Humans can leverage the power of data to combat epidemics such as the coronavirus outbreak that has put the world in siege during the first quarter of 2020.
Epidemiology is the branch of medicine which deals with the incidence, distribution, and possible control of diseases and other factors relating to health. It plays a key role in public health, informing decisions and policies and assisting evidence-based practice by identifying risk factors for disease and targets for preventive healthcare.
Two important activities in epidemiology are disease surveillance and outbreak investigation and response. In disease surveillance, epidemiologists collect data from different levels of health in routine data collection and case reporting for specific diseases. In outbreak investigation and management, epidemiologists report on suspect and confirmed cases, do contact tracing to proactively discover possible transmissions, and map the transmission chain to identify the source of the outbreak and make recommendations to control it.
How is data managed in the current epidemic?
In today’s epidemic a lot of the data is still collected on paper forms, delaying the extraction of conclusion. Below you can see forms used by CDC.
WHO and JHU are doing a great work at collecting, compiling and communicating data through their dashboards, which are used by other health organisations, by the media, and by public health authorities to make decisions and to keep the population informed.
Mapping cases is key in epidemiology, as it facilitates the identification of possible transmission chains and planning health intervention. The use of geographic information systems (GIS) gives public health teams a competitive advantage against disease, but today most of the data is collected, stored and analysed in silos, causing repeated work across organisations that could save days of work through collaboration and data sharing agreements.
If all the data currently held by health systems was digitalised and shared amongst the different health systems fighting coronavirus, an earlier detection of cases, better contact tracing and faster intervention in the suspect cases would be possible.
Where does my faith in data for epidemic control come from?
Ebola response. Liberia, 2015.
In 2015 I worked in West Africa as part of the Ebola Response. My tiny contribution was to manage a a project that consisted on developing, rolling out and improving an app to help health care workers keep track of Ebola cases and their contacts, as well as to survey the contacts for the complete length of the incubation period to assess if they had any indications of contracting the disease. What I learnt about disease transmission and epidemiology back then resonates a lot today, in the middle of another pandemic that has managed to affect countries in the developed and developing world alike. Apart from hygiene measures, avoiding crowds and limiting contacts, the early reporting of symptoms… the importance of data in the early detection and control of disease outbreaks is again top of my mind these days.
Heath Information Systems. Guinea, 2015-2016
By the end of 2015 I went back to West Africa, this time to support the design and implementation of a national health information system for the Ministry of Health in Guinea. We were challenged by a lack of agreement on the list of clinics and hospitals, the uncertainty about the number of people who lived in each city and the absence of official means of identification for patients that resulted in duplicated or triplicated medical records for individuals without a link that would allow the detection of chronic diseases or comorbidity. These, together with a young data culture, meant no zero reporting, delays in data transmission and other challenges.
Sleeping Sickness. DRC, 2015-2016.
All through 2015 and 2016 I was also involved in a project in the Democratic Republic of Congo, where we wanted to digitalise historical data and improve current data collection to end with sleeping sickness, a disease transmitted by the tse-tse fly that is close to being eradicated and could benefit greatly from data to better target the latest efforts.
In 2016 I left eHealth Africa, discovered chatbots and started working on a project called bots4health and, specifically, on Eva, a chatbot to help humans have healthier lives through conversations. I believed chatbots were a great help to collect data from humans and to share health messages with them. Messaging channels, combined with natural language processing, looked like a good solution for a lot of the problems I had seen during my work in Africa. My faith in this kind of conversation remains, that’s why I am training Eva to be able to respond to questions related to coronavirus and to send daily updates to users that opt-in to alerts.
What do others say about the potential of data in disease prevention and control?
What would happen if all health institutions took this task as seriously? Would it have been possible to prevent the epidemic instead of responding to it? I want to invite you to reflect on this thought with me, and to support my argument I’d like to share 5 online talks around the topic of humans, epidemics, and data.
How social media can predict epidemics by Nicholas Christakis
In his talk, “How social media can predict epidemics”, Nicholas Christakis, explains how data related to networks and the identification of the use of certain keywords related to diseases or symptoms, can be used to predict a flu epidemic before it happens – or the spread of an idea.
How data can predict the next pandemic by Adam Kucharski
Adam Kucharski explains how tracking patterns of human contact, such as hugs, we can actually predict how a pandemic will spread. With enough data, perhaps this could be a viable tool for preventing the spread of diseases?
Infectious Disease Forecasting… For The World by Dr. James Wilson
In this talk Dr. James Wilson reveals the real methods already in place for the forecasting and early detection and intervention of infectious diseases in places like the United States of America.
Mathematics of Epidemics by Trish Campbell
Through Trish Campbell’s talk you can better understand how mathematical models can help predict spread of infectious diseases, using the example of how videos and images can become viral on the internet.
Forecasting Infectious Disease Epidemics Using Dynamic Modelling: Ebola and Zika as Case Studies
Last but not least this is an interesting talk that talks about dynamic modelling for the forecasting outbreaks of infectious diseases, using Ebola and Zika data and epidemiology as examples.
I hope these videos have convinced you of the potential of data and of how urgent it is that all levels of healthcare in all countries go through digital transformation. All healthcare systems need to advance and connect so that we can together face and fight disease, one of the biggest enemies of humankind. Viruses and bacteria don’t understand what borders are and a strong health system is only as strong as the health systems surrounding it.