A study by two India-origin researchers at the University of Cambridge, is based on the SIR approach, but takes into account age and social contact structure to assess the impact of the lockdown and social distancing measures adopted before it.
Against the backdrop of the three-week lockdown, and 3,557 people testing positive for Covid-19 in India (as on Sunday), findings of three India-focussed Covid-19 modelling studies have been discussed in commentaries in the media.
The earliest of these — by researchers at the University of Michigan, Johns Hopkins University and Delhi School of Economics, dated March 22 (two days before the lockdown began) – is a basic susceptibility (S), infection (I) and recovery (R) model, which projects, based on Johns Hopkins University (till March 18) and Indian census data, a very narrow difference between the case counts per 100,000 population throughout March and even early April in a business-as-usual (no intervention) and a lockdown scenario.
However, it projects 161 cases per 100,000 population by the third week of May in the ‘no intervention scenario’ versus a 1 case per 100,000 in the case of a lockdown. The case counts are adjusted for the overall population based on data for states that had reported high Covid-19 incidence when the study was being done. The study, however, advocates wider testing, effective quarantining of infected cases and targeted restrictions over a blanket isolation move like the lockdown, given the economic impacts of such a move.
The second, a study by two India-origin researchers at the University of Cambridge, is based on the SIR approach, but takes into account age and social contact structure to assess the impact of the lockdown and social distancing measures adopted before it. This study finds that the impact of a 21-day lockdown may not be much, given there will be resurgence in infection within the next 3-4 weeks. However, a three-phased lockdown (21-days/28 days/18 days, with a 5-day partial lifting between the phases) and a single, continuous lockdown of 49 days will likely help India fight Covid 19 spread most effectively. With a R0 of 2.1 (every infected individual passes on the infection to a further 2.1 individuals), in a no mitigation effort scenario, the study estimates a peak case load of 150 million by the end of June, assuming all cases of infection are infective. The authors clarify the numbers are “counterfactuals” as social distancing and lockdown measures have been undertaken.
The third (revised and accepted March 31, available online April 2), a pre-proof study by researchers from the Armed Forces Medical College, Pune, and INHS Asvini, Mumbai, paints the most optimistic scenario of the lockdown. It estimates, without any non-pharmacological interventions (NPIs), India would have see over 364 million cases and 1.56 million deaths by mid-July. This model estimates a bending of the incidence curve as effective quarantine of infective cases reaches 50% or more with a lockdown – if implemented in the last week of March, this could push total active infections by May 25 to 2.4-5.7 lakh with a Covid-19 hospitalisation demand of 10,000-12,000.
Against this projection, an April 2 modelling by the Center for Disease Dynamics, Economics and Policy, talks of a total hospitalisation demand of 4.75 lakh across the country in a scenario where 0.5% of the population ends up infected in the current outbreak.
While all these models will help the country design an appropriate policy response, there are many uncertainties and variables (for instance, the impact of BCG vaccination on transmission that is being studied at the moment) that can’t be built into models.