02 May Mapping the Virus: A Research Proposal
By Anjum Altaf
Pakistan, like many other countries, is flying dark in this Coronavirus pandemic. Very little is known for certain and what little is becoming known is not being factored into the decision-making. As a result, policies are are almost entirely reactive, based largely on fear, the happenings of the previous day, the push and pull of various influential lobbies.
Given this muddle, the most blunt and expensive set of measures built around a country-wide lockdown have been imposed in the country in a knee-jerk mimicry of actions taken in countries like Italy where the virus spread with great rapidity.
There was not much contextual analysis of how feasible these set of measures might be in a country like Pakistan. To start with, a complete lockdown is an impossibility, a truth that can be confirmed by just venturing out to any market on any particular day. This is inevitable in an economy where one or more members of over half the urban households have to go out to earn what they eat that day. In theory, this could have been mitigated if provisions to meet their daily needs could have been delivered to households deprived of earning opportunities. But such arrangements on the scale required are impossible in a society without any public safety net whatsoever.
Just as infeasible was the advice to maintain social distancing. The density at which the majority of urban residents live makes that impossible. Once again, one has to venture out on the street to see the extent to which the advice is being followed — three people on a motorbike makes a mockery of social distancing. Washing hands frequently requires water and soap, the buying of which for many households is a tradeoff at the margin with an item to eat. At the very least, such advice needed to be backed up with a free or subsidised supply of the commodities.
More than a month into the so-called lockdown, which does not preclude people using their personal transport to move within cities or between cities and provinces, there is absolutely no sense if the set of measures has had any impact on the spread of the virus. And, there is no way of assessing the relative contribution, if any, of the various measures in the policy package. As a result, there is no basis to decide if and when the lockdown ought to be relaxed and which measure is to be withdrawn first if a phase-out is to be pursued. Quite to the contrary, the relaxations are being driven by industry lobbies, i.e., construction (which should not have been restricted in the first place), or the threat of social groups, i.e., clerics (whose self-serving demand merited no consideration whatsoever).
In a society with a research-minded leadership, all this could well have been factored into an intelligent research design. For example, Kohat district which has just recorded three deaths to date, could have done without the lockdown and experimented with simpler measures like masks. If the infections had spiked, social distancing and/or a lockdown could then have been introduced. It is indeed ironic that, quite without thinking, this is exactly what has been done with the relaxation for Ramzan prayers in mosques under a set of operating procedures under the proviso that the dispensation would be withdrawn if infections spiked. Indeed, given the variation in the country, such an experimental approach could even have accommodated hypotheses linking the severity of the virus to the length of women’s sleeves or men’s trousers for that matter.
Add to all the above, no benefit is being derived even from the evidence that has become available from other countries. For one, it is quite clear that Covid-19 related deaths are predominantly among the very old and, even within this group, among those with other pre-existing conditions. For another, there are a number of countries that have not gone for the blunt and expensive lockdown strategy and have not had any the worse outcomes compared to those that have. This type of evidence should factor into a review of the policies put in place at the outset. For example, does it make sense to keep parks closed while mosques are open? And should elementary schools remain closed or could they open with social distancing teaching students through a two-days-a-week rotation?
What then is to be done in the situation we are in now related both to the progress of epidemic itself and a decision-making apparatus that is incapable of data gathering and analysis on any systematic or meaningful scale? Let us start with the two numbers that are displayed daily on the dashboards of virtually all our media outlets — the running total of the number of confirmed cases and the number of deaths resulting from them. The first number is not of much use since, in the absence of an adequate number of tests, it tells nothing about how many people might be infected in the country. Its only utility is that we can get a sense of the trend in whatever reality it is measuring, most likely a subset of the number of people going to hospitals with symptoms.
The second number is not of much greater use either. It is most likely the number of people expiring in hospitals whose death certificates record the cause of death as Covid-19. We don’t know for sure if the death was due to Covid-19 or of some other condition exacerbated by the virus. We don’t know the age, sex, or income level of those who have died. Nor do we have any clue to the number who might have died of Covid-19 at home without ever making it to the hospital. Once again, the number is useful only for indicating a trend in whatever it is measuring. Both the indicators are doubling to date every seven to eight days since the first Covid-19 deaths were reported on March 18 and the country-wide lockdown was imposed. These trends should be source of continued concern. But, by itself, this number of daily deaths (20 to 30) is no comparison, by many orders of magnitude, to the total number of deaths that occur in the country every day (approximately around 4,000).
What is really needed is some idea of the prevalence of Covid-19 in the country, its lethality, the extent of the immunity that has been built up, and the identification of hotspots, if any. Only then can there emerge a sensible and targeted strategy of isolation, quarantine, and healthcare facilities on the one hand, and opening up of some activities, like elementary schools and parks, on the other.
This however is too much to expect — we will simply not have serological testing at the scale required to ascertain prevalence nor the randomized testing to identify hotspots. We have to figure out what we can do with the very sparse data that is relatively easier to obtain and understand. [As a digression, I would like to mention that as a Professor of Economics and Dean, I was always intrigued by new faculty members returning from research universities in the USA and Europe with PhD degrees based on extensive and reliable data sets available there. Within a year they were complaining that they felt handicapped because similar time-series data of good quality was not available in Pakistan. It seemed to me that this was something they should have thought through before embarking on their dissertations and focused instead on getting the most out of the kind of sparse information available in Pakistan.]
In this regard, one good suggestion I came across in a newspaper was to monitor the number of patients with acute respiratory disorders admitted to ERs or ICUs of major hospitals. This statistic, collected from a nationally representative set of hospitals, could be compared with the corresponding statistic for the same time period in the previous year. It would yield a sense for what is different during the current epidemic and the extent to which the epidemic is manifesting itself in a higher case load which would be one reasonably good indication of how different is the situation as a result of the epidemic. Is it really a crisis, a minor bump, or just hype triggered by fear?
A second, equally simple, complementary indicator could be obtained by monitoring a nationally representative set of graveyards in the country. Tracking the number of burials per week (for which permission records are kept) and comparing them to the corresponding period a year ago would yield a sense of the lethality of the virus. It would correct for the possibility that many Covid-19 affectees are not showing up at hospitals and dying unattended and unrecorded. It would also capture the possible indirect impact of the lockdown that should not be ignored — that the number of deaths due to other diseases might spike because of either lack of medical care, non-availability of transport, inability to afford medicines, shortage of funds to buy enough food, or just plain deaths of depression and despair.
Of course, this statistic would need some adjustments. Just as the average number of deaths could increase because of the indirect effects of the lockdown mentioned above, they could also go the other way because of a likely decrease in the number of traffic accidents and deaths due to air pollution which has dropped dramatically since the lockdown. However, these are adjustments that can be made relatively easily by a competent graduate student. If there are excess deaths over the comparable period of a year ago, it would be a robust signal of the seriousness of the epidemic and help pinpoint hotspots. If, however, there are few excess deaths, it could indicate that even if the prevalence of infection is high the fatality rate is so low as to not be a cause for panic.
A continued monitoring of trends in the two statistics already being reported in the newspapers complemented by data on the two indicators suggested in this article should provide sufficient information to determine if the panic induced blunt and expensive suppression measures need to be continued or whether much less expensive measures like wearing masks and washing hands might be quite enough to contain the epidemic.
Continuous monitoring of these simple indicators would be enough to tell us fairly soon if this epidemic is really something to fear or whether it is just a more difficult version of the regular flu as some have begun to claim. If the former, harsher measures could be reintroduced later. If the latter, life could start returning to normal with basic precautions merited by common sense. The millions of despairing people out of jobs deserve a better and more informed set of policies.
The writer earned a PhD from Stanford University. This opinion appeared in Sindh Courier on April 30, 2020 and is reproduced here with the author’s permission.