How data can help fight a health crisis like the coronavirus

We are facing a global crisis. The decisions leaders will make over the coming weeks will shape the world for years to come. From a public health perspective, to combat an epidemic, officials must take a number of actions, such as: build awareness, set guidelines for health professionals, target infection clusters, limit population movements, and allocate scarce resources. These decisions will influence how many people will survive and how many will die over the coming days, weeks and months. Leaders must act quickly and decisively in order to save lives.

One basis for taking these decisions is the availability of the right data. In the fight against coronavirus, insight into preventive actions, population mobility, the spread of the disease, and the resilience of people and systems to cope with the virus, can help public health and humanitarian leaders respond more effectively to the COVID-19 epidemic.

Yet, today, the public health leaders who make the hard choices are lacking high quality, high resolution data on key questions, such as: where is the disease likely to spread? Are there priority areas that we need to contain to limit further propagation? Where are the most vulnerable communities?

Monitoring population movements using mobile phone data

Aggregated mobility information from telecom data was used during the Ebola outbreak in West Africa and has been further researched by the UNICEF Innovation Lab, Flowminder and others.

Recently in Belgium, Dalberg Data Insights, one of the organisations mandated by the Belgium Government to lead the Data Against COVID-19 task force, has been analyzing aggregated and anonymised telecom data from the three telecom operators in the country. The main goal is to understand human mobility trends in regard to lockdown measures and evaluate the risk of infection increases of a specific region. Overall in Belgium, human mobility has decreased with an average of 54%, with some areas in Belgium seeing an even bigger decrease. The crisis response team in Belgium can refer to this analysis when it comes to the impact of imposed measures and indicate the risk of virus outbreak and imported cases from other regions.

Another example of effective use of data comes from South Korea. The country is watching quarantined citizens with a mobile app, developed by the Ministry of the Interior and Safety, as reported by the MIT Technology review. The country’s sense of urgency escalated after “Patient 31” became a “superspreader” and is thought to have caused the rapid rise in cases. People who are quarantined can use the app to communicate with the local government case officers and report their symptoms. Both the person, and the government case officer are notified if the person leaves the designated quarantine zone. The app is not mandatory, and people can opt out. Such measures, together with the South Korea mass coronavirus testing, has contributed to “flattening the curve” in the country. The number of daily confirmed cases peaked on February 29 and has decreased ever since.

Identifying communities at risk

Identifying the most vulnerable communities can be important for health officials to guide response efforts like health infrastructure improvements, emergency funding allocation, and preventative measures. This is especially relevant in the emerging countries where living conditions can compromise one’s ability to follow advice on how to behave. Washing your hands for 20 seconds or more with clean soap is hard to do when your main source of water is a polluted river. Self-quarantine and self-isolation is unrealistic when you share a single room with other family members. And staying at home is impossible if you live hand to mouth and have to go out twice a day to work and then stock up for the next meal.

Authorities can map areas where the ability to respond appropriately is compromised, with a high level of detail, using a combination of available primary data collection, data from national bureaus of statistics and satellite images. The Location Analytics (LOCAN) team at Dalberg Research based in Kenya is analyzing risk profiles in multiple African countries. The results are then fed back into epidemiological models as input for informed decision making on the crisis response. A similar risk model, which drew on three key risk variables – people aged over 60, regular smokers, and those who use dirty cooking fuel in their houses – was developed and applied to Nigeria where health officials continue to announce new cases despite a strong early federal response.

What about data privacy?

There are growing privacy concerns about the ways governments use data to respond to the COVID-19 crisis. As new technologies emerge that aim to collect, disseminate and use data in order to support the fight against COVID-19, we need to ensure they respect ethical best practices. Even in times of crisis, we need to comply with data privacy regulations and ensure that the data is used ethically.

One way to do that is to establish independent ethical committees or data trusts. Their role will be to create data governance mechanisms to find the balance between competing public interests, while protecting individual privacy. Examples of such rules include setting up clear guidelines on the purpose and timeline for the use of the data, defining clear processes for the access, processing and termination of use of personal data at the end of the crisis.

Tedros Adhanom Ghebreyesus, director-general of the World Health Organization, said: “You cannot fight a fire blindfolded.” The right information in the hands of the right people can save lives in a time of crisis. It will be essential to ensure that such health surveillance measures will not prevail beyond the extreme circumstances we are facing today, so that people do not feel they are losing their privacy in a new world order.