The COVID-19 Pandemic in Sri Lanka: Contextualizing it Geographically

By Dr. Nalani Hennayake and Dr. Kumuduni Kumarihamy

(Continued from Tuesday)

Part III

The mapping of COVID-19 cases at the local level (preferably at the GN division level), but at least at the MOH level, is essential for future planning. It is commendable that such a map has been produced at the Divisional Secretariat Division level for the Kandy District.

In Map 06, the DSD/MOH areas are categorized, based on the number of patients detected up to November 11, 2020, into six categories. Such mapping again runs the risk of the ‘modifiable areal unit problem’ since the data are aggregated to the DSD/MOH level. However, the dot density map of the Kandy District (Map 07) is more useful as it randomly marks the locations of positive cases within the GN division. Compared with the previous choropleth map, such a density map allows us to see the geographical patterning of COVID-19 to a certain extent and to identify the ‘hotspots’. If the density map is overlaid with the GN boundary map, the spatial heterogeneities will be highlighted more than at the DSD/MOH level.

We produced a new map (Map 08) of Kandy District with the limited data available for the GN level to see the spatial dispersion of COVID-19 as a sample exercise. By November 21, 2020, there were only 14 positive cases in the Medadumbara division, and they are distributed only in six GN divisions (Udatenna – 4, Vilamuna – 3, Waradiwela – 1, Randeniya -1, Karalliyadda 1, Kandegama 1, being updated 1), as shown in map 8. In map 06, the entire Medadumbara division is marked in red as a high-risk area giving a very different spatial impression.

If the spatial heterogeneity can be depicted at the GN division level, we can identify COVID-19 clusters/hotspots more accurately in a given area. Moreover, we can further investigate the relationship between the COVID-19 clusters/hotspots and the socioeconomic and ecological determinants such as road density, education level, poverty and connectivity (using road network). It is clear that high population density, high accessibility, and poverty, to a certain extent, correlates with the spread of the pandemic. We have done this exercise using Kandy as an example to show how such mapping and the resulting geo-visualization can be a powerful tool in understanding the severity of the situation and enabling decision-makers and implementers of control strategies. If such maps can be propagated through media, as we do with elections and weather reporting, and given their visual effect, people themselves may enforce their spatial restrictions.

If such maps can be developed for all MOH areas taking the GN division as the spatial unit of analysis, Sri Lanka would have a much better representation and geo-visualization of the COVID-19 pandemic situation, which in turn would enable one to envisage its severity and be guided by anticipatory actions. The determination of the COVID-19 positive cases’ location, for mapping, can be based on any one of the following; where it was contracted, the person’s place of work, and place of residence. The critical information needed here is not any identity-related information of the infected, but the geographical locations, the possible spaces and places of contact. Therefore, marking any one or all of the above or whatever available data would provide a map of potential localities and places for contracting the coronavirus. Travel restrictions can be enforced either by the government or exercised voluntarily by individuals when such information is provided at the GN level. Identifying such spaces and places with a potential for spreading the disease and clusters of locally developing cases is extremely important to prevent further escalation of the coronavirus. Our argument here is that mapping geographies of infection at a micro-scale, such as GN divisions, is vital. It provides a visual outlook of the crisis and its localized nature.

Geographies of vulnerability

One of the most important results of such mapping, especially if it can be done at a micro-scale level such as the GN division, is that such spatial patterns guide us to understand the social vulnerabilities. We already explained how the COVID-19 pandemic correlates with high population density, high road density, and poverty. In other words, this allows us to denaturalize the pandemic within the local context and to recognize the social, economic, and political contexts that engender the pandemic situation. Those who were infected were infected not because they were at ‘the wrong place at the wrong time,’ but have inherited or made to live/work in ‘wrong places’, socio-economically downtrodden places.

The national data provided by the Ministry of Health on the distribution of the total number of COVID-19 patients at district level and the information released from the National Operations Centre for the Prevention of COVID-19 (NOCPC) allows us to speculate that several communities are in danger. COVID-19 has become severe in certain types of localities and spaces at the outset. For example, in Colombo, high density, low-income (we prefer to call these poverty-stricken) localities are worst affected by the pandemic. Such localities are locked down to prevent from spreading outwards.

Even in Kandy, one of the localities where a higher number of cases have been reported is a poverty-stricken, high-density urban locality that provides essential labor to the city. Thus, it is clear that the urban high-density and low-end social category is a straightforward victim of this pathogen.