A recent article in The Atlantic, “k: The Overlooked Variable That’s Driving the Pandemic” makes for interesting reading particularly for Geospatial practitioners. In brief, the article observes that while the R factor is a good measure of how the contagion is affecting the population, it actually represents an average value and does not show its spread. In simple terms, if R is 1 then every infected person is infecting one more. Ideally R must go below 1 to indicate that the infection is reducing. Thus if R is 0.83 it means every 100 people are infecting 83 others, hence a downward trend in infection. in India, while the R value has fallen below 1, cases are increasing in states like West Bengal, Kerala and Maharashtra. To be able to explain this apparent anomaly it is necessary to look into the geographic spread to the virus.
This is where the k factor comes in (https://theconversation.com/is-the-k-number-the-new-r-number-what-you-need-to-know-140286). This factor pinpoints the manner in which the infection spreads. It indicates the percentage of people who are responsible for very large spread of diseases. The lower the k the lesser the number. This takes into account that one person can act as a super spreader and infect many while another person may not infect that many. This happens due to very many factors, like the person’s viral load, the way they interact with others and the environment they interact in. This is also called the Pareto Principle or the 80/20 rule which states that 80% of all outcomes are initiated by 20% of inputs. It has been observed that Covid 19 spreads in bursts due to super spreaders in contained environments and that only 19% of infected people have caused 70% of all infections so far.
Thus, it is important to identify clusters of infection and the movement of a super spreader from one cluster to another where the person may start another cluster of infections. To know this, the need is not just forward contact tracing but reverse contact tracing. If a person is infected then it is more important to know where he/she has been and contracted the infection rather than who has come into contact with the person subsequently. Because maximum infections happen in super spreader events like the Korean church incident where one infected person infected 5000 others! Two countries which avoided the spread of the pandemic without strong lockdown are Sweden and Japan and both followed this procedure. They closed down likely super spreader locations like malls and gyms but not schools.
Also Read: COVID impact: Retail industry recovering fast backed by consumer resiliency
This is where geospatial comes in. Unfortunately, contact trackers like Aarogya Setu of India or the Google-Apple API promoted by these two giants (see: https://www.technologyreview.com/2020/05/07/1000961/launching-mittr-covid-tracing-tracker/) which use geospatial and communications technologies can only inform if you are near an infected person but it cannot find out if the person is a super spreader or has come from a super spreader event. In India, the state of Kerala initially used ‘heat mapping’ of Covid cases and contact tracing to control the spread very quickly. However, this approach seem to have been given up and the dashboard of the health services does not use maps. Interestingly it traces the routes of infected people in the form of flow charts!
Most Covid dashboards are very simple with mapping of cases and Covid health facilities. They lack analytical tools which could map out super spreading locations and reverse tracing. Privacy is a key issue hence the hot spot locations have to be determined using anonymised location data. Such data is available from the governments which are tracing individuals in terms of contact tracing. For example, Anonymised Aarogya Setu data of Indian individuals is available from National Informatics Centre of the Ministry of Information Technology for use by Indian Central and State Government entities at:(https://aarogyasetu.gov.in/wp-content/uploads/2020/06/mygov-1000000000981057882.pdf). Super spreader clusters can be determined by using simple clustering techniques. Location information of individuals can be backtracked to possible clusters which can then be suitably addressed by way of containment and other administrative actions.
The k factor becomes very important as the pandemic winds down. A low k indicates the possibility of a small number of infected could infect many more as community areas open up. The spate of infections being seen in Europe and other countries is because of this situation. This is the time for dynamic location tracing of super spreader events and immediate action. This is the time for Geospatial systems to move beyond dashboards and introduce spatial analytics.
