In this work, we aim to determine the main factors driving self-initiated behavioral changes during the seasonal flu.To this end, we designed and deployed a questionnaire via Influweb, a Web platform for participatory surveillance in Italy, during the 2017 -18 and 2018 -19 seasons.We collected 599 surveys completed by 434 users.The data provide socio-demographic information, level of concerns about the flu, past experience with illnesses, and the type of behavioral changes voluntarily implemented by each participant.We describe each response with a set of features and divide them in three target categories.These describe those that report i) no (26%), ii) only moderately (36%), iii) significant (38%) changes in behaviors.In these settings, we adopt machine learning algorithms to investigate the extent to which target variables can be predicted by looking only at the set of features.Notably, 66% of the samples in the category describing more significant changes in behaviors are correctly classified through Gradient Boosted Trees.Furthermore, we investigate the importance of each feature in the classification task and uncover complex relationships between individuals' characteristics and their attitude towards behavioral change.We find that intensity, recency of past illnesses, perceived susceptibility to and perceived severity of an infection are the most significant features in the classification task and are associated to significant changes in behaviors.Overall, the research contributes to the small set of empirical studies devoted to the data-driven characterization of behavioral changes induced by infectious diseases. Author summaryHuman behavior and infectious diseases are linked by a feedback loop.While individuals might change their behavior as a response to an epidemic, such changes might influence the spreading itself.So far, our understanding and characterization of behavioral changes induced by diseases has been strongly limited by the lack of empirical data.As result, the vast majority of research has been focused on theoretical, what if, scenarios.In this work, we collected a unique dataset comprised of 599 surveys submitted by 434 users to the participatory surveillance platform Influweb over the 2017 -18 and 2018 -19 flu seasons.

Seasonal influenza surveillance is usually carried out by sentinel general practitioners (GPs) who compile weekly reports based on the number of influenza-like illness (ILI) clinical cases observed among visited patients. This traditional practice for surveillance generally presents several issues, such as a delay of one week or more in releasing reports, population biases in the health-seeking behaviour, and the lack of a common definition of ILI case. On the other hand, the availability of novel data streams has recently led to the emergence of non-traditional approaches for disease surveillance that can alleviate these issues. In Europe, a participatory web-based surveillance system called Influenzanet represents a powerful tool for monitoring seasonal influenza epidemics thanks to aid of self-selected volunteers from the general population who monitor and report their health status through Internet-based surveys, thus allowing a real-time estimate of the level of influenza circulating in the population. In this work, we propose an unsupervised probabilistic framework that combines time series analysis of self-reported symptoms collected by the Influenzanet platforms and performs an algorithmic detection of groups of symptoms, called syndromes. The aim of this study is to show that participatory web-based surveillance systems are capable of detecting the temporal trends of influenza-like illness even without relying on a specific case definition. The methodology was applied to data collected by Influenzanet platforms over the course of six influenza seasons, from 2011-2012 to 2016-2017, with an average of 34,000 participants per season. Results show that our framework is capable of selecting temporal trends of syndromes that closely follow the ILI incidence rates reported by the traditional surveillance systems in the various countries (Pearson correlations ranging from 0.69 for Italy to 0.88 for the Netherlands, with the sole exception of Ireland with a correlation of 0.38). The proposed framework was able to forecast quite accurately the ILI trend of the forthcoming influenza season (2016-2017) based only on the available information of the previous years (2011-2016). Furthermore, to broaden the scope of our approach, we applied it both in a forecasting fashion to predict the ILI trend of the 2016-2017 influenza season (Pearson correlations ranging from 0.60 for Ireland and UK, and 0.85 for the Netherlands) and also to detect gastrointestinal syndrome in France (Pearson correlation of 0.66). The final result is a near-real-time flexible surveillance framework not constrained by any specific case definition and capable of capturing the heterogeneity in symptoms circulation during influenza epidemics in the various European countries.

Traditional surveillance of seasonal influenza is generally affected by reporting lags of at least one week and by continuous revisions of the numbers initially released. As a consequence, influenza forecasts are often limited by the time required to collect new and accurate data. On the other hand, the availability of novel data streams for disease detection can help in overcoming these issues by capturing an additional surveillance signal that can be used to complement data collected by public health agencies. In this study, we investigate how combining both traditional and participatory Web-based surveillance data can provide accurate predictions for seasonal influenza in real-time fashion. To this aim, we use two data sources available in Italy from two different monitoring systems: traditional surveillance data based on sentinel doctors reports and digital surveillance data deriving from a participatory system that monitors the influenza activity through Internet-based surveys. We integrate such digital component in a linear autoregressive exogenous (ARX) model based on traditional surveillance data and evaluate its predictive ability over the course of four influenza seasons in Italy, from 2012-2013 to 2015-2016, for each of the four weekly time horizons. Our results show that by using data extracted from a Web-based participatory surveillance system, which are usually available one week in advance with respect to traditional surveillance, it is possible to obtain accurate weekly predictions of influenza activity at national level up to four weeks in advance. Compared to a model that is only based on data from sentinel doctors, our approach significantly improves real-time forecasts of influenza activity, by increasing the Pearson's correlation up to 30% and by reducing the Mean Absolute Error up to 43% for the four weekly time horizons.

The availability of novel digital data streams that can be used as proxy for monitoring infectious disease incidence is ushering in a new era for real-time forecast approaches to disease spreading. Here, we propose the first seasonal influenza forecast framework based on a stochastic, spatially structured mechanistic model (individual level microsimulation) initialized with geo-localized microblogging data. The framework provides for more than 600 census areas in the United States, Italy and Spain, the initial conditions for a stochastic epidemic computational model that generates an ensemble of forecasts for the main indicators of the epidemic season: peak time and intensity. We evaluate the forecasts accuracy and reliability by comparing the results with the data from the official influenza surveillance systems in the US, Italy and Spain in the seasons 2014/15 and 2015/16. In all countries studied, the proposed framework provides reliable results with leads of up to 6 weeks that became more stable and accurate with progression of the season. The results for the United States have been generated in real-time in the context of the Centers for Disease Control and Prevention ``Forecasting the Influenza Season Challenge''. A characteristic feature of the mechanistic modeling approach is in the explicit estimate of key epidemiological parameters relevant for public health decision-making that cannot be achieved with statistical models that do not consider the disease dynamic. Furthermore, the presented framework allows the fusion of multiple data streams in the initialization stage and can be enriched with census, weather and socioeconomic data.