How to Track Pandemics With our Medical Intelligence Solution

Within this article, we explain how to use our Medical Intelligence solution to create additional insights for current and future pandemics. We summarize all the findings of the last weeks and describe a use case of Covid-19 based on OSINT information about Yemen. The indications of this example are compared with the official numbers of the often-mentioned JHU Covid-19 map.

Insights from Corona Pandemic

• A pandemic is a global catastrophe. It is going to hit large parts of the entire world.
• We need to understand where the viruses are active and which viruses have the potential for spillover. The more comprehensive the understanding, the greater the chance of interrupting the spillover in time.
• The number of zoonotic transitions will increase because of the growing population. Especially countries suffering from insufficient hygienic conditions pose a great risk, e.g. Ebola in Africa or SARS/Corona in China.
• A large pool of viruses are circulating and we will not be aware of it until we see an increasing number of people contaminated.
• A lot of countries are not capable or not willing to provide validated numbers.
• Following the first pandemic wave, people typically increase their mobility, thus spreading the virus which replicates abundantly given the favorable environment condition. When the right ecology comes into play, it starts getting cold, and damp, then the virus becomes active. 

Consequences and Challenges for NGOs and Governments

Looking at the facts, both time and information play an essential role in the prevention or containment of a pandemic. NGOs and governments must be able to process information from different countries in the shortest possible time. Typically, information is available only in the respective national language, which makes evaluation very difficult. In addition, the information does not come from one source, but is disseminated through various sources, such as social media, TV or newspapers.

That is why prevention is an enormous challenge.

What is Medical Intelligence?

According to the US Department of Defense, Medical Intelligence is defined as follows:

That category of intelligence resulting from the collection, analysis, and interpretation of foreign medical, bio-scientific, and environmental information that is of interest to strategic planning and to military medical planning and operations for the conservation of the fighting strength of friendly forces and the formation of assessments of foreign medical capabilities in both military and civilian sectors. 

Traditionally, Medical Intelligence focuses only on military affairs. But the insights and methodology can be used as well to assure the quality of health systems for civilians. 

How does Medical Intelligence work?

As mentioned before, it is all about collecting relevant information. The initial search for medical data is quite broad creating a huge amount of results. Analysts apply a generic keyword search to multilingual data sources and refine the search query after analyzing the results.

As an example, our analysts tried to get a better understanding of the Covid-19 situation in Yemen. At the time of writing, there is only one confirmed Covid-19 infection in Yemen according to JHU Covid-19 map. This information was confirmed by a news article on the 28th of April saying that the infected person recovered and that Yemen was then Corona-free. As Yemen’s health system got seriously damaged during the latest wars, this information does not seem to be reliable. Our analysts got triggered and started with some OSINT research for the keyword “Yemen Covid19” by using the Federated Search. The first results were related to the official information.

Our analysts modified the context of the query by searching for “Yemen Fever”. The goal was to get more relevant results that might indicate another situation in Yemen. There was one result that raised our attention:

This Twitter Tweet posted on the 28th of April at 06:00 am (GMT+1) raised some rumors that the numbers transmitted to Johns Hopkins might not be correct. Another Twitter Tweet posted on the 28th of April, 11:42 pm (GMT+1) even corroborated the first rumors. 

The Yemeni Ministry of Health and Population is analyzing whether it is the well-known dengue fever or the coronavirus. However, the first tweet clearly describes the symptoms of a Covid-19 infection.

In order to track the development in Yemen and to stay up-to-date, it now makes sense to automate the search with a refined search query. Our analysts are notified in case of new relevant findings by the system and can now concentrate on the final information evaluation. Information that is collected automatically, is processed and analyzed by the system. The information will be shown on a real-time dashboard and within an intelligence feed.

Key Takeaways

• NGOs and governments should use official numbers just as an indicator and verify the assumption by using OSINT or other intelligence areas.
• Our Medical Intelligence solution made it possible to analyze Arabic information on Twitter and in local newspapers. The NLP modules, such as machine translation, sentiment estimation, … available in the Data Fusion Platform gave us a tactical advantage.
• Only 30 minutes passed from the first question to the first results. Instead of using multiple search engines in parallel and manually translating each search result, we used a single search portal for multiple data sources, applying advanced NLP processing to keywords and search results.