Covid19 and data allergy

A friend once said that life in TT sometimes feels like one divided by three, you know, 0.33 repeated ad infinitum; a recurring decimal. It is the same errors made constantly. One of these is our apparent aversion to data-driven decision-making – a “data allergy.”

In the middle of a global pandemic, when decisions are incredibly consequential, this “data allergy” is proving to be a noose around all of our necks.

There are clear questions that need to be answered, clear trade-offs that must be made, and we seemingly have little desire to make data-informed decisions. Seven months into this pandemic, we are still acting as if it were March, and our knowledge and the scientific tools of enquiry at our disposal were still very limited.

Let us ask ourselves a straightforward question:

What is the extent of spread of SARS-CoV-2 locally?

In a mid-September press conference, we were told that 99 per cent of our population was still “immunologically naive,” meaning not yet exposed to the virus and therefore susceptible to infection. One has to assume that this pronouncement was based on the number of confirmed cases at that point in time, and therefore almost certainly a grossly inaccurate claim.

According to the commentary Global Perspective of Covid19 Epidemiology for a Full-Cycle Pandemic, published in the European Journal of Clinical Investigation: “underestimation (of infections) may have been 50 to 100-fold or more in the early days of the pandemic, especially in locations with limited (PCR) testing...as of summer 2020 underestimation was apparently still 11-fold in the USA and about 30-fold in India.”

On October 5, the WHO’s best estimate was that one in ten of the global population, roughly 780 million, had already been infected, although the official case count stood at 35 million.

The trend is clear: the less PCR testing (testing for active infection) done on a per-capita basis, the less the full extent of viral spread is captured. Even today, the US and India have done approximately 18 and three times more testing per capita than TT respectively.

It is within the realm of possibility that our actual number of total infections to date is 50 times higher than our official count of 5,600, that is, roughly 280,000, or one in five people.

While many have been clamouring for more PCR testing, not enough have been asking the more consequential question: why are we yet to do a single serology (antibody) study on a random representative sample of the population?

Testing for antibodies is exactly how the WHO, and other nations – including ones as small as Iceland and Qatar, and with far less per-capita wealth than TT like Pakistan and Nigeria – have been able to get a better idea of the extent of spread and their local infection fatality rate (IFR).

Viral spread and the effective IFR of the virus vary by country. Susceptibility, while driven by age and comorbidities, is also a complex interaction of other factors, including living conditions, immune health (which can be affected by serum Vitamin D levels), and some degree of pre-existing immunity/resistance conferred by prior exposure to other circulating coronaviruses.

While global averages can give us a range of possibilities, we cannot boldly proclaim to be “following the science” when we have shown little interest in discovering what the science can say about our specific circumstance.

If the suggestion of 280,000 local infections is true, that holds huge insights and contextualises the nature of the threat that we are faced with. It implies an IFR of 0.04 per cent; an average of four deaths for every 10,000 infected.

Is this an unlikely number? Maybe, maybe not. Singapore has a case fatality rate (CFR) of 0.05 per cent. Its true IFR is likely lower, as the confirmed case count likely still underestimates the true number of infections. A recent study published in the bulletin of the WHO gives IFRs of 0.01 (Qatar) and 0.07 (Karachi, Pakistan and Mumbai, India) per cent. In other words, a local IFR of 0.04 per cent may be possible.

One would think that a low IFR and high existing spread would warrant a different public health response than a high IFR with little existing spread.

Every life cut short is a tragedy. That same compassion should remain consistent regardless of cause. According to data from the World Bank, TT had a crude death rate of 8.375 per 1,000 in 2018, that is, on average, around 30 deaths a day from all causes.

In making decisions today, it may be wise for us to consider how attempts to suppress SARS-CoV-2 may affect other causes of death today, next year, and long term.

For example, one study (of which there are many) by researchers from the Federal Reserve Bank of Chicago and Columbia University, stated, “We find that a job loss leads to a 15-20 per cent increase in the probability of dying in the 20 years following a job loss.” For a 40-year-old, that equals a loss of one to 1.5 years in life expectancy.

One would think that having an up-to-date figure on the unemployment rate (which is greatly exacerbated by varying lockdown measures) would be critical as an input into our decision-making. Unfortunately, our last official figure is two years old – “data allergy.”

Ryan Darmanie is a professional urban planning and design consultant, and an avid observer of people, their habitat, and the resulting socio-economic and political dynamics. You can connect with him at darmanieplanningdesign.com or e-mail him at ryan@darmanieplanningdesign.com