Inference and forecast of H7N9 influenza in China, 2013 to 2015

A systematic review of mechanistic models used to study avian influenza virus transmission and control

The global spread of avian influenza A viruses in domestic birds is causing increasing socioeconomic devastation. Various mechanistic models have been developed to better understand avian influenza transmission and evaluate the effectiveness of control measures in mitigating the socioeconomic losses caused by these viruses. However, the results of models of avian influenza transmission and control have not yet been subject to a comprehensive review. Such a review could help inform policy makers and guide future modeling work. To help fill this gap, we conducted a systematic review of the mechanistic models that have been applied to field outbreaks. Our three objectives were to: (1) describe the type of models and their epidemiological context, (2) list estimates of commonly used parameters of low pathogenicity and highly pathogenic avian influenza transmission, and (3) review the characteristics of avian influenza transmission and the efficacy of control strategies according to the mechanistic models. We reviewed a total of 46 articles. Of these, 26 articles estimated parameters by fitting the model to data, one evaluated the effectiveness of control strategies, and 19 did both. Values of the between-individual reproduction number ranged widely: from 2.18 to 86 for highly pathogenic avian influenza viruses, and from 4.7 to 45.9 for low pathogenicity avian influenza viruses, depending on epidemiological settings, virus subtypes and host species. Other parameters, such as the durations of the latent and infectious periods, were often taken from the literature, limiting the models' potential insights. Concerning control strategies, many models evaluated culling (n = 15), while vaccination received less attention (n = 6). According to the articles reviewed, optimal control strategies varied between virus subtypes and local conditions, and depended on the overall objective of the intervention. For instance, vaccination was optimal when the objective was to limit the overall number of culled flocks. In contrast, pre-emptive culling was preferred for reducing the size and duration of an epidemic. Early implementation consistently improved the overall efficacy of interventions, highlighting the need for effective surveillance and epidemic preparedness.

Regional Distribution of Non-human H7N9 Avian Influenza Virus Detections in China and Construction of a Predictive Model

Introduction

H7N9 avian influenza has broken out in Chinese poultry 10 times since 2013 and impacted the industry severely. Although the epidemic is currently under control, there is still a latent threat.

Material and Methods

Epidemiological surveillance data for non-human H7N9 avian influenza from April 2013 to April 2020 were used to analyse the regional distribution and spatial correlations of positivity rates in different months and years and before and after comprehensive immunisation. In addition, positivity rate monitoring data were disaggregated into a low-frequency and a high-frequency trend sequence by wavelet packet decomposition (WPD). The particle swarm optimisation algorithm was adopted to optimise the least squares support-vector machine (LS-SVM) model parameters to predict the low-frequency trend sequence, and the autoregressive integrated moving average (ARIMA) model was used to predict the high-frequency one. Ultimately, an LS-SVM-ARIMA combined model based on WPD was constructed.

Results

The virus positivity rate was the highest in late spring and early summer, and overall it fell significantly after comprehensive immunisation. Except for the year 2015 and the single month of December from 2013 to 2020, there was no significant spatiotemporal clustering in cumulative non-human H7N9 avian influenza virus detections. Compared with the ARIMA and LS-SVM models, the LS-SVM-ARIMA combined model based on WPD had the highest prediction accuracy. The mean absolute and root mean square errors were 2.4% and 2.0%, respectively.

Conclusion

Low error measures prove the validity of this new prediction method and the combined model could be used for inference of future H7N9 avian influenza virus cases. Live poultry markets should be closed in late spring and early summer, and comprehensive H7N9 immunisation continued.

Real-Time Forecasting of Hand-Foot-and-Mouth Disease Outbreaks using the Integrating Compartment Model and Assimilation Filtering

Hand-foot-and-mouth disease (HFMD) is a highly contagious viral infection, and real-time predicting of HFMD outbreaks will facilitate the timely implementation of appropriate control measures. By integrating a susceptible-exposed-infectious-recovered (SEIR) model and an ensemble Kalman filter (EnKF) assimilation method, we developed an integrated compartment model and assimilation filtering forecast model for real-time forecasting of HFMD. When applied to HFMD outbreak data collected for 2008-11 in Beijing, China, our model successfully predicted the peak week of an outbreak three weeks before the actual arrival of the peak, with a predicted maximum infection rate of 85% or greater than the observed rate. Moreover, dominant virus types enterovirus 71 (EV-71) and coxsackievirus A16 (CV-A16) may account for the different patterns of HFMD transmission and recovery observed. The results of this study can be used to inform agencies responsible for public health management of tailored strategies for disease control efforts during HFMD outbreak seasons.

Influenza A H5N1 and H7N9 in China: A spatial risk analysis

Background

Zoonotic avian influenza poses a major risk to China, and other parts of the world. H5N1 has remained endemic in China and globally for nearly two decades, and in 2013, a novel zoonotic influenza A subtype H7N9 emerged in China. This study aimed to improve upon our current understanding of the spreading mechanisms of H7N9 and H5N1 by generating spatial risk profiles for each of the two virus subtypes across mainland China.

Methods and findings

In this study, we (i) developed a refined data set of H5N1 and H7N9 locations with consideration of animal/animal environment case data, as well as spatial accuracy and precision; (ii) used this data set along with environmental variables to build species distribution models (SDMs) for each virus subtype in high resolution spatial units of 1km² cells using Maxent; (iii) developed a risk modelling framework which integrated the results from the SDMs with human and chicken population variables, which was done to quantify the risk of zoonotic transmission; and (iv) identified areas at high risk of H5N1 and H7N9 transmission. We produced high performing SDMs (6 of 8 models with AUC > 0.9) for both H5N1 and H7N9. In all our SDMs, H7N9 consistently showed higher AUC results compared to H5N1, suggesting H7N9 suitability could be better explained by environmental variables. For both subtypes, high risk areas were primarily located in south-eastern China, with H5N1 distributions found to be more diffuse and extending more inland compared to H7N9.

Conclusions

We provide projections of our risk models to public health policy makers so that specific high risk areas can be targeted for control measures. We recommend comparing H5N1 and H7N9 prevalence rates and survivability in the natural environment to better understand the role of animal and environmental transmission in human infections.