Battling Coronavirus outbreak: Covid-19 testing that is aligned with the times

In the current phase, testing should focus on community containment and reducing mortality. Increasing visibility on testing data is more important now than ever

By Ramanan Laxminarayan

As India moves into the top ten in the world in terms of number of cases, an inevitability given its large population, the question is where we go next. The lockdown has only delayed the epidemic peak. The size and shape of the peak, which tells us how many cases will flood the healthcare system during a short time window, is determined by how well we are able to slow down transmission in the absence of a lockdown. And, Covid-19 testing will assume greater importance for transmission reduction in the absence of a lockdown.

We need better data and clarity around testing. Some point out that only 5% of tests in India are positive, lower than in other countries. Could this mean that we are testing enough? To me, the low positivity rate is likely due to a combination of factors, including an inefficient testing strategy since the outset, and an incorrect interpretation of the numbers. If we were able to target high-risk populations accurately, through pooled testing, it is likely that each individual test would be far more productive at identifying Covid-19 positive cases than the current approach. Also, the current test count includes tests that are used to determine whether a patient is no longer Covid-positive.

Including this “recovery” testing artificially reduces the test positivity rate. To correctly calculate test-positivity rates, we have to divide the number of patients identified as being Covid-19 positive by the number of patients tested. However, these data have not been officially reported for a while now. Increasing the visibility on testing data is more important now than ever. Providing transparent data to the public and researchers on cases, without revealing any personal identifiers, can enable people to self-protect, which is ultimately the only way in which this epidemic will be controlled, until a vaccine arrives.

We also need to rethink our testing strategy to be consistent with the containment strategy at that point in the epidemic curve. Testing has two objectives—to identify those who are sick to isolate and treat them, and to keep track of the epidemic and anticipate new cases and deaths. To do this, we have two methods—real-time, reverse transcriptase PCR (RT-PCR) tests that look for the virus and tell us whether a patient currently has an infection, and antibody tests that tell whether a person has been infected in the past, even if they are no longer carrying the virus. We have to be clear about which of the two methods we are using, and how it helps our twin objectives of early patient identification and treatment, and disease surveillance.

In the early stages of the epidemic, it was important to widely deploy the PCR tests in high risk groups, with the objective of individual containment or slowing down the epidemic. However, during this period, testing levels in India were extremely low, which contributed to the epidemic spreading widely without anyone knowing about it.

During the second phase of community or cluster containment, it was important to carry out a combination of pooled RT-PCR testing and antibody testing, with the objective of correctly identifying neighbourhoods where the disease was spreading. During this phase, RT-PCR testing did ramp up in India, but the notion that community transmission had not begun hobbled the testing strategy. Moreover, the deployment of antibody testing could have helped correctly identify communities to be targeted for containment. Although the number of tests increased by a factor of 50 during this period, many of these tests were likely used to confirm recovery in patients who had the disease than in identifying new cases.

During the third phase, which is where we are at the moment, testing should focus on both community containment and on reducing mortality. Even if case fatality rates in India are only a quarter of what they are in some other countries, there is still reason to worry because of its large population. Current ICMR testing criteria make no mention of age as a testing criterion.

However, at this stage, we should make sure that every person above the age of 65 who has any signs of Covid-19 symptoms is tested immediately, tracked, and checked for markers of progression to severe disease, including of checking their oxygen saturation levels. Testing can no longer be a reliable means of carrying out case counts since that will require millions of PCR tests a day, which is likely beyond the total capacity of public and private sectors in India at this time. This capacity is difficult to expand at a time when other countries are competing for test equipment and kits.

Going forward, the way to keep track of infection loads is through weekly serological studies using antibody tests like ELISA (enzyme-linked immunosorbent assay) rather than through daily reports from RT-PCR tests, which could be misleading if the aim is to assess community burden. The capacity to carry out antibody tests exists in India, in both public and private sectors, without the need to import rapid tests from China.

But, thus far, there has been very little systematic effort to correctly assess the scale of exposure to the virus using serological surveys. Bi-weekly surveys can be helpful in assessing the scale and spread of the epidemic, and can help limit containment efforts to where they are absolutely needed.

More important than case-counts is the number of Covid-19-associated deaths. Deaths are the surest way of knowing where the epidemic is spreading since each Covid-19 death means that there were probably around 100 infections two to three weeks prior to that date. While states have an incentive to under-report deaths, this could backfire by blinding public health authorities, and the public, to where new cases and deaths are likely to emerge. By undercounting deaths, we are depriving ourselves of valuable information that could enable us to reduce future deaths. The public relations gains may be short-lived as the epidemic curve grows and makes today’s undercounts irrelevant when compared to future daily mortality rates.

The two-month lockdown bought valuable time for preparation and testing, but it has served its purpose as a temporary Covid-19 containment strategy. Going forward, there needs to be a clear testing strategy focused on reducing mortality. This needs to be in place across the country so that the highest-risk groups, starting with those with comorbidities above the age of 65, can be prioritised for testing, and hospitalisation, if needed. We may have won the first round of a 12-round boxing match, albeit at the large cost of an extended lockdown. But, there are 11 to go and we can be sure that the virus will exploit any strategic errors on our part.

Director, CDDEP, and senior research scholar, Princeton University. Views are personal