|  | 


Covid-19: To get an accurate estimate of spread, India must guarantee community randomized testing: analysis


In the fight against the coronavirus pandemic (Covid-19), policy makers fight blindly, because we do not know the extent of infections in the population or the death rate due to them. It is possible to track the number of deaths, cases that required hospitalization, and those that tested positive for the virus in hospitals, but those numbers present a limited picture of the pandemic. Our response to this crisis depends critically on how fast the disease is spreading outside of hospitals, in the community, and how likely the virus is to kill people who have been infected. The only way to obtain this information is through randomized population tests.

Existing methods of measuring disease prevalence are likely to give inaccurate results. India, like most countries, has a shortage of test kits. As a result, it prioritizes testing for people who have severe symptoms and present in the hospital. He misses those who have symptoms, but are not severe enough to go to the hospital, and those who are asymptomatic (lack of cough, fever, or breathing problems), but are nevertheless carriers who can infect others.

There are good reasons to think that many people are asymptomatic, which means that the true workload is undervalued. For example, the city of Vò (Italy) examined all its inhabitants and found that up to 50% of people were asymptomatic. On the Diamond Princess cruise ship, 18% of the infected population showed no symptoms, while in Iceland, 50% of those who tested positive were asymptomatic. A report recently released by Maharashtra suggests that 85% of cases in the state were asymptomatic, similar to recent Chinese findings published in The British Medical Journal. They suggest that we will never be able to measure the true prevalence of Covid-19 by testing only high-risk populations.

We lack the essential information to combat this disease due to such data limitations. Harvard economist James Stock has identified the lack of accurate infection rates as the central information gap behind efforts to combat the coronavirus. Consider the true mortality rate due to Covid-19. This is calculated by taking the number of deaths from the virus and dividing it by the total number of infections it has caused. For any level of Covid-19 deaths, the higher the infection rate, the lower the death rate. The death rate is, in turn, essential to gauge our response. If the death rate from the virus is very high, duplication of suppression methods is required to avoid mass deaths. A lower mortality rate does not mean that the situation is less problematic, but it does tell us that the disease is more prevalent in the population than currently thought, and requires a mitigation approach rather than suppression in response.

While we have estimates of the number of deaths, we are flying blind in estimating the denominator of the death rate. If, in fact, many people are walking on Covid-19 and may not even know their infection status, as some doctors suggest, we must take a very different approach to address this crisis. At the time of writing this article, India has 4,067 coronavirus cases and 109 deaths, which implies a mortality rate of 2.7%, without considering the future mortality of the currently ill. If the fraction of asymptomatic cases were 20%, this would imply a real mortality rate of 2.1%. However, if it were more than 80%, the actual death rate would be considerably lower by 0.5%. India’s location on this mortality spectrum will greatly influence how we think and combat this disease, and that is why estimating this number accurately is an urgent priority.

Accurately measuring infections will also allow specific efforts to combat the spread of the disease. Otherwise, we won’t know how many hospital beds, fans, and doctors we’ll need in different areas in the coming weeks. We cannot determine if the blockade has been effective, how long it should continue, or whether certain locations can safely be released from the blockage without causing an outbreak.

To their credit, the Indian Council for Medical Research (ICMR), at the start of the pandemic, examined 826 patients admitted to hospitals with severe acute respiratory disease but no travel history. This was a valuable effort that demonstrated that, in the early stages of the pandemic, the coronavirus had not yet spread throughout the community in India. The situation has changed since then. There is now a great deal of evidence in all countries, including India itself, that Covid-19 can be found not only in those with severe symptoms, but also in people who show few or no symptoms.

While ICMR has just expanded its testing criteria beyond symptomatic cases to include asymptomatic individuals with previous contacts with confirmed cases and antibody testing at critical points, these efforts have yet to reach actual random testing across the country.

The only way to determine the actual mortality and infection rates is through randomized community testing: identify individuals in the population at random, rather than in and around hospitals and hot spots, and testing them in the field to that we can determine the precise prevalence. Closing this information gap will be critical to making informed decisions about the best way to combat this disease. Countries like Austria and Germany have started random population testing. India must do so immediately, either at the national level or, if capacity is a constraint, at least at the sub-national level.

Arpit Gupta is an assistant professor of finance at the NYU Stern School of Business. Anup Malani is a professor at the University of Chicago School of Law and the Pritzker School of Medicine, and Reuben Abraham is CEO of the IDFC Institute, Mumbai.

The opinions expressed are personal.

Original source