Stop driving in the dark on Covid-19 spread

IN PERSPECTIVE

As lockdowns continue to be the most widely-resorted-to response strategy to the threat of Covid-19, governments the world over face a difficult trade-off between saving lives and saving the economy. In India, the lockdown has imposed debilitating economic hardship, borne disproportionately by migrant and informal sector employees, most of whom have weak safety nets at best.

Data will be critical to decision-making even in these unprecedented times. We have tested approximately 200,000 individuals for Covid-19 or approximately between 15,000-20,000 individuals per day since mid-March. However, our current set of data is neither random (as only those with a travel history, symptoms or both are eligible for testing), nor sufficient, as our capacity lies somewhere between 150,000-200,000 tests per day, across 255 public and private laboratories. To address these shortcomings, the government needs to leverage the extended lockdown period to collect a large, representative, and random risk-weighted sample of data, which it can utilise to learn about the spread of the pandemic and to help it make an informed decision about a gradual relaxation of the lockdown.

The Indian Council of Medical Research (ICMR), in a highly positive move, recently announced scaling up to 100,000 tests per day. However, moving beyond symptomatic testing of individuals to population-based testing will require a more comprehensive strategy. This includes deeper integration with the Ministry of Health and Family Welfare’s (MoHFW) ground-level containment strategy and better coordination with the larger scientific and research community to analyse the data.

Leading the charge on the ground, the MoHFW is following a “cluster containment strategy” to isolate geographic areas reporting one or more positive cases. Local authorities in Mumbai for example, have identified 382 containment zones across 24 administrative wards. While this strategy is intended to contain the spread of the infection, it is not clear that it will, because it relies on one or a group of individuals first testing positive. In this current climate, it is not clear how one gets a prescription or reaches a testing centre. Moreover, from the limited data available, we already know that around 70% of the positive cases reported in Maharashtra were asymptomatic.

A national-population based testing strategy will involve segmenting the entire country into non-overlapping clusters; determining a representative sample of the population within each cluster; screening and testing the sample on a periodic basis for Covid-19, using RT-PCR tests, potentially on pooled samples; and finally using statistical tools to infer prevalence rates with relevant error bands. This would need the government to collaborate with established statistical and scientific research institutions to outline a representative sampling strategy along with screening and testing protocols.

India has extensive experience in collecting primary data. The National Family Health Survey (NFHS), for example, is a nationally representative survey, conducted every two years. Its findings, used to inform India’s national health policies, include prevalence of anemia and diabetes, for instance, which requires door-to-door collection of blood samples. We have the expertise, it’s up to us to channelise it towards a strategic response. India’s large network of ASHA workers and primary healthcare providers can be trained to conduct screening and testing activities.

A number of governments are contemplating population-level testing measures to better understand how their populations are affected. Germany, for example, recently announced plans to implement a nation-wide antibody testing programme to learn how many Germans are immune to the virus, the progression of the virus, and to eventually decide how to phase out their lockdown which has been in place since March 22.

As with any policy recommendation, this comes with its limitations. The biggest constraint to scaling-up testing will be imposed by a shortage of testing kits given the surge in global demand. To relax this, there is a need to accelerate the approval process for domestically manufactured as well as imported kits, including both the PCR and rapid anti-body varieties. There is also a need to make the approval process more transparent and streamlined, thus improving incentives to participate for India’s advanced biotechnology sector. Finally, the current strategy of empaneling a large number of private and public labs could be coupled with experimenting with pooled testing methods, and utilising tuberculosis testing machines which are bound to have a wider geographic spread.

To questions on whether sampling and testing of the kind suggested here are feasible at all in the current scenario, we must note that the notion of large is variable. It has to be large-enough depending upon the risk-base and the availability of test-kits. The smaller the sample, the wider the error bands, but the approach would have to be of the type described here, failing which we will simply not have enough information about when and how to open up the economy. What we recommend is the most parsimonious way of using the available test kits, as opposed to the ad hoc individualised testing currently implemented, which leaves us entirely in the dark about the extent to which the virus has spread.

India should be prepared for a long and arduous journey towards a complete recovery. The worse-hit amongst us globally, like the Chinese and the Italians, have experienced either complete or intermittent lockdowns for over three months now. Committing to short-term, ambitious goals like implementing population-based testing measures can be a first step, enabling the government to make evidence-based decisions about a phased re-opening of the economy.

(Madraswala studies Economics and Public policy at the Harvard Kennedy School of Government; Mor is a Visiting Scientist at The Banyan Academy of Leadership in Mental Health) (The Billion Press)