As appeared in the Netra News — নেত্র নিউজ
The initial projections seem overblown, but whether they remain so depends on how the government deals with the lockdown.
by David Bergman 22 May 2020
When the first epidemiological projections of the numbers of people who could die in Bangladesh from the Covid-19 virus started to be published, it seemed that the country faced a tragedy of biblical proportions.
First there was the study by five academics from Brac University, North South University and Johns Hopkins University which projected that, without any government action to mitigate and suppress the spread of the virus, half a million could die in Bangladesh by the end of May. Then, there was a United Nation inter-agency memo with projections suggesting that the death toll, in a “no intervention scenario”, could rise to as high as two million. And then there was the Imperial College projection which suggested that while the worst case scenario may be 637,191 deaths in Bangladesh in an “unmitigated” scenario, this number could be reduced by about 92% to 52,000 if the government organised a sustained lockdown in the country until a vaccine was found.
Yet, looking at Bangladesh’s official death numbers from Covid-19, it would seem that even Imperial College’s best case scenario figures were hugely overblown. 64 days after the country’s first reported death on March 18th, there are a total of only 432 officially reported deaths. This equates to two deaths per million, and is one of the lowest death rates from Covid-19 in the world — compare it to Belgium with 784 deaths per million, Spain with 591 deaths per million, the United Kingdom with 513 deaths per million, the United States with 275 deaths per million, and Germany, seen as a European success story, with 96 deaths per million.
There are of course reasons to be highly sceptical about the official numbers of deaths in Bangladesh, and to ask if the reported numbers bear much resemblance to the real numbers.
The official death toll only includes those people who die in a Covid-19 designated hospital where the person has already tested positive for the virus. It leaves out those who die — whether at home, on the way to a hospital, or even in a hospital — after suffering from symptoms similar to those who suffer from Covid-19. Bangladesh Peace Observatory — which is part of the Centre for Genocide Studies at Dhaka University — has collated newspaper reports of 622 deaths between March 8th to May 16th. If this number was added to the official figure of 314 deaths on that date, this would triple the total number of Covid-19 deaths to 936. Interestingly, however, the number of these “excess deaths” is decreasing as the number of official deaths increase, and as the levels of testing have grown. While in the week of April 6th to 11th, there were 25 official deaths and 117 “non-official” deaths, in the week of May 10th to 16th, trend has reversed and there were 117 official deaths and only 58 “non-official” ones.
Even if these additional cases were added into the tally of Covid-19 deaths, the death rate in Bangladesh would be only six deaths per million of the population — though this does not take into account other similar deaths not reported in newspapers.
However, using numbers of unofficial Covid-19 symptom related deaths to question the official numbers assumes that all or most of those people would in fact have tested positive for the virus. Research by Bangla Tribune raises doubts about that assumption. The online newspaper looked at 180 of these deaths reported in a period between March 17th and April 14th. It found that in 52 cases no test sample was taken, but that in only 13 out of 128 cases, where a test was undertaken, it was found to be positive — a rate of only 10%. If accurate, this would suggest only 62 of the 622 excess deaths should be categorised as Covid-19 deaths.
Yet, there are also reasons why one should be somewhat sceptical about the results of this Bangla Tribune study. It can not be assumed, for example, that the sampling of those who have died was taken quickly enough after death, or that the sample was appropriately stored prior to being tested to ensure an accurate result. Nonetheless it is the only study looking at test results from these unofficial Covid-19 symptom related deaths.
One comparative statistic that does suggest Bangladesh’s official number may not be too far from the reality is the death rate in India — its neighbour and the country which most matches Bangladesh culturally as well as in terms of social-economic conditions. India’s death rate also hovers around two deaths per million. Indeed, other countries in South Asia have a similarly low death rate, in some cases much lower: Pakistan and Afghanistan are at four deaths per million, while Sri Lanka and Nepal are lower at 0.4 and 0.07 deaths per million respectively, suggesting that for whatever reason the virus is having a less virulent impact in South Asia than in Europe and North America — benefiting Bangladesh. It is also notable that the initial “lockdowns” in each of these countries took place at a much earlier moment in the epidemic than most countries in Europe or North America.
One other observation that might suggest that the toll of the Covid-19 epidemic has not hit Bangladesh anywhere as near as hard as elsewhere in the world is that the Covid-19 designated hospitals in the country do not seem to be under particular pressure. These hospitals have around 14,000 available beds, yet according to the Directorate General of Health Services (DGHS), on May 20th only 3,897 beds, less than a third, were occupied. However, some argue that one should treat the low Covid-19 hospital bed occupancy rate with some caution. They point out that the low occupancy rate may simply reflect the difficulty people have to geographically accessing the hospitals, the stigma attached to admitting Covid-19 illness, and that most of these hospitals only admit patients if they bring with them a Covid-19 positive test certificate, which is not always straightforward to obtain, as testing remains limited.
Assuming, for the sake of argument, that the official death numbers are close to the actual numbers of deaths, it is not time for the Bangladesh government to think that the death toll will inevitably remain low. Right now, the daily official death toll is on an upward trajectory. The first reported death was on March 18th and if we total up the officially reported deaths from each of the eight weeks since then, one can see the trend.
The MRC Centre for Global Infectious Disease Analysis which is based at Imperial College and has a close working relationship with the World Health Organization, forecasts a continued increase in the numbers of deaths in Bangladesh. Its country specific forecast suggests that there will be a steady increase in the daily numbers, so that by the beginning of June the daily official number of deaths will be around 27 deaths — though it could be as high as over 40 (see diagram below). In comparison, last week’s daily average of deaths was 17, with 21 being the highest number on a single day. Moreover, the MRC Centre also estimates that the Effective Reproduction Number of the virus in Bangladesh — the average number of new infections arising from each infected person — was as of May 17th, one of highest in the world at 1.65.
The number of deaths are of course still relatively low, but these can increase quickly. This is indicated by other modelling done by the MRC Centre, as reported earlier by Netra News, that projects, as of May 20th, that if the government were to relax its lockdown by half, by the beginning of June there could be an 18 fold increase in the daily number of infections from an estimated 3,612 to 65,843. This would in turn have a significant impact on the future numbers of reported deaths.
The Bangladesh government has already taken steps in recent weeks to open up the “lockdown”, through partial opening of garment factories, mosques and malls. The government is also due to end the “lockdown” — which it refers to as a national holiday — at the end of the month. If it does so, there remains a real risk that Bangladesh’s relatively small numbers of deaths could in just a matter of weeks move steadily towards the earlier projection numbers which right now seem to be overblown.●
?Netra News: Without Government Action, over 500,000 may die
?Netra News: UN Interagency Memo: Without interventions up to 2 million people may die
?Netra News: Imperial College – Sustained Bangladesh lockdown could reduce levels of death by upto 92 percent
?Bangladesh Peace Observatory: “Covid-19 Related Incident Mapping in Bangladesh”
?Dhaka Tribune: Only 10 percent of patients with died with Coronavirus like symptoms tested positive
?MRC Centre for Global Infectious Disease Analysis: Short-term forecasts of COVID-19 deaths in multiple countries
?MRC Centre for Global Infectious Disease Analysis: Situation Report for COVID-19: Bangladesh, 2020-05-20 (Please note this report is updated on a daily basis, so numbers will likely be different if you access this page on subsequent dates )