Scoping review on search queries and social media for disease surveillance: a chronology of innovation

Virtual Zika transmission after the first U.S. case: who said what and how it spread on Twitter

BACKGROUND

This paper goes beyond detecting specific themes within Zika-related chatter on Twitter, to identify the key actors who influence the diffusive process through which some themes become more amplified than others.

METHODS

We collected all Zika-related tweets during the 3 months immediately after the first U.S. case of Zika. After the tweets were categorized into 12 themes, a cross-section were grouped into weekly datasets, to capture 12 amplifier/user groups, and analyzed by 4 amplification modes: mentions, retweets, talkers, and Twitter-wide amplifiers.

RESULTS

We analyzed 3,057,130 tweets in the United States and categorized 4997 users. The most talked about theme was Zika transmission (~58%). News media, public health institutions, and grassroots users were the most visible and frequent sources and disseminators of Zika-related Twitter content. Grassroots users were the primary sources and disseminators of conspiracy theories.

CONCLUSIONS

Social media analytics enable public health institutions to quickly learn what information is being disseminated, and by whom, regarding infectious diseases. Such information can help public health institutions identify and engage with news media and other active information providers. It also provides insights into media and public concerns, accuracy of information on Twitter, and information gaps. The study identifies implications for pandemic preparedness and response in the digital era and presents the agenda for future research and practice.

Data Science Priorities for a University Hospital-Based Institute of Infectious Diseases: A Viewpoint

Automation of laboratory tests, bioinformatic analysis of biological sequences, and professional data management are used routinely in a modern university hospital-based infectious diseases institute. This dates back to at least the 1980s. However, the scientific methods of this 21st century are changing with the increased power and speed of computers, with the "big data" revolution having already happened in genomics and environment, and eventually arriving in medical informatics. The research will be increasingly "data driven," and the powerful machine learning methods whose efficiency is demonstrated in daily life will also revolutionize medical research. A university-based institute of infectious diseases must therefore not only gather excellent computer scientists and statisticians (as in the past, and as in any medical discipline), but also fully integrate the biologists and clinicians with these computer scientists, statisticians, and mathematical modelers having a broad culture in machine learning, knowledge representation, and knowledge discovery.

Humanitarian health computing using artificial intelligence and social media: A narrative literature review

INTRODUCTION

According to the World Health Organization (WHO), over 130 million people are in constant need of humanitarian assistance due to natural disasters, disease outbreaks, and conflicts, among other factors. These health crises can compromise the resilience of healthcare systems, which are essential for achieving the health objectives of the sustainable development goals (SDGs) of the United Nations (UN). During a humanitarian health crisis, rapid and informed decision making is required. This is often challenging due to information scarcity, limited resources, and strict time constraints. Moreover, the traditional approach to digital health development, which involves a substantial requirement analysis, a feasibility study, and deployment of technology, is ill-suited for many crisis contexts. The emergence of Web 2.0 technologies and social media platforms in the past decade, such as Twitter, has created a new paradigm of massive information and misinformation, in which new technologies need to be developed to aid rapid decision making during humanitarian health crises.

OBJECTIVE

Humanitarian health crises increasingly require the analysis of massive amounts of information produced by different sources, such as social media content, and, hence, they are a prime case for the use of artificial intelligence (AI) techniques to help identify relevant information and make it actionable. To identify challenges and opportunities for using AI in humanitarian health crises, we reviewed the literature on the use of AI techniques to process social media.

METHODOLOGY

We performed a narrative literature review aimed at identifying examples of the use of AI in humanitarian health crises. Our search strategy was designed to get a broad overview of the different applications of AI in a humanitarian health crisis and their challenges. A total of 1459 articles were screened, and 24 articles were included in the final analysis.

RESULTS

Successful case studies of AI applications in a humanitarian health crisis have been reported, such as for outbreak detection. A commonly shared concern in the reviewed literature is the technical challenge of analyzing large amounts of data in real time. Data interoperability, which is essential to data sharing, is also a barrier with regard to the integration of online and traditional data sources. Human and organizational aspects that might be key factors for the adoption of AI and social media remain understudied. There is also a publication bias toward high-income countries, as we identified few examples in low-income countries. Further, we did not identify any examples of certain types of major crisis, such armed conflicts, in which misinformation might be more common.

CONCLUSIONS

The feasibility of using AI to extract valuable information during a humanitarian health crisis is proven in many cases. There is a lack of research on how to integrate the use of AI into the work-flow and large-scale deployments of humanitarian aid during a health crisis.

#Deathbedlive: the end-of-life trajectory, reflected in a cancer patient's tweets

Background

Understanding physical and psycho-social illness trajectories towards the end of life can help in the planning of palliative and supportive care. With terminal patients increasingly seeking and sharing health information and support via social media, it is timely to examine whether these trajectories are reflected in their digital narratives. In this exploratory study, we analysed the Twitter feed of prominent cancer sufferer and physician, Kate Granger, over the final 6 months of her life.

Methods

With the consent of Kate’s widower, Chris Pointon, 1628 Twitter posts from @GrangerKate were manually screened. The 550 tweets judged relevant to her disease were qualitatively content analysed with reference to the six modifiable dimensions of the patient experience in Emanuel and Emanuel’s ‘framework for a good death’. The frequency of each tweet category was charted over time and textual content was examined and cross-referenced with key events, to obtain a deeper understanding of its nature and significance.

Results

Tweets were associated with physical symptoms (N = 270), psychological and cognitive symptoms (N = 213), social relationships and support (N = 85), economic demands and care giving needs (N = 85), hopes and expectations (N = 51) and spiritual beliefs (N = 7). While medical treatments and procedures were discussed in detail, medical information-seeking was largely absent, likely reflecting Kate clinical expertise. Spirituality was expressed more as hope in treatments or “someone out there listening”, than in religious terms. The high value of Kate’s palliative care team was a dominant theme in the support category, alongside the support she received from her online community of fellow sufferers, friends, family and colleagues. Significant events, such as medical procedures and hospital stays generated the densest Twitter engagement. Transitions between trajectory phases were marked by changes in the relative frequency of tweet-types.

Conclusions

In Kate’s words, “the power of patient narrative cannot be underestimated”. While this analysis spanned only 6 months, it yielded rich insights. The results reflect theorised end-of-life dimensions and reveal the potential of social media data and digital bio-ethnography to shine a light on terminal patients’ lived experiences, coping strategies and support needs, suggesting new opportunities for enhancing personalised palliative care and avenues for further research.

HARNESSING SOCIAL MEDIA FOR HEALTH INFORMATION MANAGEMENT

The remarkable upsurge of social media has dramatic impacts on health care research and practice in the past decade. Social media are reshaping health information management in a variety of ways, ranging from providing cost-effective ways to improve clinician-patient communication and exchange health-related information and experience, to enabling the discovery of new medical knowledge and information. Despite some demonstrated initial success, social media use and analytics for improving health as a research field is still at its infancy. Information systems researchers can potentially play a key role in advancing the field. This study proposes a conceptual framework for social media-based health information management by drawing on multi-disciplinary research. With the guidance of the framework, this research presents related research challenges, identifies important yet under-explored research issues, and discusses promising directions for future research.

Discrepancies Between Classic and Digital Epidemiology in Searching for the Mayaro Virus: Preliminary Qualitative and Quantitative Analysis of Google Trends

Background

Mayaro virus (MAYV), first discovered in Trinidad in 1954, is spread by the Haemagogus mosquito. Small outbreaks have been described in the past in the Amazon jungles of Brazil and other parts of South America. Recently, a case was reported in rural Haiti.

Objective

Given the emerging importance of MAYV, we aimed to explore the feasibility of exploiting a Web-based tool for monitoring and tracking MAYV cases.

Methods

Google Trends is an online tracking system. A Google-based approach is particularly useful to monitor especially infectious diseases epidemics. We searched Google Trends from its inception (from January 2004 through to May 2017) for MAYV-related Web searches worldwide.

Results

We noted a burst in search volumes in the period from July 2016 (relative search volume [RSV]=13%) to December 2016 (RSV=18%), with a peak in September 2016 (RSV=100%). Before this burst, the average search activity related to MAYV was very low (median 1%). MAYV-related queries were concentrated in the Caribbean. Scientific interest from the research community and media coverage affected digital seeking behavior.

Conclusions

MAYV has always circulated in South America. Its recent appearance in the Caribbean has been a source of concern, which resulted in a burst of Internet queries. While Google Trends cannot be used to perform real-time epidemiological surveillance of MAYV, it can be exploited to capture the public’s reaction to outbreaks. Public health workers should be aware of this, in that information and communication technologies could be used to communicate with users, reassure them about their concerns, and to empower them in making decisions affecting their health.

Systematic review of surveillance by social media platforms for illicit drug use

Background

The use of social media (SM) as a surveillance tool of global illicit drug use is limited. To address this limitation, a systematic review of literature focused on the ability of SM to better recognize illicit drug use trends was addressed.

Methods

A search was conducted in databases: PubMed, CINAHL via Ebsco, PsychINFO via Ebsco, Medline via Ebsco, ERIC, Cochrane Library, Science Direct, ABI/INFORM Complete and Communication and Mass Media Complete. Included studies were original research published in peer-reviewed journals between January 2005 and June 2015 that primarily focused on collecting data from SM platforms to track trends in illicit drug use. Excluded were studies focused on purchasing prescription drugs from illicit online pharmacies.

Results

Selected studies used a range of SM tools/applications, including message boards, Twitter and blog/forums/platform discussions. Limitations included relevance, a lack of standardized surveillance systems and a lack of efficient algorithms to isolate relevant items.

Conclusion

Illicit drug use is a worldwide problem, and the rise of global social networking sites has led to the evolution of a readily accessible surveillance tool. Systematic approaches need to be developed to efficiently extract and analyze illicit drug content from social networks to supplement effective prevention programs.

Internet-based biosurveillance methods for vector-borne diseases: Are they novel public health tools or just novelties?

Internet-based surveillance methods for vector-borne diseases (VBDs) using "big data" sources such as Google, Twitter, and internet newswire scraping have recently been developed, yet reviews on such "digital disease detection" methods have focused on respiratory pathogens, particularly in high-income regions. Here, we present a narrative review of the literature that has examined the performance of internet-based biosurveillance for diseases caused by vector-borne viruses, parasites, and other pathogens, including Zika, dengue, other arthropod-borne viruses, malaria, leishmaniasis, and Lyme disease across a range of settings, including low- and middle-income countries. The fundamental features, advantages, and drawbacks of each internet big data source are presented for those with varying familiarity of "digital epidemiology." We conclude with some of the challenges and future directions in using internet-based biosurveillance for the surveillance and control of VBD.

[Patterns of access to information on protection against UV during the Brazilian summer: is there such a thing as the "summer effect"?]

Internet search patterns associated with "windows" of collective interest have been increasingly investigated in the field of public health. This article sets out to identify search patterns relating to the quest for information on skin protection after the perception of excessive exposure to UV radiation - the so-called "summer effect" as it is commonly referred to in Brazil. To calculate the number of hits on the Brazilian National Cancer Institute website - a renowned source of information resources on prevention - log analyzer software was used to measure the volume of hits on specific content pages. The pages on skin protection and self-examination (pages of interest) were monitored over a 48-month period. It was seen that, although the monthly average of hits on pages of interest revealed statistically significant annual growth, the results for the analysis of variance showed no significant differences between the number of hits in the summer compared with other months (p = 0.7491). In short, the perception of intense exposure to the summer sun did not encourage further interest to search for information on prevention.

A Systematic Review of the Use of Social Media for Food Safety Risk Communication

This article covers the current published literature related to the use of social media in food safety and infectious disease communication. The aim was to analyze literature recommendations and draw conclusions about how best to utilize social media for food safety risk communication going forward. A systematic literature review was conducted, and 24 articles were included for analysis. The inclusion criteria were (i) original peer-reviewed articles and (ii) primary focus on communication through social media about food safety and/or infectious diseases. Studies were coded for themes about social media applications, benefits, limitations, and best practices. Trust and personal beliefs were important drivers of social media use. The wide reach, immediacy, and information gathering capacities of social media were frequently cited benefits. Suggestions for social media best practices were inconsistent among studies, and study designs were highly variable. More evidence-based suggestions are needed to better establish guidelines for social media use in food safety and infectious disease risk communication. The information gleaned from this review can be used to create effective messages for shaping food safety behaviors.

Utility and potential of rapid epidemic intelligence from internet-based sources

OBJECTIVES

Rapid epidemic detection is an important objective of surveillance to enable timely intervention, but traditional validated surveillance data may not be available in the required timeframe for acute epidemic control. Increasing volumes of data on the Internet have prompted interest in methods that could use unstructured sources to enhance traditional disease surveillance and gain rapid epidemic intelligence. We aimed to summarise Internet-based methods that use freely-accessible, unstructured data for epidemic surveillance and explore their timeliness and accuracy outcomes.

METHODS

Steps outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist were used to guide a systematic review of research related to the use of informal or unstructured data by Internet-based intelligence methods for surveillance.

RESULTS

We identified 84 articles published between 2006-2016 relating to Internet-based public health surveillance methods. Studies used search queries, social media posts and approaches derived from existing Internet-based systems for early epidemic alerts and real-time monitoring. Most studies noted improved timeliness compared to official reporting, such as in the 2014 Ebola epidemic where epidemic alerts were generated first from ProMED-mail. Internet-based methods showed variable correlation strength with official datasets, with some methods showing reasonable accuracy.

CONCLUSION

The proliferation of publicly available information on the Internet provided a new avenue for epidemic intelligence. Methodologies have been developed to collect Internet data and some systems are already used to enhance the timeliness of traditional surveillance systems. To improve the utility of Internet-based systems, the key attributes of timeliness and data accuracy should be included in future evaluations of surveillance systems.

A novel use of social media to evaluate the occurrence of skin lesions affecting wild dusky grouper, Epinephelus marginatus (Lowe, 1834), in Libyan coastal waters

The social media network Facebook^™ was used to gather information on the occurrence and geographical distribution of dusky grouper dermatitis, a skin lesion affecting the dusky grouper, Epinephelus marginatus. Dusky grouper are common targets for spear fishermen in the Mediterranean and by monitoring spearfishing activity in Libyan waters, it was possible to document skin lesions from their entries on Facebook. Thirty-two Facebook accounts and 8 Facebook groups posting from 23 Libyan coastal cities provided a retrospective observational data set comprising a total of 382 images of dusky grouper caught by spearfishing between December 2011 and December 2015. Skin lesions were observable on 57/362 fish, for which images were of sufficient quality for analysis, giving a minimal prevalence for lesions of 15.75%. Only dusky grouper exceeding an estimated 40 cm total length exhibited lesions. The ability to collect useful data about the occurrence and geographical distribution of pathological conditions affecting wild fish using social media networks demonstrates their potential utility as a tool to support epidemiological studies and monitor the health of populations of aquatic animals. To our knowledge, this represents the first time that such an approach has been applied for assessing health in a wild population of fish.

Using Google Trends Data to Study Public Interest in Breast Cancer Screening in Brazil: Why Not a Pink February?

Background

One of the major challenges of the Brazilian Ministry of Health is to foster interest in breast cancer screening (BCS), especially among women at high risk. Strategies have been developed to promote the early identification of breast cancer mainly by Pink October campaigns. The massive number of queries conducted through Google creates traffic data that can be analyzed to show unrevealed interest cycles and their seasonalities.

Objectives

Using Google Trends, we studied cycles of public interest in queries toward mammography and breast cancer along the last 5 years. We hypothesize that these data may be correlated with collective interest cycles leveraged by national BCS campaigns such as Pink October.

Methods

Google Trends was employed to normalize traffic data on a scale from 0 (<1% of the peak volume) to 100 (peak of traffic) presented as weekly relative search volume (RSV) concerning mammography and breast cancer as search terms. A time series covered the last 261 weeks (November 2011 to October 2016), and RSV of both terms were compared with their respective annual means. Polynomial trendlines (second order) were employed to estimate overall trends.

Results

We found an upward trend for both terms over the 5 years, with almost parallel trendlines. Remarkable peaks were found along Pink October months— mammography and breast cancer searches were leveraged up reaching, respectively, 119.1% (2016) and 196.8% (2015) above annual means. Short downward RSVs along December-January months were also noteworthy along all the studied period. These trends traced an N-shaped pattern with higher peaks in Pink October months and sharp falls along subsequent December and January.

Conclusions

Considering these findings, it would be reasonable to bring Pink October to the beginning of each year, thereby extending the beneficial effect of the campaigns. It would be more appropriate to start screening campaigns at the beginning of the year, when new resolutions are taken and new projects are added to everyday routines. Our work raises attention to the study of traffic data to encourage health campaign analysts to undertake better analysis based on marketing practices.

Whiplash Syndrome Reloaded: Digital Echoes of Whiplash Syndrome in the European Internet Search Engine Context

Background

In many Western countries, after a motor vehicle collision, those involved seek health care for the assessment of injuries and for insurance documentation purposes. In contrast, in many less wealthy countries, there may be limited access to care and no insurance or compensation system.

Objective

The purpose of this infodemiology study was to investigate the global pattern of evolving Internet usage in countries with and without insurance and the corresponding compensation systems for whiplash injury.

Methods

We used the Internet search engine analytics via Google Trends to study the health information-seeking behavior concerning whiplash injury at national population levels in Europe.

Results

We found that the search for “whiplash” is strikingly and consistently often associated with the search for “compensation” in countries or cultures with a tort system. Frequent or traumatic painful injuries; diseases or disorders such as arthritis, headache, radius, and hip fracture; depressive disorders; and fibromyalgia were not associated similarly with searches on “compensation.”

Conclusions

In this study, we present evidence from the evolving viewpoint of naturalistic Internet search engine analytics that the expectations for receiving compensation may influence Internet search behavior in relation to whiplash injury.

The Potential of Social Media and Internet-Based Data in Preventing and Fighting Infectious Diseases: From Internet to Twitter

Health threats due to infectious diseases used to be a major public health concerns around the globe till mid of twentieth century when effective public health interventions helped in eradicating a number of infectious diseases around the world. Over the past 15 years, there has been a rise in the number of emerging and reemerging infectious diseases being reported such as the Acute Respiratory Syndrome (SARS) in 2002, HINI in 2009, Middle East Respiratory Syndrome (MERS) in 2012, Ebola in 2014, and Zika in 2016. These emerging viral infectious diseases have led to serious public health concerns leading to death and causing fear and anxiety among the public. More importantly, at the moment, the prevention and control of viral infectious diseases is difficult due to a lack of effective vaccines. Thus having real-time reporting tools are paramount to alert relevant public health surveillance systems and authorities about taking the right and necessary actions to control and minimize the potential harmful effects of viral infectious diseases. Social media and Internet-based data can play a major role in real-time reporting to empower active public health surveillance systems for controlling and fighting infectious diseases.

Interoperability of Information Systems Managed and Used by the Local Health Departments

BACKGROUND

In the post-Affordable Care Act era marked by interorganizational collaborations and availability of large amounts of electronic data from other community partners, it is imperative to assess the interoperability of information systems used by the local health departments (LHDs).

OBJECTIVES

To describe the level of interoperability of LHD information systems and identify factors associated with lack of interoperability.

DATA AND METHODS

This mixed-methods research uses data from the 2015 Informatics Capacity and Needs Assessment Survey, with a target population of all LHDs in the United States. A representative sample of 650 LHDs was drawn using a stratified random sampling design. A total of 324 completed responses were received (50% response rate). Qualitative data were used from a key informant interview study of LHD informatics staff from across the United States. Qualitative data were independently coded by 2 researchers and analyzed thematically. Survey data were cleaned, bivariate comparisons were conducted, and a multivariable logistic regression was run to characterize factors associated with interoperability.

RESULTS

For 30% of LHDs, no systems were interoperable, and 38% of LHD respondents indicated some of the systems were interoperable. Significant determinants of interoperability included LHDs having leadership support (adjusted odds ratio [AOR] = 3.54), control of information technology budget allocation (AOR = 2.48), control of data systems (AOR = 2.31), having a strategic plan for information systems (AOR = 1.92), and existence of business process analysis and redesign (AOR = 1.49).

CONCLUSION

Interoperability of all systems may be an informatics goal, but only a small proportion of LHDs reported having interoperable systems, pointing to a substantial need among LHDs nationwide.

Evaluating Google, Twitter, and Wikipedia as Tools for Influenza Surveillance Using Bayesian Change Point Analysis: A Comparative Analysis

Background

Traditional influenza surveillance relies on influenza-like illness (ILI) syndrome that is reported by health care providers. It primarily captures individuals who seek medical care and misses those who do not. Recently, Web-based data sources have been studied for application to public health surveillance, as there is a growing number of people who search, post, and tweet about their illnesses before seeking medical care. Existing research has shown some promise of using data from Google, Twitter, and Wikipedia to complement traditional surveillance for ILI. However, past studies have evaluated these Web-based sources individually or dually without comparing all 3 of them, and it would be beneficial to know which of the Web-based sources performs best in order to be considered to complement traditional methods.

Objective

The objective of this study is to comparatively analyze Google, Twitter, and Wikipedia by examining which best corresponds with Centers for Disease Control and Prevention (CDC) ILI data. It was hypothesized that Wikipedia will best correspond with CDC ILI data as previous research found it to be least influenced by high media coverage in comparison with Google and Twitter.

Methods

Publicly available, deidentified data were collected from the CDC, Google Flu Trends, HealthTweets, and Wikipedia for the 2012-2015 influenza seasons. Bayesian change point analysis was used to detect seasonal changes, or change points, in each of the data sources. Change points in Google, Twitter, and Wikipedia that occurred during the exact week, 1 preceding week, or 1 week after the CDC’s change points were compared with the CDC data as the gold standard. All analyses were conducted using the R package “bcp” version 4.0.0 in RStudio version 0.99.484 (RStudio Inc). In addition, sensitivity and positive predictive values (PPV) were calculated for Google, Twitter, and Wikipedia.

Results

During the 2012-2015 influenza seasons, a high sensitivity of 92% was found for Google, whereas the PPV for Google was 85%. A low sensitivity of 50% was calculated for Twitter; a low PPV of 43% was found for Twitter also. Wikipedia had the lowest sensitivity of 33% and lowest PPV of 40%.

Conclusions

Of the 3 Web-based sources, Google had the best combination of sensitivity and PPV in detecting Bayesian change points in influenza-related data streams. Findings demonstrated that change points in Google, Twitter, and Wikipedia data occasionally aligned well with change points captured in CDC ILI data, yet these sources did not detect all changes in CDC data and should be further studied and developed.

Disease Monitoring and Health Campaign Evaluation Using Google Search Activities for HIV and AIDS, Stroke, Colorectal Cancer, and Marijuana Use in Canada: A Retrospective Observational Study

Background

Infodemiology can offer practical and feasible health research applications through the practice of studying information available on the Web. Google Trends provides publicly accessible information regarding search behaviors in a population, which may be studied and used for health campaign evaluation and disease monitoring. Additional studies examining the use and effectiveness of Google Trends for these purposes remain warranted.

Objective

The objective of our study was to explore the use of infodemiology in the context of health campaign evaluation and chronic disease monitoring. It was hypothesized that following a launch of a campaign, there would be an increase in information seeking behavior on the Web. Second, increasing and decreasing disease patterns in a population would be associated with search activity patterns. This study examined 4 different diseases: human immunodeficiency virus (HIV) infection, stroke, colorectal cancer, and marijuana use.

Methods

Using Google Trends, relative search volume data were collected throughout the period of February 2004 to January 2015. Campaign information and disease statistics were obtained from governmental publications. Search activity trends were graphed and assessed with disease trends and the campaign interval. Pearson product correlation statistics and joinpoint methodology analyses were used to determine significance.

Results

Disease patterns and online activity across all 4 diseases were significantly correlated: HIV infection (r=.36, P<.001), stroke (r=.40, P<.001), colorectal cancer (r= −.41, P<.001), and substance use (r=.64, P<.001). Visual inspection and the joinpoint analysis showed significant correlations for the campaigns on colorectal cancer and marijuana use in stimulating search activity. No significant correlations were observed for the campaigns on stroke and HIV regarding search activity.

Conclusions

The use of infoveillance shows promise as an alternative and inexpensive solution to disease surveillance and health campaign evaluation. Further research is needed to understand Google Trends as a valid and reliable tool for health research.

High correlation of Middle East respiratory syndrome spread with Google search and Twitter trends in Korea

The Middle East respiratory syndrome coronavirus (MERS-CoV) was exported to Korea in 2015, resulting in a threat to neighboring nations. We evaluated the possibility of using a digital surveillance system based on web searches and social media data to monitor this MERS outbreak. We collected the number of daily laboratory-confirmed MERS cases and quarantined cases from May 11, 2015 to June 26, 2015 using the Korean government MERS portal. The daily trends observed via Google search and Twitter during the same time period were also ascertained using Google Trends and Topsy. Correlations among the data were then examined using Spearman correlation analysis. We found high correlations (>0.7) between Google search and Twitter results and the number of confirmed MERS cases for the previous three days using only four simple keywords: "MERS", "" ("MERS (in Korean)"), "" ("MERS symptoms (in Korean)"), and "" ("MERS hospital (in Korean)"). Additionally, we found high correlations between the Google search and Twitter results and the number of quarantined cases using the above keywords. This study demonstrates the possibility of using a digital surveillance system to monitor the outbreak of MERS.

Chinese Public Attention to the Outbreak of Ebola in West Africa: Evidence from the Online Big Data Platform

Objective: The outbreak of the Ebola epidemic in West Africa in 2014 exerted enormous global public reaction via the Internet and social media. This study aimed to investigate and evaluate the public reaction to Ebola in China and identify the primitive correlation between possible influence factors caused by the outbreak of Ebola in West Africa and Chinese public attention via Internet surveillance. Methods: Baidu Index (BDI) and Sina Micro Index (SMI) were collected from their official websites, and the disease-related data were recorded from the websites of the World Health Organization (WHO), U.S. Centers for Disease Control and Prevention (CDC), and U.S. National Ministries of Health. The average BDI of Internet users in different regions were calculated to identify the public reaction to the Ebola outbreak. Spearman’s rank correlation was used to check the relationship of epidemic trends with BDI and SMI. Additionally, spatio-temporal analysis and autocorrelation analysis were performed to detect the clustered areas with the high attention to the topic of “Ebola”. The related news reports were collected from authoritative websites to identify potential patterns. Results: The BDI and the SMI for “Ebola” showed a similar fluctuating trend with a correlation coefficient = 0.9 (p < 0.05). The average BDI in Beijing, Tibet, and Shanghai was higher than other cities. However, the disease-related indicators did not identify potential correlation with both indices above. A hotspot area was detected in Tibet by local autocorrelation analysis. The most likely cluster identified by spatiotemporal cluster analysis was in the northeast regions of China with the relative risk (RR) of 2.26 (p ≤ 0.01) from 30 July to 14 August in 2014. Qualitative analysis indicated that negative news could lead to a continuous increase of the public’s attention until the appearance of a positive news report. Conclusions: Confronted with the risk of cross-border transmission of the infectious disease, online surveillance might be used as an innovative approach to perform public communication and health education through examining the public’s reaction and attitude.

Correlation between National Influenza Surveillance Data and Search Queries from Mobile Devices and Desktops in South Korea

BACKGROUND

Digital surveillance using internet search queries can improve both the sensitivity and timeliness of the detection of a health event, such as an influenza outbreak. While it has recently been estimated that the mobile search volume surpasses the desktop search volume and mobile search patterns differ from desktop search patterns, the previous digital surveillance systems did not distinguish mobile and desktop search queries. The purpose of this study was to compare the performance of mobile and desktop search queries in terms of digital influenza surveillance.

METHODS AND RESULTS

The study period was from September 6, 2010 through August 30, 2014, which consisted of four epidemiological years. Influenza-like illness (ILI) and virologic surveillance data from the Korea Centers for Disease Control and Prevention were used. A total of 210 combined queries from our previous survey work were used for this study. Mobile and desktop weekly search data were extracted from Naver, which is the largest search engine in Korea. Spearman's correlation analysis was used to examine the correlation of the mobile and desktop data with ILI and virologic data in Korea. We also performed lag correlation analysis. We observed that the influenza surveillance performance of mobile search queries matched or exceeded that of desktop search queries over time. The mean correlation coefficients of mobile search queries and the number of queries with an r-value of ≥ 0.7 equaled or became greater than those of desktop searches over the four epidemiological years. A lag correlation analysis of up to two weeks showed similar trends.

CONCLUSION

Our study shows that mobile search queries for influenza surveillance have equaled or even become greater than desktop search queries over time. In the future development of influenza surveillance using search queries, the recognition of changing trend of mobile search data could be necessary.

Estimating Influenza Outbreaks Using Both Search Engine Query Data and Social Media Data in South Korea

BACKGROUND

As suggested as early as in 2006, logs of queries submitted to search engines seeking information could be a source for detection of emerging influenza epidemics if changes in the volume of search queries are monitored (infodemiology). However, selecting queries that are most likely to be associated with influenza epidemics is a particular challenge when it comes to generating better predictions.

OBJECTIVE

In this study, we describe a methodological extension for detecting influenza outbreaks using search query data; we provide a new approach for query selection through the exploration of contextual information gleaned from social media data. Additionally, we evaluate whether it is possible to use these queries for monitoring and predicting influenza epidemics in South Korea.

METHODS

Our study was based on freely available weekly influenza incidence data and query data originating from the search engine on the Korean website Daum between April 3, 2011 and April 5, 2014. To select queries related to influenza epidemics, several approaches were applied: (1) exploring influenza-related words in social media data, (2) identifying the chief concerns related to influenza, and (3) using Web query recommendations. Optimal feature selection by least absolute shrinkage and selection operator (Lasso) and support vector machine for regression (SVR) were used to construct a model predicting influenza epidemics.

RESULTS

In total, 146 queries related to influenza were generated through our initial query selection approach. A considerable proportion of optimal features for final models were derived from queries with reference to the social media data. The SVR model performed well: the prediction values were highly correlated with the recent observed influenza-like illness (r=.956; P<.001) and virological incidence rate (r=.963; P<.001).

CONCLUSIONS

These results demonstrate the feasibility of using search queries to enhance influenza surveillance in South Korea. In addition, an approach for query selection using social media data seems ideal for supporting influenza surveillance based on search query data.

The wisdom of crowds in action: Forecasting epidemic diseases with a web-based prediction market system

BACKGROUND

The quest for an effective system capable of monitoring and predicting the trends of epidemic diseases is a critical issue for communities worldwide. With the prevalence of Internet access, more and more researchers today are using data from both search engines and social media to improve the prediction accuracy. In particular, a prediction market system (PMS) exploits the wisdom of crowds on the Internet to effectively accomplish relatively high accuracy.

OBJECTIVE

This study presents the architecture of a PMS and demonstrates the matching mechanism of logarithmic market scoring rules. The system was implemented to predict infectious diseases in Taiwan with the wisdom of crowds in order to improve the accuracy of epidemic forecasting.

METHODS

The PMS architecture contains three design components: database clusters, market engine, and Web applications. The system accumulated knowledge from 126 health professionals for 31 weeks to predict five disease indicators: the confirmed cases of dengue fever, the confirmed cases of severe and complicated influenza, the rate of enterovirus infections, the rate of influenza-like illnesses, and the confirmed cases of severe and complicated enterovirus infection.

RESULTS

Based on the winning ratio, the PMS predicts the trends of three out of five disease indicators more accurately than does the existing system that uses the five-year average values of historical data for the same weeks. In addition, the PMS with the matching mechanism of logarithmic market scoring rules is easy to understand for health professionals and applicable to predict all the five disease indicators.

CONCLUSIONS

The PMS architecture of this study affords organizations and individuals to implement it for various purposes in our society. The system can continuously update the data and improve prediction accuracy in monitoring and forecasting the trends of epidemic diseases. Future researchers could replicate and apply the PMS demonstrated in this study to more infectious diseases and wider geographical areas, especially the under-developed countries across Asia and Africa.

"Googling" for Cancer: An Infodemiological Assessment of Online Search Interests in Australia, Canada, New Zealand, the United Kingdom, and the United States

Background

The infodemiological analysis of queries from search engines to shed light on the status of various noncommunicable diseases has gained increasing popularity in recent years.

Objective

The aim of the study was to determine the international perspective on the distribution of information seeking in Google regarding “cancer” in major English-speaking countries.

Methods

We used Google Trends service to assess people’s interest in searching about “Cancer” classified as “Disease,” from January 2004 to December 2015 in Australia, Canada, New Zealand, the United Kingdom, and the United States. Then, we evaluated top cities and their relative search volumes (SVs) and country-specific “Top searches” and “Rising searches.” We also evaluated the cross-country correlations of SVs for cancer, as well as rank correlations of SVs from 2010 to 2014 with the incidence of cancer in 2012 in the abovementioned countries.

Results

From 2004 to 2015, the United States (relative SV [from 100]: 63), Canada (62), and Australia (61) were the top countries searching for cancer in Google, followed by New Zealand (54) and the United Kingdom (48). There was a consistent seasonality pattern in searching for cancer in the United States, Canada, Australia, and New Zealand. Baltimore (United States), St John’s (Canada), Sydney (Australia), Otaika (New Zealand), and Saint Albans (United Kingdom) had the highest search interest in their corresponding countries. “Breast cancer” was the cancer entity that consistently appeared high in the list of top searches in all 5 countries. The “Rising searches” were “pancreatic cancer” in Canada and “ovarian cancer” in New Zealand. Cross-correlation of SVs was strong between the United States, Canada, and Australia (>.70, P<.01).

Conclusions

Cancer maintained its popularity as a search term for people in the United States, Canada, and Australia, comparably higher than New Zealand and the United Kingdom. The increased interest in searching for keywords related to cancer shows the possible effectiveness of awareness campaigns in increasing societal demand for health information on the Web, to be met in community-wide communication or awareness interventions.

A novel web informatics approach for automated surveillance of cancer mortality trends

Cancer surveillance data are collected every year in the United States via the National Program of Cancer Registries (NPCR) and the Surveillance, Epidemiology and End Results (SEER) Program of the National Cancer Institute (NCI). General trends are closely monitored to measure the nation's progress against cancer. The objective of this study was to apply a novel web informatics approach for enabling fully automated monitoring of cancer mortality trends. The approach involves automated collection and text mining of online obituaries to derive the age distribution, geospatial, and temporal trends of cancer deaths in the US. Using breast and lung cancer as examples, we mined 23,850 cancer-related and 413,024 general online obituaries spanning the timeframe 2008-2012. There was high correlation between the web-derived mortality trends and the official surveillance statistics reported by NCI with respect to the age distribution (ρ=0.981 for breast; ρ=0.994 for lung), the geospatial distribution (ρ=0.939 for breast; ρ=0.881 for lung), and the annual rates of cancer deaths (ρ=0.661 for breast; ρ=0.839 for lung). Additional experiments investigated the effect of sample size on the consistency of the web-based findings. Overall, our study findings support web informatics as a promising, cost-effective way to dynamically monitor spatiotemporal cancer mortality trends.

Social media in Ebola outbreak

The West African 2014 Ebola outbreak has highlighted the need for a better information network. Hybrid information networks, an integration of both hierarchical and formalized command control-driven and community-based, or ad hoc emerging networks, could assist in improving public health responses. By filling the missing gaps with social media use, the public health response could be more proactive rather than reactive in responding to such an outbreak of global concern. This article provides a review of the current social media use specifically in this outbreak by systematically collecting data from ProQuest Newsstand, Dow Jones Factiva, Program for Monitoring Emerging Diseases (ProMED) as well as Google Trends. The period studied is from 19 March 2014 (first request for information on ProMED) to 15 October 2014, a total of 31 weeks. The term 'Ebola' was used in the search for media reports. The outcome of the review shows positive results for social media use in effective surveillance response mechanisms - for improving the detection, preparedness and response of the outbreak - as a complement to traditional, filed, work-based surveillance approach.

The utility of web mining for epidemiological research: studying the association between parity and cancer risk

BACKGROUND

The World Wide Web has emerged as a powerful data source for epidemiological studies related to infectious disease surveillance. However, its potential for cancer-related epidemiological discoveries is largely unexplored.

METHODS

Using advanced web crawling and tailored information extraction procedures, the authors automatically collected and analyzed the text content of 79 394 online obituary articles published between 1998 and 2014. The collected data included 51 911 cancer (27 330 breast; 9470 lung; 6496 pancreatic; 6342 ovarian; 2273 colon) and 27 483 non-cancer cases. With the derived information, the authors replicated a case-control study design to investigate the association between parity (i.e., childbearing) and cancer risk. Age-adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for each cancer type and compared to those reported in large-scale epidemiological studies.

RESULTS

Parity was found to be associated with a significantly reduced risk of breast cancer (OR = 0.78, 95% CI, 0.75-0.82), pancreatic cancer (OR = 0.78, 95% CI, 0.72-0.83), colon cancer (OR = 0.67, 95% CI, 0.60-0.74), and ovarian cancer (OR = 0.58, 95% CI, 0.54-0.62). Marginal association was found for lung cancer risk (OR = 0.87, 95% CI, 0.81-0.92). The linear trend between increased parity and reduced cancer risk was dramatically more pronounced for breast and ovarian cancer than the other cancers included in the analysis.

CONCLUSION

This large web-mining study on parity and cancer risk produced findings very similar to those reported with traditional observational studies. It may be used as a promising strategy to generate study hypotheses for guiding and prioritizing future epidemiological studies.

Forecasting the Incidence of Dementia and Dementia-Related Outpatient Visits With Google Trends: Evidence From Taiwan

Background

Google Trends has demonstrated the capability to both monitor and predict epidemic outbreaks. The connection between Internet searches for dementia information and dementia incidence and dementia-related outpatient visits remains unknown.

Objective

This study aimed to determine whether Google Trends could provide insight into trends in dementia incidence and related outpatient visits in Taiwan. We investigated and validated the local search terms that would be the best predictors of new dementia cases and outpatient visits. We further evaluated the nowcasting (ie, forecasting the present) and forecasting effects of Google Trends search trends for new dementia cases and outpatient visits. The long-term goal is to develop a surveillance system to help early detection and interventions for dementia in Taiwan.

Methods

This study collected (1) dementia data from Taiwan’s National Health Insurance Research Database and (2) local Internet search data from Google Trends, both from January 2009 to December 2011. We investigated and validated search terms that would be the best predictors of new dementia cases and outpatient visits. We then evaluated both the nowcasting and the forecasting effects of Google Trends search trends through cross-correlation analysis of the dementia incidence and outpatient visit data with the Google Trends data.

Results

The search term “dementia + Alzheimer’s disease” demonstrated a 3-month lead effect for new dementia cases and a 6-month lead effect for outpatient visits (r=.503, P=.002; r=.431, P=.009, respectively). When gender was included in the analysis, the search term “dementia” showed 6-month predictive power for new female dementia cases (r=.520, P=.001), but only a nowcasting effect for male cases (r=.430, P=.009). The search term “neurology” demonstrated a 3-month leading effect for new dementia cases (r=.433, P=.008), for new male dementia cases (r=.434, P=.008), and for outpatient visits (r=.613, P<.001).

Conclusions

Google Trends established a plausible relationship between search terms and new dementia cases and dementia-related outpatient visits in Taiwan. This data may allow the health care system in Taiwan to prepare for upcoming outpatient and dementia screening visits. In addition, the validated search term results can be used to provide caregivers with caregiving-related health, skills, and social welfare information by embedding dementia-related search keywords in relevant online articles.

Using Social Media for Actionable Disease Surveillance and Outbreak Management: A Systematic Literature Review

OBJECTIVE

Research studies show that social media may be valuable tools in the disease surveillance toolkit used for improving public health professionals' ability to detect disease outbreaks faster than traditional methods and to enhance outbreak response. A social media work group, consisting of surveillance practitioners, academic researchers, and other subject matter experts convened by the International Society for Disease Surveillance, conducted a systematic primary literature review using the PRISMA framework to identify research, published through February 2013, answering either of the following questions: Can social media be integrated into disease surveillance practice and outbreak management to support and improve public health?Can social media be used to effectively target populations, specifically vulnerable populations, to test an intervention and interact with a community to improve health outcomes?Examples of social media included are Facebook, MySpace, microblogs (e.g., Twitter), blogs, and discussion forums. For Question 1, 33 manuscripts were identified, starting in 2009 with topics on Influenza-like Illnesses (n = 15), Infectious Diseases (n = 6), Non-infectious Diseases (n = 4), Medication and Vaccines (n = 3), and Other (n = 5). For Question 2, 32 manuscripts were identified, the first in 2000 with topics on Health Risk Behaviors (n = 10), Infectious Diseases (n = 3), Non-infectious Diseases (n = 9), and Other (n = 10).

CONCLUSIONS

The literature on the use of social media to support public health practice has identified many gaps and biases in current knowledge. Despite the potential for success identified in exploratory studies, there are limited studies on interventions and little use of social media in practice. However, information gleaned from the articles demonstrates the effectiveness of social media in supporting and improving public health and in identifying target populations for intervention. A primary recommendation resulting from the review is to identify opportunities that enable public health professionals to integrate social media analytics into disease surveillance and outbreak management practice.

MACVIA-ARIA Sentinel NetworK for allergic rhinitis (MASK-rhinitis): the new generation guideline implementation

Several unmet needs have been identified in allergic rhinitis: identification of the time of onset of the pollen season, optimal control of rhinitis and comorbidities, patient stratification, multidisciplinary team for integrated care pathways, innovation in clinical trials and, above all, patient empowerment. MASK-rhinitis (MACVIA-ARIA Sentinel NetworK for allergic rhinitis) is a simple system centred around the patient which was devised to fill many of these gaps using Information and Communications Technology (ICT) tools and a clinical decision support system (CDSS) based on the most widely used guideline in allergic rhinitis and its asthma comorbidity (ARIA 2015 revision). It is one of the implementation systems of Action Plan B3 of the European Innovation Partnership on Active and Healthy Ageing (EIP on AHA). Three tools are used for the electronic monitoring of allergic diseases: a cell phone-based daily visual analogue scale (VAS) assessment of disease control, CARAT (Control of Allergic Rhinitis and Asthma Test) and e-Allergy screening (premedical system of early diagnosis of allergy and asthma based on online tools). These tools are combined with a clinical decision support system (CDSS) and are available in many languages. An e-CRF and an e-learning tool complete MASK. MASK is flexible and other tools can be added. It appears to be an advanced, global and integrated ICT answer for many unmet needs in allergic diseases which will improve policies and standards.

Big Data and the Global Public Health Intelligence Network (GPHIN)

Background

Globalization and the potential for rapid spread of emerging infectious diseases have heightened the need for ongoing surveillance and early detection. The Global Public Health Intelligence Network (GPHIN) was established to increase situational awareness and capacity for the early detection of emerging public health events.

Objective

To describe how the GPHIN has used Big Data as an effective early detection technique for infectious disease outbreaks worldwide and to identify potential future directions for the GPHIN.

Findings

Every day the GPHIN analyzes over more than 20,000 online news reports (over 30,000 sources) in nine languages worldwide. A web-based program aggregates data based on an algorithm that provides potential signals of emerging public health events which are then reviewed by a multilingual, multidisciplinary team. An alert is sent out if a potential risk is identified. This process proved useful during the Severe Acute Respiratory Syndrome (SARS) outbreak and was adopted shortly after by a number of countries to meet new International Health Regulations that require each country to have the capacity for early detection and reporting. The GPHIN identified the early SARS outbreak in China, was credited with the first alert on MERS-CoV and has played a significant role in the monitoring of the Ebola outbreak in West Africa. Future developments are being considered to advance the GPHIN's capacity in light of other Big Data sources such as social media and its analytical capacity in terms of algorithm development.

Conclusion

The GPHIN's early adoption of Big Data has increased global capacity to detect international infectious disease outbreaks and other public health events. Integration of additional Big Data sources and advances in analytical capacity could further strengthen the GPHIN's capability for timely detection and early warning.

Using Twitter to Measure Public Discussion of Diseases: A Case Study

Background

Twitter is increasingly used to estimate disease prevalence, but such measurements can be biased, due to both biased sampling and inherent ambiguity of natural language.

Objective

We characterized the extent of these biases and how they vary with disease.

Methods

We correlated self-reported prevalence rates for 22 diseases from Experian’s Simmons National Consumer Study (n=12,305) with the number of times these diseases were mentioned on Twitter during the same period (2012). We also identified and corrected for two types of bias present in Twitter data: (1) demographic variance between US Twitter users and the general US population; and (2) natural language ambiguity, which creates the possibility that mention of a disease name may not actually refer to the disease (eg, “heart attack” on Twitter often does not refer to myocardial infarction). We measured the correlation between disease prevalence and Twitter disease mentions both with and without bias correction. This allowed us to quantify each disease’s overrepresentation or underrepresentation on Twitter, relative to its prevalence.

Results

Our sample included 80,680,449 tweets. Adjusting disease prevalence to correct for Twitter demographics more than doubles the correlation between Twitter disease mentions and disease prevalence in the general population (from .113 to .258, P <.001). In addition, diseases varied widely in how often mentions of their names on Twitter actually referred to the diseases, from 14.89% (3827/25,704) of instances (for stroke) to 99.92% (5044/5048) of instances (for arthritis). Applying ambiguity correction to our Twitter corpus achieves a correlation between disease mentions and prevalence of .208 ( P <.001). Simultaneously applying correction for both demographics and ambiguity more than triples the baseline correlation to .366 ( P <.001). Compared with prevalence rates, cancer appeared most overrepresented in Twitter, whereas high cholesterol appeared most underrepresented.

Conclusions

Twitter is a potentially useful tool to measure public interest in and concerns about different diseases, but when comparing diseases, improvements can be made by adjusting for population demographics and word ambiguity.

Age-related differences in the accuracy of web query-based predictions of influenza-like illness

BACKGROUND

Web queries are now widely used for modeling, nowcasting and forecasting influenza-like illness (ILI). However, given that ILI attack rates vary significantly across ages, in terms of both magnitude and timing, little is known about whether the association between ILI morbidity and ILI-related queries is comparable across different age-groups. The present study aimed to investigate features of the association between ILI morbidity and ILI-related query volume from the perspective of age.

METHODS

Since Google Flu Trends is unavailable in Italy, Google Trends was used to identify entry terms that correlated highly with official ILI surveillance data. All-age and age-class-specific modeling was performed by means of linear models with generalized least-square estimation. Hold-out validation was used to quantify prediction accuracy. For purposes of comparison, predictions generated by exponential smoothing were computed.

RESULTS

Five search terms showed high correlation coefficients of > .6. In comparison with exponential smoothing, the all-age query-based model correctly predicted the peak time and yielded a higher correlation coefficient with observed ILI morbidity (.978 vs. .929). However, query-based prediction of ILI morbidity was associated with a greater error. Age-class-specific query-based models varied significantly in terms of prediction accuracy. In the 0-4 and 25-44-year age-groups, these did well and outperformed exponential smoothing predictions; in the 15-24 and ≥ 65-year age-classes, however, the query-based models were inaccurate and highly overestimated peak height. In all but one age-class, peak timing predicted by the query-based models coincided with observed timing.

CONCLUSIONS

The accuracy of web query-based models in predicting ILI morbidity rates could differ among ages. Greater age-specific detail may be useful in flu query-based studies in order to account for age-specific features of the epidemiology of ILI.

Resilient information networks for coordination of foodborne disease outbreaks

Foodborne disease outbreaks are increasingly being seen as a greater concern by public health authorities. It has also become a global research agenda to identify improved pathways to coordinating outbreak detection. Furthermore, a significant need exists for timely coordination of the detection of potential foodborne disease outbreaks to reduce the number of infected individuals and the overall impact on public health security. This study aimed to offer an effective approach for coordinating foodborne disease outbreaks. First, we identify current coordination processes, complexities, and challenges. We then explore social media surveillance strategies, usage, and the power of these strategies to influence decision-making. Finally, based on informal (social media) and formal (organizational) surveillance approaches, we propose a hybrid information network model for improving the coordination of outbreak detection.

Social media in adolescent health literacy education: a pilot study

Background

While health literacy has gained notice on a global stage, the initial focus on seeking associations with medical conditions may have overlooked its impact across generations. Adolescent health literacy, specifically in dentistry, is an underexplored area despite the significance of this formative stage on an individual’s approach to healthy lifestyles and behaviors.

Objective

The aim is to conduct a pilot study to evaluate the efficacy of three major social media outlets - Twitter, Facebook, and YouTube - in supporting adolescents’ oral health literacy (OHL) education.

Methods

A random sample of 22 adolescents (aged 14-16 years) from an English-medium international school in Hong Kong provided informed consent. Sociodemographic information, including English language background, social media usage, and dental experience were collected via a questionnaire. A pre- and post-test of OHL (REALD-30) was administered by two trained, calibrated examiners. Following pre-test, participants were randomly assigned to one of three social media outlets: Twitter, Facebook, or YouTube. Participants received alerts posted daily for 5 consecutive days requiring online accessing of modified and original OHL education materials. One-way ANOVA ( analysis of variance) was used to compare the mean difference between the pre- and the post-test results among the three social media.

Results

No associations were found between the social media allocated and participants’ sociodemographics, including English language background, social media usage, and dental experience. Of the three social media, significant differences in literacy assessment scores were evident for participants who received oral health education messages via Facebook (P=.02) and YouTube (P=.005).

Conclusions

Based on the results of the pilot study, Facebook and YouTube may be more efficient media outlets for OHL promotion and education among adolescent school children when compared to Twitter. Further analyses with a larger study group is warranted.

Cumulative query method for influenza surveillance using search engine data

BACKGROUND

Internet search queries have become an important data source in syndromic surveillance system. However, there is currently no syndromic surveillance system using Internet search query data in South Korea.

OBJECTIVES

The objective of this study was to examine correlations between our cumulative query method and national influenza surveillance data.

METHODS

Our study was based on the local search engine, Daum (approximately 25% market share), and influenza-like illness (ILI) data from the Korea Centers for Disease Control and Prevention. A quota sampling survey was conducted with 200 participants to obtain popular queries. We divided the study period into two sets: Set 1 (the 2009/10 epidemiological year for development set 1 and 2010/11 for validation set 1) and Set 2 (2010/11 for development Set 2 and 2011/12 for validation Set 2). Pearson's correlation coefficients were calculated between the Daum data and the ILI data for the development set. We selected the combined queries for which the correlation coefficients were .7 or higher and listed them in descending order. Then, we created a cumulative query method n representing the number of cumulative combined queries in descending order of the correlation coefficient.

RESULTS

In validation set 1, 13 cumulative query methods were applied, and 8 had higher correlation coefficients (min=.916, max=.943) than that of the highest single combined query. Further, 11 of 13 cumulative query methods had an r value of ≥.7, but 4 of 13 combined queries had an r value of ≥.7. In validation set 2, 8 of 15 cumulative query methods showed higher correlation coefficients (min=.975, max=.987) than that of the highest single combined query. All 15 cumulative query methods had an r value of ≥.7, but 6 of 15 combined queries had an r value of ≥.7.

CONCLUSIONS

Cumulative query method showed relatively higher correlation with national influenza surveillance data than combined queries in the development and validation set.

MoSAIC: Mobile Surveillance for Acute Respiratory Infections and Influenza-Like Illness in the Community

Surveillance for acute respiratory infection (ARI) and influenza-like illness (ILI) relies primarily on reports of medically attended illness. Community surveillance could mitigate delays in reporting, allow for timely collection of respiratory tract samples, and characterize cases of non–medically attended ILI representing substantial personal and economic burden. Text messaging could be utilized to perform longitudinal ILI surveillance in a community-based sample but has not been assessed. We recruited 161 households (789 people) in New York City for a study of mobile ARI/ILI surveillance, and selected reporters received text messages twice weekly inquiring whether anyone in the household was ill. Home visits were conducted to obtain nasal swabs from persons with ARI/ILI. Participants were primarily female, Latino, and publicly insured. During the 44-week period from December 2012 through September 2013, 11,282 text messages were sent. In responses to 8,250 (73.1%) messages, a household reported either that someone was ill or no one was ill; 88.9% of responses were received within 4 hours. Swabs were obtained for 361 of 363 reported ARI/ILI episodes. The median time from symptom onset to nasal swab was 2 days; 65.4% of samples were positive for a respiratory pathogen by reverse-transcriptase polymerase chain reaction. In summary, text messaging promoted rapid ARI/ILI reporting and specimen collection and could represent a promising approach to timely, community-based surveillance.

A case study of the New York City 2012-2013 influenza season with daily geocoded Twitter data from temporal and spatiotemporal perspectives

BACKGROUND

Twitter has shown some usefulness in predicting influenza cases on a weekly basis in multiple countries and on different geographic scales. Recently, Broniatowski and colleagues suggested Twitter's relevance at the city-level for New York City. Here, we look to dive deeper into the case of New York City by analyzing daily Twitter data from temporal and spatiotemporal perspectives. Also, through manual coding of all tweets, we look to gain qualitative insights that can help direct future automated searches.

OBJECTIVE

The intent of the study was first to validate the temporal predictive strength of daily Twitter data for influenza-like illness emergency department (ILI-ED) visits during the New York City 2012-2013 influenza season against other available and established datasets (Google search query, or GSQ), and second, to examine the spatial distribution and the spread of geocoded tweets as proxies for potential cases.

METHODS

From the Twitter Streaming API, 2972 tweets were collected in the New York City region matching the keywords "flu", "influenza", "gripe", and "high fever". The tweets were categorized according to the scheme developed by Lamb et al. A new fourth category was added as an evaluator guess for the probability of the subject(s) being sick to account for strength of confidence in the validity of the statement. Temporal correlations were made for tweets against daily ILI-ED visits and daily GSQ volume. The best models were used for linear regression for forecasting ILI visits. A weighted, retrospective Poisson model with SaTScan software (n=1484), and vector map were used for spatiotemporal analysis.

RESULTS

Infection-related tweets (R=.763) correlated better than GSQ time series (R=.683) for the same keywords and had a lower mean average percent error (8.4 vs 11.8) for ILI-ED visit prediction in January, the most volatile month of flu. SaTScan identified primary outbreak cluster of high-probability infection tweets with a 2.74 relative risk ratio compared to medium-probability infection tweets at P=.001 in Northern Brooklyn, in a radius that includes Barclay's Center and the Atlantic Avenue Terminal.

CONCLUSIONS

While others have looked at weekly regional tweets, this study is the first to stress test Twitter for daily city-level data for New York City. Extraction of personal testimonies of infection-related tweets suggests Twitter's strength both qualitatively and quantitatively for ILI-ED prediction compared to alternative daily datasets mixed with awareness-based data such as GSQ. Additionally, granular Twitter data provide important spatiotemporal insights. A tweet vector-map may be useful for visualization of city-level spread when local gold standard data are otherwise unavailable.

Online reports of foodborne illness capture foods implicated in official foodborne outbreak reports

OBJECTIVE

Traditional surveillance systems capture only a fraction of the estimated 48 million yearly cases of foodborne illness in the United States. We assessed whether foodservice reviews on Yelp.com (a business review site) can be used to support foodborne illness surveillance efforts.

METHODS

We obtained reviews from 2005 to 2012 of 5824 foodservice businesses closest to 29 colleges. After extracting recent reviews describing episodes of foodborne illness, we compared implicated foods to foods in outbreak reports from the U.S. Centers for Disease Control and Prevention (CDC).

RESULTS

Broadly, the distribution of implicated foods across five categories was as follows: aquatic (16% Yelp, 12% CDC), dairy-eggs (23% Yelp, 23% CDC), fruits-nuts (7% Yelp, 7% CDC), meat-poultry (32% Yelp, 33% CDC), and vegetables (22% Yelp, 25% CDC). The distribution of foods across 19 more specific food categories was also similar, with Spearman correlations ranging from 0.60 to 0.85 for 2006-2011. The most implicated food categories in both Yelp and CDC were beef, dairy, grains-beans, poultry and vine-stalk.

CONCLUSIONS

Based on observations in this study and the increased usage of social media, we posit that online illness reports could complement traditional surveillance systems by providing near real-time information on foodborne illnesses, implicated foods and locations.

Recommendations for the role of social science research in One Health

The social environment has changed rapidly as technology has facilitated communication among individuals and groups in ways not imagined 20 years ago. Communication technology increasingly plays a role in decision-making about health and environmental behaviors and is being leveraged to influence that process. But at its root is the fundamental need to understand human cognition, communication, and behavior. The concept of 'One Health' has emerged as a framework for interdisciplinary work that cuts across human, animal, and ecosystem health in recognition of their interdependence and the value of an integrated perspective. Yet, the science of communication, information studies, social psychology, and other social sciences have remained marginalized in this emergence. Based on an interdisciplinary collaboration, this paper reports on a nascent conceptual framework for the role of social science in 'One Health' issues and identifies a series of recommendations for research directions that bear additional scrutiny and development.

Big data in medicine is driving big changes

OBJECTIVES

To summarise current research that takes advantage of "Big Data" in health and biomedical informatics applications.

METHODS

Survey of trends in this work, and exploration of literature describing how large-scale structured and unstructured data sources are being used to support applications from clinical decision making and health policy, to drug design and pharmacovigilance, and further to systems biology and genetics.

RESULTS

The survey highlights ongoing development of powerful new methods for turning that large-scale, and often complex, data into information that provides new insights into human health, in a range of different areas. Consideration of this body of work identifies several important paradigm shifts that are facilitated by Big Data resources and methods: in clinical and translational research, from hypothesis-driven research to data-driven research, and in medicine, from evidence-based practice to practice-based evidence.

CONCLUSIONS

The increasing scale and availability of large quantities of health data require strategies for data management, data linkage, and data integration beyond the limits of many existing information systems, and substantial effort is underway to meet those needs. As our ability to make sense of that data improves, the value of the data will continue to increase. Health systems, genetics and genomics, population and public health; all areas of biomedicine stand to benefit from Big Data and the associated technologies.

Demand-based web surveillance of sexually transmitted infections in Russia

OBJECTIVES

To investigate the possibility of using HIV- and syphilis-related web queries to predict incident diagnosis rates of sexually transmitted infections in Russia.

METHODS

The regional volume of HIV/syphilis queries, normalized to the total number of queries submitted to the most popular search engine, was used to predict the notification rates of HIV/syphilis in each region by applying both global non-spatial and spatial statistics.

RESULTS

Nationwide, both search volumes and regional HIV/syphilis diagnosis rates were positively spatially auto-correlated, indicating a clustered pattern of spatial distribution. A high positive correlation between notification rates and search volume was observed. Compared with linear models, spatially explicit geographically weighted models adjusted for broadband Internet diffusion proved superior in predicting the regional level of the HIV/syphilis epidemic on the basis of their search volume.

CONCLUSIONS

Timeliness, easy availability, low cost, and transparency make HIV- and syphilis-related web queries a promising addition to traditional methods of disease surveillance in Russia. Geographically weighted regression provides useful insights, as it is able to capture the spatial heterogeneity of the relationship between search volume and disease incidence.

Comparative analysis of online health queries originating from personal computers and smart devices on a consumer health information portal

Background

The number of people using the Internet and mobile/smart devices for health information seeking is increasing rapidly. Although the user experience for online health information seeking varies with the device used, for example, smart devices (SDs) like smartphones/tablets versus personal computers (PCs) like desktops/laptops, very few studies have investigated how online health information seeking behavior (OHISB) may differ by device.

Objective

The objective of this study is to examine differences in OHISB between PCs and SDs through a comparative analysis of large-scale health search queries submitted through Web search engines from both types of devices.

Methods

Using the Web analytics tool, IBM NetInsight OnDemand, and based on the type of devices used (PCs or SDs), we obtained the most frequent health search queries between June 2011 and May 2013 that were submitted on Web search engines and directed users to the Mayo Clinic’s consumer health information website. We performed analyses on “Queries with considering repetition counts (QwR)” and “Queries without considering repetition counts (QwoR)”. The dataset contains (1) 2.74 million and 3.94 million QwoR, respectively for PCs and SDs, and (2) more than 100 million QwR for both PCs and SDs. We analyzed structural properties of the queries (length of the search queries, usage of query operators and special characters in health queries), types of search queries (keyword-based, wh-questions, yes/no questions), categorization of the queries based on health categories and information mentioned in the queries (gender, age-groups, temporal references), misspellings in the health queries, and the linguistic structure of the health queries.

Results

Query strings used for health information searching via PCs and SDs differ by almost 50%. The most searched health categories are “Symptoms” (1 in 3 search queries), “Causes”, and “Treatments & Drugs”. The distribution of search queries for different health categories differs with the device used for the search. Health queries tend to be longer and more specific than general search queries. Health queries from SDs are longer and have slightly fewer spelling mistakes than those from PCs. Users specify words related to women and children more often than that of men and any other age group. Most of the health queries are formulated using keywords; the second-most common are wh- and yes/no questions. Users ask more health questions using SDs than PCs. Almost all health queries have at least one noun and health queries from SDs are more descriptive than those from PCs.

Conclusions

This study is a large-scale comparative analysis of health search queries to understand the effects of device type (PCs vs SDs) used on OHISB. The study indicates that the device used for online health information search plays an important role in shaping how health information searches by consumers and patients are executed.

Analysing Twitter and web queries for flu trend prediction

Background

Social media platforms encourage people to share diverse aspects of their daily life. Among these, shared health related information might be used to infer health status and incidence rates for specific conditions or symptoms. In this work, we present an infodemiology study that evaluates the use of Twitter messages and search engine query logs to estimate and predict the incidence rate of influenza like illness in Portugal.

Results

Based on a manually classified dataset of 2704 tweets from Portugal, we selected a set of 650 textual features to train a Naïve Bayes classifier to identify tweets mentioning flu or flu-like illness or symptoms. We obtained a precision of 0.78 and an F-measure of 0.83, based on cross validation over the complete annotated set. Furthermore, we trained a multiple linear regression model to estimate the health-monitoring data from the Influenzanet project, using as predictors the relative frequencies obtained from the tweet classification results and from query logs, and achieved a correlation ratio of 0.89 (p < 0.001). These classification and regression models were also applied to estimate the flu incidence in the following flu season, achieving a correlation of 0.72.

Conclusions

Previous studies addressing the estimation of disease incidence based on user-generated content have mostly focused on the english language. Our results further validate those studies and show that by changing the initial steps of data preprocessing and feature extraction and selection, the proposed approaches can be adapted to other languages. Additionally, we investigated whether the predictive model created can be applied to data from the subsequent flu season. In this case, although the prediction result was good, an initial phase to adapt the regression model could be necessary to achieve more robust results.

Stroke experiences in weblogs: a feasibility study of sex differences

Background

Research on cerebral stroke symptoms using hospital records has reported that women experience more nontraditional symptoms of stroke (eg, mental status change, pain) than men do. This is an important issue because nontraditional symptoms may delay the decision to get medical assistance and increase the difficulty of correct diagnosis. In the present study, we investigate sex differences in the stroke experience as described in stories on weblogs.

Objective

The goal of this study was to investigate the feasibility of using the Internet as a source of data for basic research on stroke experiences.

Methods

Stroke experiences described in blogs were identified by using StoryUpgrade, a program that searches blog posts using a fictional prototype story. In this study, the prototype story was a description of a stroke experience. Retrieved stories coded by the researchers as relevant were used to update the search query and retrieve more stories using relevance feedback. Stories were coded for first- or third-person narrator, traditional and nontraditional patient symptoms, type of stroke, patient sex and age, delay before seeking medical assistance, and delay at hospital and in treatment.

Results

There were 191 relevant stroke stories of which 174 stories reported symptoms (52.3% female and 47.7% male patients). There were no sex differences for each traditional or nontraditional stroke symptom by chi-square analysis (all Ps>.05). Type of narrator, however, affected report of traditional and nontraditional symptoms. Female first-person narrators (ie, the patient) were more likely to report mental status change (56.3%, 27/48) than male first-person narrators (36.4%, 16/44), a marginally significant effect by logistic regression (P=.056), whereas reports of third-person narrators did not differ for women (27.9%, 12/43) and men (28.2%, 11/39) patients. There were more reports of at least 1 nontraditional symptom in the 92 first-person reports (44.6%, 41/92) than in the 82 third-person reports (25.6%, 21/82, P=.006). Ischemic or hemorrhagic stroke was reported in 67 and 29 stories, respectively. Nontraditional symptoms varied with stroke type with 1 or more nontraditional symptoms reported for 79.3% (23/29) of hemorrhagic stroke patients and 53.7% (36/67) of ischemic stroke patients (P=.001).

Conclusions

The results replicate previous findings based on hospital interview data supporting the reliability of findings from weblogs. New findings include the effect of first- versus third-person narrator on sex differences in the report of nontraditional symptoms. This result suggests that narrator is an important variable to be examined in future studies. A fragmentary data problem limits some conclusions because important information, such as age, was not consistently reported. Age trends strengthen the feasibility of using the Internet for stroke research because older adults have significantly increased their Internet use in recent years.

Guess who's not coming to dinner? Evaluating online restaurant reservations for disease surveillance

Background

Alternative data sources are used increasingly to augment traditional public health surveillance systems. Examples include over-the-counter medication sales and school absenteeism.

Objective

We sought to determine if an increase in restaurant table availabilities was associated with an increase in disease incidence, specifically influenza-like illness (ILI).

Methods

Restaurant table availability was monitored using OpenTable, an online restaurant table reservation site. A daily search was performed for restaurants with available tables for 2 at the hour and at half past the hour for 22 distinct times: between 11:00 am-3:30 pm for lunch and between 6:00-11:30 PM for dinner. In the United States, we examined table availability for restaurants in Boston, Atlanta, Baltimore, and Miami. For Mexico, we studied table availabilities in Cancun, Mexico City, Puebla, Monterrey, and Guadalajara. Time series of restaurant use was compared with Google Flu Trends and ILI at the state and national levels for the United States and Mexico using the cross-correlation function.

Results

Differences in restaurant use were observed across sampling times and regions. We also noted similarities in time series trends between data on influenza activity and restaurant use. In some settings, significant correlations greater than 70% were noted between data on restaurant use and ILI trends.

Conclusions

This study introduces and demonstrates the potential value of restaurant use data for event surveillance.

Electronic word of mouth on twitter about physical activity in the United States: exploratory infodemiology study

Background

Twitter is a widely used social medium. However, its application in promoting health behaviors is understudied.

Objective

In order to provide insights into designing health marketing interventions to promote physical activity on Twitter, this exploratory infodemiology study applied both social cognitive theory and the path model of online word of mouth to examine the distribution of different electronic word of mouth (eWOM) characteristics among personal tweets about physical activity in the United States.

Methods

This study used 113 keywords to retrieve 1 million public tweets about physical activity in the United States posted between January 1 and March 31, 2011. A total of 30,000 tweets were randomly selected and sorted based on numbers generated by a random number generator. Two coders scanned the first 16,100 tweets and yielded 4672 (29.02%) tweets that they both agreed to be about physical activity and were from personal accounts. Finally, 1500 tweets were randomly selected from the 4672 tweets (32.11%) for further coding. After intercoder reliability scores reached satisfactory levels in the pilot coding (100 tweets separate from the final 1500 tweets), 2 coders coded 750 tweets each. Descriptive analyses, Mann-Whitney U tests, and Fisher exact tests were performed.

Results

Tweets about physical activity were dominated by neutral sentiments (1270/1500, 84.67%). Providing opinions or information regarding physical activity (1464/1500, 97.60%) and chatting about physical activity (1354/1500, 90.27%) were found to be popular on Twitter. Approximately 60% (905/1500, 60.33%) of the tweets demonstrated users’ past or current participation in physical activity or intentions to participate in physical activity. However, social support about physical activity was provided in less than 10% of the tweets (135/1500, 9.00%). Users with fewer people following their tweets (followers) (P=.02) and with fewer accounts that they followed (followings) (P=.04) were more likely to talk positively about physical activity on Twitter. People with more followers were more likely to post neutral tweets about physical activity (P=.04). People with more followings were more likely to forward tweets (P=.04). People with larger differences between number of followers and followings were more likely to mention companionship support for physical activity on Twitter (P=.04).

Conclusions

Future health marketing interventions promoting physical activity should segment Twitter users based on their number of followers, followings, and gaps between the number of followers and followings. The innovative application of both marketing and public health theory to examine tweets about physical activity could be extended to other infodemiology or infoveillance studies on other health behaviors (eg, vaccinations).

Internet search patterns of human immunodeficiency virus and the digital divide in the Russian Federation: infoveillance study

Background

Human immunodeficiency virus (HIV) is a serious health problem in the Russian Federation. However, the true scale of HIV in Russia has long been the subject of considerable debate. Using digital surveillance to monitor diseases has become increasingly popular in high income countries. But Internet users may not be representative of overall populations, and the characteristics of the Internet-using population cannot be directly ascertained from search pattern data. This exploratory infoveillance study examined if Internet search patterns can be used for disease surveillance in a large middle-income country with a dispersed population.

Objective

This study had two main objectives: (1) to validate Internet search patterns against national HIV prevalence data, and (2) to investigate the relationship between search patterns and the determinants of Internet access.

Methods

We first assessed whether online surveillance is a valid and reliable method for monitoring HIV in the Russian Federation. Yandex and Google both provided tools to study search patterns in the Russian Federation. We evaluated the relationship between both Yandex and Google aggregated search patterns and HIV prevalence in 2011 at national and regional tiers. Second, we analyzed the determinants of Internet access to determine the extent to which they explained regional variations in searches for the Russian terms for “HIV” and “AIDS”. We sought to extend understanding of the characteristics of Internet searching populations by data matching the determinants of Internet access (age, education, income, broadband access price, and urbanization ratios) and searches for the term “HIV” using principal component analysis (PCA).

Results

We found generally strong correlations between HIV prevalence and searches for the terms “HIV” and “AIDS”. National correlations for Yandex searches for “HIV” were very strongly correlated with HIV prevalence (Spearman rank-order coefficient [rs]=.881, P≤.001) and strongly correlated for “AIDS” (rs=.714, P≤.001). The strength of correlations varied across Russian regions. National correlations in Google for the term “HIV” (rs=.672, P=.004) and “AIDS” (rs=.584, P≤.001) were weaker than for Yandex. Second, we examined the relationship between the determinants of Internet access and search patterns for the term “HIV” across Russia using PCA. At the national level, we found Principal Component 1 loadings, including age (-0.56), HIV search (-0.533), and education (-0.479) contributed 32% of the variance. Principal Component 2 contributed 22% of national variance (income, -0.652 and broadband price, -0.460).

Conclusions

This study contributes to the methodological literature on search patterns in public health. Based on our preliminary research, we suggest that PCA may be used to evaluate the relationship between the determinants of Internet access and searches for health problems beyond high-income countries. We believe it is in middle-income countries that search methods can make the greatest contribution to public health.