Categories
Analytical Insights

Correlation Between Mobility Restriction and Spread of COVID-19

The top 30 countries in UNDP’s Human Development Index (HDI)1 were analysed to uncover the dynamic relationships, if any, on human movement2 to the transmission of COVID-19. It was observed that when Governments in these countries announce and implement various kinds of movement control, there is a noticeable correlation in COVID-19 confirmed cases declining after about 2-3 weeks.

Figure 1 The Yellow Column shows the point where mobility greatly decreases for transit stations, workplaces, retail and recreation, which is then followed by a decrease in confirmed number of COVID-19 cases

This relationship in declining cases appears to be strongly correlated when community mobility in transit stations, workplaces, retail and recreation reduces by 40% or more compared to pre-Government-announced measures.

Figure 2 As shown by the Yellow column, on 22nd September 2020 Prime Minister Boris Johnson set out a raft of new coronavirus restrictions for England, which immediately resulted in significant reduction in mobility at parks

The first movement indicator to usually show an immediate response or compliance towards the Government’s call to reduce movement in public spaces was observed, showing a decline in mobility at parks. These observations appear to indicate that compliance to Government-announced social distancing measures have a direct correlation to the effectiveness in mitigating the spread of COVID-19.

Credit to: Dr Chook Jack Bee, Prof David Bradley, Prof Teo Kok Lay, Dr Lai Kee Huong, Dr Jane Teh Kimm Lii and Dr Peh Suat Cheng, Sunway University

Note: The above analysis was made possible by the collaboration of Sunway University with the Global COVID-19 Index (GCI) initiative.

1 The UNDP HDI ranks countries according to life expectancy, education levels and GNI per capita. Countries analysed were Norway, Ireland, Switzerland, Germany, Sweden, Netherland, Australia, Denmark, Singapore, Finland, United Kingdom, Belgium, New Zealand, Canada, USA, Austria, Israel, Japan, Slovenia, Korea, Luxembourg, Spain, France, Czech Republic, Malta, Estonia, Italy, United Arab Emirates, Greece
2 Mobility datasets were obtained from Google and covers mobility observations in the categories of retail and recreation, groceries and pharmacies, parks, transit stations, workplaces, and residential relative to pre-epidemic period.

Categories
Analytical Insights

Unique Cross-Practice Analysis reveals 23 Countries which are likely to experience COVID19 Second Wave

What do you get when you apply price analysis methods commonly practiced by financial analysts to COVID19 data? Sunway University has discovered that applying the concepts of support and resistance in the analysis of case trend data can enable the identification of countries that are in the process of facing second waves of COVID19 infections.

The concept of support level in price analysis is when a variable is expected to worsen, but a myriad of factors causes the variable to slow in its worsening progress. On the contrary, a resistance level works in the opposite where an improvement in the variable is met with various factors that cause the improvement to be slowed down.

Figure 1. Global trend of COVID-19. Three supports (indicated by yellow bars) predict increased number of COVID-19 cases.

Using these concepts on COVID19 case data, Sunway University first analysed Total Case data globally in the initial stage of the pandemic as a control mechanism. Based on Figure 1, it is observed that daily confirmed cases vary day-to-day but are generally on a rising trend. This is where the Sunway University team led by Dr Chook Jack Bee, and supported by Prof David Bradley, Prof Teo Kok Lay, Dr Lai Kee Huong, Dr Jane Teh Kimm Lii and Dr Peh Suat Cheng, introduced the idea of supports in the analysis (indicated by the yellow bars).

Figure 2. Trends of COVID-19 infection in United Kingdom. Three supports (indicated by yellow bars) predict increased number of COVID-19 cases. Three resistances (indicated by red bars) predict decreased number of COVID-19 cases.

Applying this simple concept to individual countries reveals more revealing observations. Figure 2 illustrates an example undertaken on United Kingdom. “These support and resistance thresholds are determined from the observed ‘valleys’ of the data plot. The first ‘valley’ will serve as baseline. If the second and third ‘valley’ are higher than the baseline, then it is likely that there is support for a rising trend and vice versa. The red line shows stringency and in respect of incidence would seem to show that even a relatively small relaxation in the former can produce a steep rise in incidence (subsequent to a period of latency). The UK data do not represent a singularity, there being a good many other examples of national data that support such an emergent picture,” explains Dr Chook.

Using 90-day COVID-19 data up till 27th September, the Sunway University team have identified 23 countries that are likely on a trend of new waves of COVID19 in the very near future. These include Bulgaria, Canada, Czech Republic, Denmark, Finland, Georgia, Iceland, Indonesia, Jordan, Malaysia, Montenegro, Mozambique, Myanmar, Netherland, Poland, Portugal, Russia, Uganda, Ukraine, United Arab Emirates, United Kingdom, United States and Uruguay.

Sunway University is currently applying machine learning methods to be able to quickly ascertain similar patterns as the infection continues to evolve. “We hope this analysis will prompt respective countries the urgency to undertake proper measures to counter the rising trend of COVID-19 cases,” added Dr Chook.

Credit to: Dr Chook Jack Bee, Prof David Bradley, Prof Teo Kok Lay, Dr Lai Kee Huong, Dr Jane Teh Kimm Lii and Dr Peh Suat Cheng, Sunway University

Note: The above analysis was made possible by the collaboration of Sunway University with the Global COVID-19 Index (GCI) initiative.