Research Excerpts

Lower Case-Fatality Rates for Economic Blocks which have both High Ambient Temperatures and Low Latitude

Sunway University has identified that countries that possess both high ambient temperatures and low latitudes have low COVID-19 Case Fatality rates. These findings appear to support epidemiological hypothesis that temperatures do play a role in managing the impact COVID-19 may have, particularly on mortality rates.

Excerpt from Sunway University:

We started by applying SARS-coronavirus 2 (SARS-CoV2) as the causative agent of COVID-19. It has been suggested that high ambient temperature disfavours coronavirus infection. This is supported by two lines of evidence:

  1. Elevated surface temperatures reduces viral viability;
  2. Infectivity of coronavirus decreases towards deeper, hotter airways.

Thus the question arises, ‘Can high ambient temperature limit the number of fatalities in COVID-19 cases?

Case-fatality, ambient temperatures and the latitude of 184 economic blocks were retrieved on 21 May 2020. After excluding 123 of those blocks with fewer than 2,000 confirmed cases, we proceeded with analysis of the remaining 83 economic blocks. We calculated case-fatality has by taking the total number of deaths divided by the total number of confirmed cases.

Figure 1. Influence of ambient temperature and latitude on case-fatality

Case-fatality was divided into four groups:
• 0-2%
• >2-4%
• >4%-8%
• >8%

We analysed the average monthly lowest and highest temperatures from the month of which the first confirmed case-fatality case(s) were reported through to May 2020. Because of the very high case-fatality observed in a number of European countries including Belgium, France and the United Kingdom, we also looked at the association between latitude and case-fatality. Kruskal Wallis Test analysis has revealed the average monthly ambient temperature (lowest, P = 0.025; highest, P = 0.001) to be inversely associated with case-fatality, whereas conversely latitude is seen to be directly associated with case-fatality (P = 0.037) (Figure 1).

In simpler terms, we found a correlation that the case fatality rates were highest in areas of lower mean temperatures and higher latitudes, whilst the inverse can be noted in countries that have higher meant temperatures and lower latitudes (i.e. closer to the Equator).

We found this to be a key insight and Sunway University are looking to find other unique correlations that could be gleaned from looking at COVID-19 using a data-driven approach.

Credit to: Dr Chook Jack Bee, Prof David Bradley, Prof Peh Suat Cheng from School of Healthcare and Medical Science, Sunway University; and Dr Jane Teh Kimm Lii and Prof Teo Kok Lay from School of Mathematical Science, Sunway University

Research Excerpts

Fourteen unique COVID-19 trend groupings identified by Sunway University following analysis on confirmed cases of 184 economic blocks

Using the Global COVID-19 Index (GCI) model as a supporting data engine, Sunway University has identified a unique way of looking at the spread of COVID-19 which it hopes to unveil new insights into country-level epidemic management. Via a dendogram analysis which takes into consideration multivariate factors, the researchers at Sunway University have identified 14 unique trend groupings from the 156 countries covered by the GCI.

Research Excerpt from Sunway University:

We attempted to study the trends of the first 60-day cumulative confirmed cases of COVID-19. We started by looking at the full sample set available from the GCI. After filtering out economic blocks with incomplete data and total confirmed cases fewer than 50 within the first 60 days, we proceeded to perform cluster analysis on 156 economic blocks.

A 5-day gradient was calculated for the cumulative confirmed cases of COVID-19. We then utilised a Hierarchical Clustering (Average Linkage) algorithm to construct a dendrogram. From the dendrogram, we identified 14 distinct clusters of countries that showed similar distribution trends of cumulative confirmed cases (Table 1).

These clusters cover a total of 58 economic blocks (38% of the filtered sample size). The remaining 98 blocks like China and Norway have standalone trends which potentially can be case studies of their own.

It will be interesting to see if the observed epidemic management of economic blocks within the same cluster share commonalities that could lead to greater insights as to what caused these trends to be similar. We hope to use these findings to enable data-driven research from a different perspective to the more commonly explored parameters of regional, cultural, income, healthcare systems and population density. Of particular interest would be the measures undertaken or not undertaken to contain the spread of COVID-19.

Table 1. Distribution trends of cumulative confirmed cases of COVID-19

Credit to: Dr Chook Jack Bee, Prof David Bradley, Prof Peh Suat Cheng from School of Healthcare and Medical Science, Sunway University; and Dr Jane Teh Kimm Lii and Prof Teo Kok Lay from School of Mathematical Science, Sunway University