COVID-19: points to ponder

Local ward council workers inform residents of Purba Rajabazar in Dhaka on PA system about dos and don’ts of the lockdown enforced in the area. — New Age/Sourav Lasker

THE Barisal division had 297 COVID-19 patients by the end of May; the number shot to 2,609 by June 28. This is 2.93 times higher every 14 days. In Bogura, the number of cases was 216 till May 31; this increased to 2,782 by June 29, indicating 3.6 times higher every two weeks. In Khulna, COVID-19 patients were only 76 on May 31. In a month, it increased to 1,551, or 4.5 times higher every two weeks. Barisal probably has the lowest immigration and emigration while Bogura is known to have experienced a high number of immigration compared with the districts in the Barishal or the Khulna division. These figures are too wide to say that they surrogate a basic reproduction number. It is an oversimplification and not an accurate process of estimating R0.

In the 12th week, May 24–30, since the first reported COVID-19 cases in Bangladesh, the daily number of cases had been 1,790. Two weeks later, in June 7–13,, it was 3,056 — about 1.7 times higher within a gap of one incubation period. In the 16th week, June 21–27, the number of cases was 3,600, which is 1.18 times higher than that of the 14th week. Are these correct reflections of R0? No, because cases were not decreasing and, rather, started increasing after the 12th week. If the R0 estimations at the district level seem unacceptable, why should R0, estimated in the same manner for the national level, be acceptable?

How to estimate R0

FOR R0 estimation, we need to identify index cases, in a defined static population in a defined geographical or administrative area to begin with, where people neither enter from outside nor leave it. The next step is to find out the people who came in contact with index cases and follow the contacts for about two weeks, the incubation period of COVID-19, assuming that everyone in the population is susceptible theoretically, to see if the contacts are infected — with or without any clinical features within 14 days. The no-in migration or no-out migration in the defined population and geographical or administrative area needs explanation. Emigration would reduce the number of people who either had a chance of developing the infection in the defined area or have already been infected but could not be measured for their infection status because of their emigration. While immigration would swell the number of infected people, who might have contacted it from the place where they moved in from, but without prior information on the status of infection and were not infected from index cases. A more practical proposition would be, however, to exclude those who moved in and show clinical features before 14 days and follow those up who moved out for 14 days.

The other reasons we cannot estimate R0 in Bangladesh is that the present figures available do not count all those who might have been infected actually, eg, infected without clinical features and infected with mild-to-moderate clinical features. Unless and until we know the full expanse of infection, how can we say how many people an index case transmitted the infection to within the incubation period of the disease?

Purba Razabazar lockdown

The 2011 census put the population of the neighbourhood in ward 40 u8nder the Tejgaon police jurisdiction area as 29,820. Based on the rate of increase of population in urban areas in Bangladesh, the projected population of Purba Razabazar is supposed to be 40,045 in 2020. This is about 20 per cent of the population of ward 40, ie only 0.22 per cent of the population of the Dhaka city.

It has been claimed that because of the lockdown, the number of cases fell at Purba Rajabazar. This begs a fundamental question: what is the aim of lockdown? The answer is to keep other people safe from the lockdown area. So, the effect should rather be measured in terms of how much the infection rate fell or did not increase among people outside Purba Rajabazar who were not in areas in lockdwon. This bit of information is not known to us. We do not know what the infection rate was at Purba Rajbazar to begin with. Infection among 60 per 100,000 people, applied to colour a location red, has been based on the infection among people with considerable clinical features of COVID-19 and their contacts which was, and still is, until the last bit of information that we have come across, only one per patient on an average which a figure hardly believable. One might argue that the way the infection is estimated now gives clue to the overall incidence. But we are not certain unless a population-based estimation is done on several locations at least once. One additional point in this regard is how locations with different administrative and population sizes may be compared with each other even if in terms of rates as the transmission potential should be different based on these differentials alone.

The fact that those who were without any overt features or who had mild symptoms do not appear in the estimation process proves that the estimation process does not give any incidence rate. Besides, were there not other areas where 99.88 per cent of the population of the city live which could also be coloured red at the same time? How can one expect that just by locking down only one of these areas will prevent the spread of the disease to other parts of the city?

Cases of COVID-19 fell at Purba Rajabazar; so was claimed. Nonetheless, it will be interesting to know if the fall was significant in statistical terms. We do not know. These questions and the issue of the actual incidence in the area will remain unanswered forever as the actual expanse or percentage of infection in the area was unknown ab initio. By the way, did the rate of COVID-19 infection not fall or at least not increase remarkably in other parts of the city? Was it due to the lockdown of a small 0.22 per cent of population of Dhaka city only?

When might a disease fall in a population?

HOW may the fall of a disease on a location be explained? The non-possibility of exposure of the people of a location, eg Purba Rajabazar, from extraneous sources, which, however, should be insignificant, as people from other areas of the city supposedly had a much lower infection rate with much lower possibility to transmit the virus to an area which supposedly had already a much higher rate. The other possibility is that everyone in the lockdown area became immune during the lockdown period, which, to be true, would mean that everyone was infected within the area due to interaction between them, chances are slim though. Yet another is that the virus died down within those who were infected and that it did not infect uninfected people in the area. This would be possible only if those who were ill from COVID-19 remained in isolation or the suspected ones kept them in quarantine and everyone else practised personal protective measures meticulously, which protected them from asymptomatic or mildly symptomatic carriers. Needless to say, this later measures could be equally effective if adopted in any other areas without lockdown. There may be yet another possibility that by the time the lockdown was applied, too many people were already infected and not many were left to be infected in the area, knowledge of which is, however, beyond our reach. A theory to be acceptable, namely, lockdown reduced the infection rate at Purba Rajabazar, a feasible explanation should be in order.

Fundamental effect of lockdown

ONE aim of lockdown is to keep the transmission of a disease under leash to buy time, which notwithstanding also reduces the number of infection, but does not bring it down to zero. The number starts climbing after a time to some extent, after the lifting of the lockdown. Could the pocket of new cases at Purba Rajabazar be a testimony to this fact, as is seen in many other countries?

 Fawning of reasoning

THE basic question at the end: how does putting a population of only 0.22 per cent in a city, with so much of population mobility, so many not bothering about personal protective measures and practices, can effect a fall in infection in other parts of the city? Are there not other sources and avenues of engendering infection? What will happen to the people of Purba Rajabazar who are free to mingle with other people now with a new found confidence that they are infallible? If all of them have really become immune already, there is no concern. But what if they are not? Who will ensure that those who were not infected at Purba Rajabazar already will neither get infected nor infect others in city areas which are not in lockdown. A two-way domino effect, in fact, is in the offing, now uninfected, residing in lockdown area, getting infected from non-lockdown areas after lifting the lockdown and vice versa.

Many states and countries needed re-imposition of lockdown. To be effective, lockdown has to be applied simultaneously to all applicable places based on an accurate rate and size of incidence.

The final question is whether we have compared the effect of lockdown with the adoption of strict personal practices and measures including quarantine and isolation?

AM Zakir Hussain is a former director, Primary Health Care and Disease Control, former director of IEDCR, DGHS, former regional adviser of SEARO, WHO, former staff consultant, Asian Development Bank, Bangladesh and Working Group member of Bangladesh Health Watch (BHW).