Transmission dynamics of the COVID-19 epidemic in India and modelling optimal lockdown exit strategies

Abstract

Background: The coronavirus disease 2019 (COVID-19) has caused over 3 200 000 cases and 230 000 deaths as on 2 May 2020, and has quickly become an unprecedented global health threat. India, with its unique challenges in fighting this pandemic, imposed one of the worlds strictest and largest population-wide lockdown on 25 March 2020. Here, we estimated key epidemiological parameters and evaluated the effect of control measures on the COVID-19 epidemic in India. Through a modelling approach, we explored various strategies to exit the lockdown. Methods: We obtained data from 140 confirmed COVID-19 patients at a tertiary care hospital in India to estimate the delay from symptom onset to confirmation and the proportion of cases without symptoms. We estimated the basic reproduction number (R0) and time-varying effective reproduction number (Rt) after adjusting for imported cases and reporting lag, using incidence data from 4 March to 25 April 2020 for India. We built upon the SEIR model to account for underreporting, reporting delays, and varying asymptomatic proportion and infectivity. Using this model, we simulated lockdown relaxation under various scenarios to evaluate its effect on the second wave, and also modelled increased detection through testing. We hypothesised that increased testing after lockdown relaxation will decrease the epidemic growth enough to allow for a greater resumption of normal social mixing thus minimising the social and economic fallout. Findings: The median delay from symptom onset to confirmation (reporting lag) was estimated to be 2·68 days (95% CI 2·00−3·00) with an IQR of 2·03 days (95% CI 1·00−3·00). 60·7% of confirmed COVID-19 cases (n=140) were found to be asymptomatic. The R0 for India was estimated to be 2·083 (95% CI 2·044−2·122 ; R2 = 0·972), while the Rt gradually down trended from 1·665 (95%CI 1·539−1·789) on 30 March to 1·159 (95% CI 1·128−1·189) on 22 April. In the modelling, we observed that the time lag from date of lockdown relaxation to start of second wave increases as lockdown is extended farther after the first wave peak. This benefit was greater for a gradual relaxation as compared to a sudden lifting of lockdown. We found that increased detection through testing decreases the number of total infections and symptomatic cases, and the benefit of detecting each extra case was higher when prevailing transmission rates were higher (as when restrictions are relaxed). Lower levels of social restrictions when coupled with increased testing, could achieve similar outcomes as an aggressive social distancing regime where testing was not increased. Interpretation: The aggressive control measures in India since 25 March have produced measurable reductions in transmission, although suppression needs to be maintained to achieve sub-threshold Rt. Additional benefits for mitigating the second wave can be achieved if lockdown can be feasibly extended farther after the peak of active cases has passed. Aggressive measures like lockdowns may inherently be enough to suppress the epidemic, however other measures need to be scaled up as lockdowns are relaxed. Expanded testing is expected to play a pivotal role in the lockdown exit strategy and will determine the degree of return to normalcy that will be possible. Increased testing coverage will also ensure rapid feedback from surveillance systems regarding any resurgence in cases, so that geo-temporally targeted measures can be instituted at the earliest. Considering that asymptomatics play an undeniable role in transmission of COVID-19, it may be prudent to reduce the dependence on presence of symptoms for implementing control strategies, behavioral changes and testing.

Predicting COVID-19 spread in the face of control measures in West Africa

“…Finally, our study suggests that mass application of control measures coupled with appropriate testing and timely isolation of asymptomatic humans will go a long way in keeping disease numbers low, which is consistent with results in [41] . However, with the fragile economic conditions and health care systems in the region, interventions like complete lock-down as was there in some European countries, most US states, and China for a couple of months is not feasible [40] and can be more catastrophic than the disease itself.…”

“… [39] for three US states and the entire US and Gupta et al. [41] for India suggest that increased detection through testing can reduce the burden of the disease. Although, contact tracing is already being implemented in West Africa, the measure is not very effective due to the lack of resources and high community transmission in the region.…”

“…We obtained a basic reproduction number of 1.854 (CI ), which is relatively low compared to values recently reported for two West African countries—Nigeria (2.29) and Senegal (2.36) [40] . Also, a basic reproduction number of 2.083 (CI ) was reported for India in [41] . The control reproduction number obtained for West Africa was 1.37 (CI: ], which is close to the value reported for South Korea (1.5) [42] but lower than the value reported for Wuhan one week before travel restrictions were introduced (2.35) [43] .…”

Comprehensive public health evaluation of lockdown as a non-pharmaceutical intervention on COVID-19 spread in India: national trends masking state-level variations

“… 11 Similarly, Gupta estimated key epidemiological parameters and evaluated the effect of control measures on the COVID-19 epidemic in India and its states using a dynamic compartment-based susceptible-exposed-infected-removed (SEIR) modelling approach, reiterating that state-specific R 0 values exhibit high variability with respect to the national value of R 0 . 12 However, much is left to be done now that the nationwide lockdown has ended and a targeted lockdown phase is ongoing. All epidemiological projections suggest that current gains may be reversed rapidly if air travel and social mixing resume.…”

Transitivity of COVID-19, based on parameters R_0 and R_((t)) in Pombos/PE municipality, Brazil

“…To determine the mean and standard deviation, it was chosen the parameters found in Ainslie et al (2020); Nishiura, Linton, & Akhmetzhanov (2020); and Gupta et al (2020), which are determined using Monte Carlo's simulations. Using minimums and maximums established in the previous parameters, uncertain parameterization was performed to obtain more exact results.…”