Digital Epidemiology and How It Transform Public Health
Sun, 31 Mar 2019 || By Anggika Rahmadiani

Along with technology innovation, the scientific methodologies in population studies has transformed itself. Big data has enormously enriched data availability and methods in data collection, allow researchers to do digital tracing in their data collection instead of relying solely on the traditional population studies method. Public health is one of the sectors that exposed to this auspicious innovation through digital epidemiology. By benefiting from the availability of extensive data set, public health claimed to improve health surveillance and detection of disease outbreaks. This commentary will try to expand the discussion of digital epidemiology from its methodological concern into the privacy consequences.

 

Digital Epidemiology and Its Exemplary Cases

Epidemiology alone has a purpose of comprehending the population health condition and disease pattern, which later can be utilized to make detection and preventive strategy. The use of digital trace, range from algorithm analysis of social media usage or any internet-based behavior, that initially not generated for epidemical purposes gave birth to the digital epidemiology[1]. The method went mainstream after US Centers for Disease Control and Prevention (CDC), and Google started to use online search query to detecting influenza breakout around the United States in 2009[2]. Here are two cases on how digital epidemiology has contributed epistemically to the scientific progression.    

The research from Max Planck Institute for Demographic Research has concluded a hypothetical relation that fertility measures based on Facebook data are highly correlated with the mean age of childbearing (MAC) data as conventional indicators based on traditional data from UNDESA and UNICEF[3]. By extracting data from Facebook Advertising Platform that consists of society group age from 15-49 years old who had a child in the last 12 months, this research able to generate the parenthood related-data. The significant scientific contribution is this research successfully fill the gap of incomplete male fertility data in 79 developing countries, a data set that UNDESA and UNICEF once did not have.

Besides social media, search engine machine also played a remarkable role in executing more rapid epidemiological surveillance. Current research from Universitas Gadjah Mada showed how Google Trends data can help Indonesia as a hyperendemic country for dengue to expanding their disease early detection and monitoring capacity[4]. By comparing the traditional data surveillance from the Indonesia Ministry of Health and keyword search as information-seeking behavior pattern in 2012-2016, the research can detect the dengue case one to three months earlier.[5]

 

Is it a prospective method for future public health?

Digital epidemiology came up as a promising method that focused on the individualized-based information with a rapid process and resulted a large data set, compared to the traditional epidemiology which relied on the hypothesis-driven information[6]. Xihong Lin, Chair of Biostatistics at Harvard emphasized how digital epidemiology has enabled researcher access of vast data set with rich details in a short, even real-time,[7]  ranged from geolocations as the most general information to individual biogenome or phenomes. As Medtech industries continue to innovate in producing more health-related devices like fitness trackers and its consumers are currently emerging in our digital society[8], the individual health behavior data and pattern from these IoTs can also expand the possibilities of more complex digital epidemiology research.

Contrary to the previous statement, in its scientific debate, digital epidemiology rises an epistemic question on how this digital method significantly brings any changes to the studies. From statistical studies' point of view, they conclude that there is no significant change that digital epidemiology brings to the studies because it still stagnantly stay in the data-driven method[9] (Hohle, 2017:13). From the epistemic concern before, what can be done to improve digital epidemiology is to encourage public health institution in generating their digital data streams without dependently rely on the internet corporations' data.[10] The exemplary work came from HealthMap, a digital information system about global infectious disease threat that acquired data from public health media sources.  

On the other side, digital epidemiology started to emerge when human behavior towards disease transformed in a digital age. It also raises the ethical debate on whether personal health data sharing from online behavior can lead to another privacy breach. Upholding the context-sensitive regulation[11] moreover, anonymizing the personal health data used in digital epidemiological research is urgently needed as it is the main an ethical ground for the research that will be used for the public goods interest in population health, which makes it differs from any user-generated content companies' business interest. 

Editor: Anisa Pratita Mantovani

Read another article written by Anggika Rahmadiani

 

[1] Salathé, M. (2017). ‘Digital epidemiology: what is it, and where is it going?’ Life Sciences, Society and Policy 14 (1) pp 1-5.

[2] Ginsberg, J., Mohebbi, Matthew H., Patel, Rajan S., Brammer, L., Smolinski, Mark S., Brilliant, L. (2009). ‘Detecting influenza epidemics using search engine query data’ Nature 457 pp 1102-1104

[3] Rampazzo, F., Zagheni, E., Weber, I., Rita Testa, M., Billari, F. (2018) ‘Mater Certa Est, Pater Numquam: What Can Facebook Advertising Data Tell Us about Male Fertility Rates?’ Proceedings of the Twelfth  International AAAI Conference on Web and Social Media 2018 pp. 672-675

[4] Husnayain, A., Fuad, A., & Lazuardi, L. (2019) ‘Correlation between Google Trends on dengue fever and national surveillance report in Indonesia’, Global Health Action, 12 (1) DOI:10.1080/16549716.2018.1552652

[5] Husnaiyain, A & Anis, F. (2019) Menggunakan Google Trends untuk dukung sistem monitoring demam berdarah, bagaimana caranya? [Online] The Conversation Available at https://theconversation.com/menggunakan-google-trends-untuk-dukung-sistem-monitoring-demam-berdarah-bagaimana-caranya-110278. Accessed by 14 March 2018

[6] Velasco E. (2018) ‘Disease detection, epidemiology and outbreak response: the digital future of public health practice’ Life Science Society Policy. 14(1):7

[7] Lin, X. (2015) Off the cuff: It’s the data first, hypotheses second: Harvard Public Health [Online] Available at https://www.hsph.harvard.edu/magazine/magazine_article/off-the-cuff-its-the-data-first-hypotheses-second/. Accessed 14 March 2019

[8] Evans, J., Licking, E., Hillenbach, J., Spence, P. (2018) Pulse of the industry 2018: When the human body is the most significant data platform, how will Medtech companies capture value? [Online] Ernest and Young. Available at https://www.ey.com/Publication/vwLUAssets/ey-pulse-of-the-industry-2018/$FILE/ey-pulse-of-the-industry-2018.pdf Accessed by 15 March 2019

[9] Höhle, M. (2017) ‘A statistician's perspective on digital epidemiology.' Life Sciences, Society and Policy 13 (17) DOI 10.1186/s40504-017-0063-9

[10] Salathé, M. (2017). ‘Digital epidemiology: what is it, and where is it going?’ Life Sciences, Society and Policy 14 (1) pp 1-5.

[11] Vayena, E., Salathé, M., Madoff, L., Brownstein, J. (2015) ‘Ethical Challenges of Big Data in Public Health’ PLoS Computational Biology 11(2)