Inference of seasonal and pandemic influenza transmission dynamics

Abstract

The inference of key infectious disease epidemiological parameters is critical for characterizing disease spread and devising prevention and containment measures. The recent emergence of surveillance records mined from big data such as health-related online queries and social media, as well as model inference methods, permits the development of new methodologies for more comprehensive estimation of these parameters. We use such data in conjunction with Bayesian inference methods to study the transmission dynamics of influenza. We simultaneously estimate key epidemiological parameters, including population susceptibility, the basic reproductive number, attack rate, and infectious period, for 115 cities during the 2003-2004 through 2012-2013 seasons, including the 2009 pandemic. These estimates discriminate key differences in the epidemiological characteristics of these outbreaks across 10 y, as well as spatial variations of influenza transmission dynamics among subpopulations in the United States. In addition, the inference methods appear to compensate for observational biases and underreporting inherent in the surveillance data.influenza | transmission dynamics | data assimilation | asymptomatic infections | spatial patterns

Transmission network of the 2014–2015 Ebola epidemic in Sierra Leone

“…depletion of actual susceptibles) or from an increase in awareness of the disease and precautions and intervention measures taken that reduce the chance of transmission and effectively remove individuals from the susceptible pool [43]. In addition, our previous study suggests that population susceptibility and the basic reproductive number tend to compensate for each other, whereas estimates of the effective reproductive number (calculated as R e ¼ R 0 S/N for a non-network model) are generally more accurate [23]. As such, we here focus on the effective reproductive number R e , and did not analyse the population susceptibility or the basic reproductive number R 0 .…”

“…Accordingly, we model the propagation of Ebola through a population using a susceptible-exposed-infectious-removed (SEIR) compartmental model [21]. This choice is also based on our previous work indicating that more parsimonious model structures, such as the SEIR model, tend to be more easily optimized than more complex modelling frameworks [22,23]. In a network of multiple districts, residents within each district may contract the disease locally or when traveling to other districts.…”

“…S, E, I, for each district) and model parameters, including the transmission rate b for each district, the incubation period, the infectious period and the three exponents (t 1 , t 2 and r) for the gravity model. The EAKF [15,23,[25][26][27]] is a data assimilation method that uses an ensemble of system replicas to represent the distribution of possible model state and parameter values. It approximates a Bayesian update to the model states and parameters using the observational data and an estimate of the errors in the data.…”

Real Time Influenza Monitoring Using Hospital Big Data in Combination with Machine Learning Methods: Comparison Study (Preprint)

“…Outbreaks cause up to 5 million severe cases and 500,000 deaths per year worldwide. [1][2][3][4][5] During influenza peaks, the large increase of visits to general practitioners and to emergency departments causes healthcare system disruption. To reduce its impact and to help organizing adapted sanitary responses, it is necessary to monitor influenza-like illness (ILI; any acute respiratory infection with fever ≥ 38 °C, cough and onset within the last 10 days) activity.…”

“…They can be from social networks (e.g, Facebook,Twitter), viewing sites, (e.g, YouTube, Netflix), shopping sites, (e.g, Amazon, Cdiscount), but also from sales or rentals website between particulars (e.g, Craigslist, Airbnb). In the case of influenza, some studies used data from Google [2,4,9,[13][14][15][16] Twitter [17,18] or Wikipedia [19][20][21] The biggest advantage of web data is that they are produced in real time. One of the first and most famous studies on the use of internet data for detecting influenza epidemics is Google Flu Trends, [13,22] a web service operated by Google.…”