Spatio‑temporal epidemiology and associated indicators of COVID‑19 (wave‑I and II) in India

The spatio-temporal epidemiology of COVID-19 in India was never uniform. Cases and deaths varied sharply between states, union territories and districts, and the reasons for that uneven spread were not obvious. This study mapped where the burden fell and which underlying indicators were associated with it.

It was published in Scientific Reports in 2024 by Karuppusamy Balasubramani, Venkatesh Ravichandran, Kumar Arun Prasad and colleagues, including Rakesh Kumar Sharma.

The data behind this COVID-19 epidemiology study

The authors downloaded and curated district-level daily data on COVID-19 cases and deaths for the whole country, from February 2020 to August 2021, covering the first and second waves. The figures were normalised against the projected population for 2020 before the space–time trends were analysed.

India’s first wave peaked in September 2020 and declined gradually until February 2021, and was relatively mild, helped by the nationwide lockdown. The second wave, from February 2021, was far more damaging, with widespread mortality and morbidity. During that period the Delta variant, first identified in Maharashtra, became the dominant strain.

Where the burden fell

States and districts in southern India were the worst affected. Coastal districts, and districts adjoining the large urban regions of Mumbai, Chennai, Bengaluru, Goa and New Delhi, recorded more than 50,001 cases per million population.

Which indicators shaped COVID-19 epidemiology

A negative binomial regression was run with 21 independent variables, identified through multicollinearity analysis, covering demography, socio-economic status, environment and health. It was applied separately to the first wave, the second wave and the combined totals for cases and deaths.

The wealth index, derived from household amenities data, showed a high positive risk ratio with both cases and deaths. Other socio-economic factors were also positively associated at district level, including literacy rate, health services, the rate of other workers, alcohol use in men, tobacco use in women, overweight or obesity in women, and rainfall. These variables were closely interconnected in the districts that became hotspots.

Among them, the wealth index, the literacy rate and health services stand out, because all three are key indices of socio-economic development within a state.

Why this matters for planning

Identifying district-level space–time trends, and the indicators that go with them, gives policymakers something concrete to work from during a public health emergency. Mapping the spatial differences across districts allows public health policy to be shaped region by region rather than applied uniformly.

This kind of district-level geospatial analysis is close to CDPRR’s own work on risk assessment and mapping. Related material on this site includes computational screening of phytoligands against SARS-CoV-2 and the Mumbai Port Trust chlorine leak case study.

The full paper is available at doi.org/10.1038/s41598-023-50363-2.

(Read the full paper in PDF format)