Cross-correlation and time series analysis of rabies in different animal species in Nepal from 2005 to 2018

Abstract

Rabies is a fatal zoonotic disease, resulting in human and livestock deaths. In Nepal, animal rabies has posed a significant challenge to public health. Because animals are the primary source of rabies in humans, a better understanding of rabies epidemiology in animals is necessary. The objectives of this study were to determine the correlation between rabies occurrences in dogs and livestock animals and to detect the trends and change points of the disease using longitudinal data. The nationwide rabies dataset from 2005 to 2018 was analyzed using cross-correlation, multiple change points, and time series methods. Autoregressive Integrated Moving Average (ARIMA) and Neural Network Autoregression (NNAR) were applied to the time series data. The results show a positive correlation between canine rabies and livestock rabies occurrences. Three significant change points were detected in the time series data, demonstrating that the occurrences were high in the initial years but stabilized before peaking to an upward trend in the final years of the study period. Nonetheless, there was no seasonality pattern in rabies occurrences. The most suitable models were ARIMA (2,1,2) and NNAR (5,1,4) (12). Based on the study findings, both locals and tourists in Nepal need to have enhanced awareness of the potential dangers posed by rabies in canines and livestock. This study offers much-needed insight into the patterns and epidemiology of animal rabies which will be helpful for policymakers in drafting rabies control plans for Nepal.

Keywords

• Cross-correlation
• Forecasting
• Nepal
• Rabies
• Time series