KUALA LUMPUR, April 7 – Nearly four of 10 coronavirus clusters reported since late February originated from factories, followed by community spread, construction sites, educational centres, and retail outlets.
A total of 314 Covid-19 clusters were reported nationwide from last February 22 to April 2, of which factories comprised 38.54 per cent, community spread (15.29 per cent), construction sites (8.6 per cent), educational sites (7.96 per cent), shopping areas (7.01 per cent), among other venues, as reported by the Ministry of Health (MOH).
With the biggest proportion of Malaysia’s population, workforce and industries coming from Selangor, it is no surprise that the majority of clusters (both workplaces and community) comes from Selangor (24.2 per cent), followed in a close second by the state of Johor (23.57 per cent), and Sarawak (10.19 per cent) in third place.
Many of these clusters were contributing factors to the rise in Covid-19 cases, thus the three states that have a rise in cases commonly are Selangor, Sarawak and Johor (though again it must be reiterated that the cumulative number means very little when comparing between states. The cases per 100,000 population is the one indicator that puts everyone on par for comparison).
But let us leave the states and their cumulative numbers aside. Is there anything that we can learn from clusters? Of course there is! By looking at the clusters, we can determine if our screening can be improved in terms of the areas in where we should be screening. Let’s take a look at the chart below:
This pie chart tells us the distribution of clusters from February 22 to April 2 this year in Malaysia. Although this chart is important to show where common places of clusters may be found, it doesn’t do justice to the magnitude of the clusters.
In other words, a factory cluster may only contain 25 infected people and this cannot be compared to another factory cluster with 1,250 cases. This mean that the proportion of clusters does not account for the number of cases the clusters contribute.
Thus, we conducted another analysis for comparison.
The graph above describes the number of cases within each type of cluster that were reported on the day the cluster was announced. From February 22 to April 2, there were 9,316 infected people within the 314 clusters.
We can see that most Covid-19 cases came from factories (48.06 per cent), followed by community spread (12.55 per cent), construction sites (11.56 per cent), detention centres (5.62 per cent), and then only educational sites (5.53 per cent).
So this serves as a reminder that although there are more Covid-19 clusters originating from educational sites (7.96 per cent) than detention centres (2.55 per cent), there are more cases within detention centres (5.62 per cent) than educational facilities (5.53 per cent). This is a different way of looking at the data.
Another way of looking at the data is to obtain the number of Covid-19 cases according to a broader classification — work clusters, community or institutionalised (either detention centres, old folks’ homes etc). As we can see from the graph above, the majority of Covid-19 clusters originate from workplaces and from the previous graphs, the majority of them came from factories and construction sites.
The most accurate way to determine the troublesome clusters are to get the total number of cases within each cluster to quantify which cluster (as in the second graph) is the most infectious one. That is something this research is not able to do as data reported daily on government websites is very limited (and sometimes not made available).
However, the message we intend to share here is that we cannot be comparing the presentation of clusters as a direction of where we are going to test. We need to see the weightage of where the clusters are coming from by weighing up the total cases within the clusters to allow us to judge the best mitigation for each area (each will differ from another).
Another worrying area that merits more focus is the sudden increase of Covid-19 cases among plantation workers. The government must strongly consider screening such workers (both registered and unregistered ones). This is also the best time to register plantation workers for Covid-19 vaccination.
Note: CodeBlue is publishing this analysis anonymously because the author says: “Malaysia today punishes those who want to put things right”.
