Bengaluru reportedly has the highest active Covid case-load ever seen in any Indian city. Hospital beds and oxygen are in short supply. By late April several hospitals in the city were already facing oxygen crises. As in Delhi, the lack of oxygen seems likely to lead to tragic and avoidable deaths.

How bad is the surge?

To say that Bengaluru’s epidemic is huge is an understatement. According to the city’s bulletins, in the week from April 29 to May 5 it registered 1.5 lakh cases. This is a staggering case-load in a city with an estimated 2021 population of 12.7 million. It is more than two times Mumbai’s highest-ever weekly new cases. This one week’s cases are, in fact, 18% of the city’s total cases during the pandemic so far.

To put the current numbers in perspective, weekly cases are five times higher than during last year’s peak, and an astonishing one hundred times higher than three months ago.

Test positivity has also been rising fast. Thirty seven percent of all Covid-19 tests returned positive during the week ending May 5. This compares to 20% the previous week, and just 12% the week before that. Many, many infections are surely going undetected as testing fails to match spread.

A surge in Covid-19 deaths, some of them unrecorded

In the week prior to May 5, Bengaluru recorded 867 Covid-19 deaths. This is very close to the per capita recorded death rate seen in Delhi, and considerably higher than the worst seen in Mumbai during this wave. This one week saw around 12% of the city’s total recorded deaths from the coronavirus during the pandemic so far.

Bengaluru’s recorded fatalities could, moreover, be a considerable undercount: an investigation by The News Minute found that during one week in April, there were at least 60% more Covid-19 protocol funerals than official deaths from the virus. As hospitals and testing become overwhelmed, the unrecorded toll could also shoot up.

Rewinding a little, we don’t have a clear picture of how many Covid-19 deaths occurred during 2020. All-cause mortality data for the city does not seem to be available. However, a report in The Hindu suggested that the first seven months of 2020 saw 12,000 more deaths than during the same period in 2019. Total recorded Covid-19 deaths in the city numbered just 1000 by the end of July, 2020, a possible sign of major death undercounting during the early epidemic.

How did things get so bad?

“R” is the average number of new infections resulting from one infection. It is often used to express how fast an epidemic is growing. A value less than 1 means that the epidemic is subsiding, while a value of more than 1 means that it is growing.

We can estimate R from cases, although the process involves some assumptions. Bengaluru’s estimated R crossed 1 in late February and has stayed above 1 since. We can see this as the beginning of this surge. From the second week of March until late April, R hovered around 1.5.

This was a long period of rapid growth, during which cases doubled every nine days or so. Residents of the city must ask: what steps did the city authorities take to slow the epidemic during most of this period? There appears to be a dangerous pattern of ignoring rapid growth when case numbers are relatively low, and waiting until numbers are high to act. By this time it is too late.

What has Bengaluru’s epidemic looked like so far?

To understand the current surges it’s important to look back. How many city dwellers were infected during 2020? Seroprevalence surveys (or “serosurveys” for short) estimate how many people in a sampled population have developed antibodies to SARS-CoV-2, the virus responsible for Covid-19.

A study in September 2020 that measured both antibody levels and active infection estimated that around 30% of city dwellers had been infected. This came, however, before the city’s 2020 peak in cases. During the last three months of the year cases more than doubled, while recorded deaths increased by almost 50%. Based on the rise in fatalities, perhaps 40%-45% of the city had been infected by the end of the year.

There is high uncertainty in this estimate, but most data points to Bengaluru not being hit quite as badly and Mumbai or Delhi during 2020. Similar to Mumbai and Delhi, the city saw uneven spread. It seems likely the slums were hit earlier and harder: a serosurvey in a slum during late August and early September found 58% with antibodies, much higher than the value for the city as a whole.

How uneven spread might be affecting the current wave is not yet clear, but we would in general expect pockets of the city that saw lower spread during 2020 to be more vulnerable this time around.

New variants of the virus

Whatever the exact level of prior infection in the city, the recent speed at which the disease has been spreading is quite remarkable. Could new variants of the virus be playing a part?

This seems likely, based on limited data and the patterns we are seeing across the country. As of May 6, 2021, the public repository GISAID includes genome sequences for 130 samples collected from Karnataka since March 1. Of these, 26% are of B.1.1.7 (the “U.K. variant”), while 44% involve B.1.617 (the “Indian variant”).

GISAID data suggests that the fractions of both B.1.1.7 and B.1.617 grew significantly in the state from March. In particular, nearly all the B.1.617 samples on GISAID date from March or later. But the total numbers are small and for the state as a whole, rather than Bengaluru. So, we can only speculate that these variants might be contributing to Bengaluru’s huge surge.

B.1.1.7 is known to be highly transmissible, spreading more easily from person to person; and B.1.617 possibly even more so. More transmissible variants could allow the disease to rage through areas where efforts to slow the spread were successful during the first wave. There may also be some reinfections occurring – the chief minister was reportedly infected with the coronavirus for a second time during this wave – though whether the variants make reinfections more likely is an open question.

Could variants explain the high death toll? Limited sequencing means we don’t know. There is evidence that B.1.1.7 causes more severe disease – and this could be playing some part in increasing hospitalisations. On the other hand, there is currently no evidence for greater lethality of B.1.617 – a small relief.

What next?

Time after time, we are witnessing cities that were badly affected last year suffering again during this wave. Bengaluru is the latest. This pattern may reflect uneven spread, waning immunity or the presence of variants that are more transmissible or more able to cause reinfections. But, whatever the reasons, naive herd immunity arguments are not holding up.

During the first week of May the growth in cases has slowed. This could signal the first effects of a lockdown that began on April 27. But rapidly rising test positivity casts doubt on this explanation. Until both cases and test positivity start to fall there is little room for optimism.

What about vaccines? An estimated 16% of Bengaluru’s population have received at least one dose of a vaccine, and hopefully the effects will be seen in reducing severe disease in the city, as probably witnessed in Mumbai. It might be hard to see a vaccine effect in the midst of such a rapid surge. For now the city remains in critical condition.

Case, fatality and testing data used in this piece is taken from the BBMP Covid bulletins and from

Murad Banaji is a mathematician with an interest in disease modelling.

Corrections and clarifications: A previous version of this article said that the new cases registered in Bengaluru between April 29 to May 5 was three times Mumbai’s highest-ever weekly new cases. The city’s figure is actually more than two times Mumbai’s highest weekly cases.