter season and July month as periods with higher incidences of CPVE outbreaks. Monthly data analysis revealed a positive correlation between CPVE occurrence and maximum temperature (lagged by 11 months), morning relative humidity (lagged by 1 month) and rainfall (lagged by 10 months). Conversely, negative correlations were found with maximum temperature (lagged by 1 & 6 months), minimum temperature (lagged by 4 months), morning relative humidity (lagged by 7 months), evening relative humidity (lagged by 7 months), rainfall (lagged by 2 & 7 months) and wind speed (lagged by 4 months). For predictive analysis, the Autoregressive Integrated Moving Average with eXogenous variable (ARIMAX) model performed best, incorporating two and seven-month lagged rainfall values and four-month lagged wind speed data. Similarly, the Extreme Gradient Boosting (XGBoost) and Recurrent Neural Network (RNN) models, though not considering climatic data, also demonstrated optimal performance. This study emphasizes the critical need for continuous global monitoring of CPVE and underscores the significant impact of climate on its outbreaks. These findings are crucial for shaping effective intervention strategies to mitigate the impact of CPVE on canine populations.