Infodemiology and infoveillance: framework for contagious exanthematous diseases, of childhood in Italy

ABSTRACT

Background

Contagious exanthematous diseases are becoming a major public health problem. The purpose of this study was to evaluate the potential epidemiological trend of four infectious exanthematous diseases in Italy through the searches on the internet.

Methods

We used the following Italian search term: ‘Sesta malattia’ (Sixth Disease, in English), ‘Eritema Infettivo’ (also knows ‘Quinta malattia’ in Italian; Fifth Disease in English), ‘Quarta malattia’ (Fourth Disease in English) and ‘Scarlattina’ (Scarlet fever in English). We overlapped Google Trends and Wikipedia data to perform a linear regression and correlation analysis. Statistical analyses were performed using the Spearman’s rank correlation coefficient (rho). The study period is between July 2015 and December 2022

Results

The diseases considered have a seasonal trend and the search peaks between GT and Wikipedia overlap. A temporal correlation was observed between GT and Wikipedia search trends. Google Trends Internet search data showed strong correlation with Wikipedia with a rho statistically significant for Fifth disease (rho = 0.78), Fourth disease (rho = 0.76) and Scarlet-fever (rho = 0.77), moderate correlation for Sixth disease (rho = 0.32).

Conclusions

Infectious disease searches using Google and Wikipedia can be useful for public health surveillance and help policy makers implement prevention and information programs for the population, in addition to the fact that increases in searches could represent an early warning in the detection of outbreaks.

1. Introduction

Contagious exanthematous diseases (CEDs) are a group of infections characterized by skin rash (exanthem) and fever, caused by a variety of pathogens; including viruses, rickettsia or bacteria [1]. Among them, Measles, Scarlet Fever, Rubella, Fourth, Fifth and Sixth children’s Diseases are the most frequently diagnosed among children and young adults [1,2]. Despite the variability in terms of pathogens, common symptoms are shared, making appropriate diagnosis difficult. Moreover, clinical manifestations can range from mild to severe, with potentially serious complications when not promptly diagnosed and appropriately treated. In fact, even typically if benign and self-limited, severe and potentially life-threatening complications can be observed particularly in subgroups, such as infants, pregnant women, and immunocompromised individuals. Understanding the epidemiology, clinical presentation, pathogenesis, and management of CEDs is essential for prevention, diagnosis, and treatment.

In this contest, National Surveillance Systems (NSSs) are one of the main Public Health tools to monitor, control, and prevent the occurrence and spread of notifiable diseases [3]. In Italy, NSS was set up in 1990 (Ministerial Decree of 15 December 1990) with the aim of grouping notifiable infectious diseases in five classes, establishing times and ways of reporting for all probable and/or confirmed cases [4]. An efficient NSS is vital for public health in terms of outbreaks control and for saving lives and cost [5]. Nevertheless, most NSSs are affected by a certain level of delay (between cases occurrence and notification) and by a degree of underestimation. These two limits might impact on timely information and in underestimating the real epidemiological picture [6]. As an example, Herdman and colleagues found that Scarlet Fever was associated with a delayed diagnosis, especially during seasonal peaks and outbreaks, which results in delay notification and underestimation [7]. This intrinsic limitation of the traditional surveillance systems might be encompassed by taking advantage of technologies. In this perspective, previous evidence showed the applicability and performance of online epidemiological surveillance in supporting traditional surveillance systems [8]. Internet research volumes have been proven to be a very important source of data that could be used for trend analysis of various health-related topics [9]. Actually, they represent an invaluable source of clinical data capturing internet research activities on disease symptoms and medical treatments. They could be a helpful tool for public health surveillance and disease management [10]. This new approach to surveillance systems, also known as digital epidemiology, combine traditional surveillance data with Internet-based sources. By combining data on public’s interest with traditional surveillance, digital epidemiology is able to develop mathematical models that can deeply understand (and in some cases predict) outbreaks [11]. Infoveillance and infodemiology are related concepts that involve the monitoring and analysis of information, especially in the context of public health.

Infodemiology:

• Definition: Infodemiology refers to the science of distribution and determinants of information in an electronic medium, specifically the Internet, with the ultimate aim to inform public health and public policy.

• Focus: It focuses on studying patterns of information dissemination, user-generated content, and the dynamics of information flow related to health issues.

• Examples: Monitoring social media for trends related to health topics, analyzing search engine queries for health-related information, and studying online discussions to understand public perceptions of health issues.

Infoveillance:

• Definition: Infoveillance is the systematic collection, analysis, interpretation, and dissemination of health-related information from the internet, social media, and other sources for public health surveillance.

• Focus: It is a broader concept that encompasses the monitoring of various types of information sources, including traditional media, social media, blogs, and forums, to track and detect health events or trends.

Examples: Tracking disease outbreaks through online news articles, monitoring social media for reports of symptoms or disease clusters, and using online data to assess public sentiment and response to health interventions.

While there is a close relationship between infodemiology and infoveillance, the key difference lies in their focus. Infodemiology specifically looks at the distribution and determinants of information, especially on the Internet, while infoveillance is a more comprehensive term that includes the surveillance of various information sources for public health purposes.

Both concepts play crucial roles in leveraging digital information to enhance public health efforts, whether it’s monitoring the spread of diseases or understanding public perceptions and reactions to health-related events. The overlap between the two terms often occurs, and they are frequently used interchangeably in literature and discussions related to digital public health surveillance [12]. Considering the beneficial effects that can be envisioned in healthcare field (including timely and cost saving), digital epidemiology has become even more attractive; with a flourishing body of evidence explored different health related topic and internet research volume, mainly using research data generated using Wikipedia [13–15] or Google Trends [16,17].

In light of the above, and considering that despite the mandate to notify some CEDs, in Italy there are no data available regarding CEDs surveillance and epidemiology; we performed this cross-sectional analysis aimed to assess the potential epidemiological trend of notifiable CEDs in Italy through searches on the internet, assessing whether a correlation/association between users’ searches on CEDs in Google and Wikipedia exists.

2. Materials and methods

A cross-sectional study design was used. Data on Internet searches have been obtained from Google Trends (GT) based on Google Search, the most widely used internet search engine [18]. We used the following Italian search term: ‘Sesta malattia’ (Sixth Disease, in English), ‘Eritema Infettivo’ (also knows ‘Quinta malattia’ in Italian; Fifth Disease in English), ‘Quarta malattia’ (Fourth Disease in English) and ‘Scarlattina’ (Scarlet fever in English). One monthly time-frame elapsing has been extracted from July 2015 (since the beginning of Wikipedia coverage that has been used as a comparison) to December 2022. GT produces relative search volume (RSV) scaled to the highest search proportion week, which is computed as the percentage of queries concerning a particular term for a specific location and time period, where 100 is the maximum value and 0 is the minimum value. Thus, RSV allows for directly comparing search volume across search terms. From Wikipedia [19] it is possible to know how many times a specific page is viewed by users; data were extracted as monthly data corresponding to the monthly report of Google’s RSV. Using this tool, the number of monthly views by users from the July 2015 to December 2022 of the Wikipedia’ pages ‘Sesta malattia’ (Sixth disease, in English), ‘Eritema Infettivo’ (also knows ‘Quinta malattia’ in Italian; Fifth disease in English), ‘Quarta malattia’ (Fourth disease in English) and ‘Scarlattina’ (Scarlet fever in English) were extracted. The files in ‘.CSV’ format has been downloaded.

We overlapped Google Trends and Wikipedia data to perform a linear regression and correlation analysis. Cross-correlation results are obtained as product-moment correlations between the two-time series. The advantage of using cross-correlations is that it accounts for time dependence between two time-series variables. Statistical analyses were performed using the Spearman’s rank correlation coefficient (rho). According to a rule of thumb there is a strong correlation if rho > 0.7, moderate correlation if the value of rho is between 0.3 and 0.7 and weak correlation if rho < 0.3 [20]. A linear regression was performed considering Wikipedia searches as dependent variable and Google trends RSV as independent variable, results are expressed as coefficient with 95% confidence intervals (95% CI). Potential autocorrelation was ascertained through the calculation of the Durbin-Watson (DW) statistics. The DW test is a statistic test used to detect the presence of autocorrelation in the residuals (prediction errors) from a regression analysis [21]. The DW test statistic or d always lies between 0 and 4. If the d is substantially less than 2, there is evidence of positive serial correlation, while values greater than 2 suggest no autocorrelation. Representative linear model and correlation chart of the data were calculated, also calculating the R² of the model. The statistical significance level for the analyses was 0.05. The data were analyzed using the STATA statistical software, version 14 [22] and Microsoft Excel ®. The data download and analyses have been done the 13th of March, 2023.

3. Results

A temporal correlation was observed between GT and Wikipedia search trends. Google Trends Internet search data showed strong correlation with Wikipedia with a rho statistically significant for Fifth disease (rho = 0.78), Fourth disease (rho = 0.76) and Scarlet-fever (rho = 0.77), moderate correlation for Sixth disease (rho = 0.32) (see Table 1). Linear regression models show associations for the same disease that are statistically significant, and there is no autocorrelation as evidenced by the results of the Durbin Watson test (see Table 2). The raw data for GT and Wikipedia are shown in Figure 1. The search peaks for both Wikipedia and GT are in the same periods for the considering diseases. Figures 2 and 3 show the correlation chart between Wikipedia searches and Google’s RSV and linear regression between Wikipedia searches and Google’s RSV for the search terms. Figure 2 shows the correlation charts that represented the data, the model shows an R² equal to 0.1046 for Sixth disease, an R² equal to 0.5898 for Fifth disease, an R² equal to 0.5771 for Fourth disease and an R² equal to 0.5946 for Scarlet Fever. Figure 3 shows the linear regression that represented the data, the model shows an R² equal to 0.1384 for Sixth disease, an R² equal to 0.8079 for Fifth disease, an R² equal to 0.6395 for Fourth disease and an R² equal to 0.5904 for Scarlet Fever.

Table 1.

Spearman’s rank correlation coefficient.

	Google RSV	Wikipedia
Sesta malattia_Google RSV (Sixth disease)	1
Sesta malattia_Wikipedia (Sixth disease)	0.32**	1
Quinta malattia_Google RSV (Fifth disease)	1
Quinta malattia_Wikipedia (Fifth disease)	0.78*	1
Quarta malattia_Google RSV (Fourth disease)	1
Quarta malattia_Wikipedia (Fourth disease)	0.76*	1
Scarlattina_Google RSV (Scarlet Fever)	1
Scarlattina_Wikipedia (Scarlet Fever)	0.77*	1

*p-value<0.001, **p-value<0.005, Scarlattina: Scarlet Fever, Quarta malattia: Fourth disease, Quinta malattia: Fifth disease, Sesta malattia: Sixth disease.

Table 2.

Linear regression models.

Sesta malattia (Sixth disease)	Dependent variable: Wikipedia
Independent variable	Coefficient	95% CI	p-value	Durbin Watson
Google RSV	176.73	83.32–270.14	<0.001	0.169
Quinta malattia (Fifth disease)	Dependent variable: Wikipedia
Independent variable	Coefficient	95% CI	p-value	Durbin Watson
Google RSV	67.85	60.84–74.86	<0.001	0.424
Quarta malattia (Fourth disease)	Dependent variable: Wikipedia
Independent variable	Coefficient	95% CI	p-value	Durbin Watson
Google RSV	24.83	20.88–28.78	<0.001	1.245
Scarlattina (Scarlet Fever)	Dependent variable: Wikipedia
Independent variable	Coefficient	95% CI	p-value	Durbin Watson
Google RSV	395.13	325.41–464.84	<0.001	0.478

Figure 1.

Search trend of Wikipedia page and Google RSV.

Figure 2.

Correlation chart between Wikipedia searches and Google’s RSV. Spearman’s rank correlation coefficient was used. (a) Sixth disease, (b) fifth disease, (c) fourth disease, (d) scarlet fever.

Figure 3.

Linear regression between Wikipedia searches and Google’s RSV. (a) Sixth disease, (b) fifth disease, (c) fourth disease, (d) scarlet fever.

4. Discussion

In the current study we assessed the temporal correlation between GT and Wikipedia Italian search term associated with CEDs (including Sixth Disease, Fifth Disease, Fourth Disease, and Scarlet-fever). Our results show a strong-moderate correlation among the two search engines. Moreover, search volumes show the same temporal trends with spikes in the same periods for both Wikipedia and GT. Based on these results we can speculate that people who search on Wikipedia and GT are those who already received a clinical diagnosis from a physician. We assume that, only those who have already received a diagnosis are aware of the scientific name of the disease and, based on that information, are interested in seeking health-related information on the Internet. Actually, in this study we analyzed research volume directly related with the name of the disease, instead of the symptoms, as previous studies [14,17]. With this in mind, our results allow us to estimate the burden of diseases, particularly for those without surveillance system in place. In our view, this type of analysis allows for an assessment closer to the real number of cases compared to analyzing the research volumes related to disease symptoms. This is because of several reasons; firstly, infectious exanthematous diseases share similar symptoms, making differential diagnosis difficult for unqualified people. Secondly, infectious exanthematous diseases have nonspecific symptoms that could be attributable to many other diseases (even not infectious), potentially increasing background noise. These hypotheses are in line with the observation that the search volume for both Wikipedia and GT show a flattening of the curve in the period March 2020-July 2022. This interesting data should be interpretate in light of the COVID-19 pandemic. Actually, due to the application of diverse preventive measures (physical distancing, face mask, use of hydro-alcoholic solution, lock-down and so on) many infectious diseases (mainly air-borne diseases) dramatically drop down. Our data could suggest that, during COVID-19 pandemic, and thanks to the above-mentioned preventive measures, many cases of infectious exanthematous diseases have been avoided, consequently reducing the research volume generated. This result calls an important consideration. The reduced viral circulation (and consequently the estimated total number of cases) has potentially increased the number of people (especially newborns) immunologically virgins against these infectious exanthematous diseases. Consequently, an increment of cases could be expected in the future, and public health experts and physicians should be aware of the possible rebound effect. In light of this, infodemiological research might have important advances and many areas of application. Actually, previous evidence shows that infodemiological studies can reflect or anticipate the epidemiological characteristics of some disorders. On this line of research, we previously assessed the correlation between other infectious diseases (as influenza, pertussis, arboviruses) [12,14,15,23], chronic diseases (as cancers, and rheumatoid arthritis) [24,25] and GT and Wikipedia research volume, confirming the hypothesis.

The importance of our work is linked to the use of Internet research volume for assessing and estimating the burden of diseases not otherwise known. In fact, in previous studies we compared the Internet search volume with existing surveillance data, that does not exist for infectious exanthematous diseases. In this view, our pioneering study opens up further application areas of the infodemiological research. As a matter of fact, and to the best of our knowledge, this was the first study exploring whether GT and Wikipedia searches on infectious exanthematous diseases could represent a reliable proxy for the actual time course of these infections, in Italy. Of course, due to the pioneering nature of our study, results need to be interpreted with caution. Indeed, referring to our data, and solely for the purpose of providing an example, it is interesting to note the increase in search volume related to the sixth disease and scarlet fever. In both cases, there is an observed rise in search volume (from 2020 and throughout 2022, respectively). These data could be interpreted either in consideration of an actual increase in the circulation of the pathogen, indicating a real rise in cases, or conversely, it could be secondary to media (and non-media) events that may have influenced the general population’s interest in the topic, prompting users to search for information on the internet [24]. In this case, the absence of traditional surveillance systems for making comparisons makes it difficult to draw definite conclusions

Limitations of the study

Before to generalize, some limitations should be taken into account. Firstly, only GT and Wikipedia data were analyzed, not considering those generated from other search engines. However, 80% of worldwide internet users refers to Google [26]. Secondly, digital knowledge is not equally distributed among the population (because of age, level of education or social status), and for this reason our results might not be generalizable to the whole population. Third, as previously shown mediatic events might impact on Internet searches [24]. Fourth, in our study the data analyzed included the COVID-19 pandemic which may have influenced the quality and quantity of searches. Fifth, our models could not be adjusted for potential confounders due to the anonymity of the data. Nevertheless, as mentioned before, this is a pilot study that can offer a new point of view and new application of infodemiological data.

5. Conclusions

Given that to date digital epidemiology is not yet able to replace a classic surveillance system with this study we were able to trace the digital epidemiology of the diseases taken into consideration through research, the current trend of the research would suggest that, as happens in other countries with a temperate climate, the digital epidemiology of the disease is comparable to the real one [27,28]. The study described search trends and associated and correlated Google searches of the disease with those of Wikipedia. We found an atypical research trend during the period of restrictions due to the COVID-19 pandemic, probably due to the fact that being diseases transmitted by direct contact with mucus and saliva thanks to social distancing and the use of masks, the possibilities of contagion in that period were significantly reduced. Infectious disease searches using Google and Wikipedia can be useful for public health surveillance and help policy makers implement prevention and information programs for the population, in addition to the fact that increases in searches could represent an early warning in the detection of outbreaks.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Authors contribution

OES conceptualized, OES designed the study, OES performed data analysis and data extraction. SP, OES, VG wrote the first draft. All authors have read and agreed to the published version of the manuscript. All authors reviewed the manuscript.