Resurgence of SARS-CoV-2 in India: Potential role of the B.1.617.2 (Delta) variant and delayed interventions

Maxwell Salvatore, Rupam Bhattacharyya, Soumik Purkayastha, Lauren Zimmermann, Debashree Ray, Aditi Hazra, Michael Kleinsasser, Thomas A. Mellan, Charles Whittaker, Seth Flaxman, Samir Bhatt, Swapnil Mishra, Bhramar MukherjeemedRxiv 2021 · DOI

TL;DR

This paper investigates the resurgence of SARS-CoV-2 in India during early 2021, attributing it to the B.1.617.2 (Delta) variant's altered epidemiological properties and delayed public health interventions. The study highlights that timely interventions could have significantly mitigated cases and deaths, emphasizing the critical need for enhanced genomic surveillance and proactive pandemic response.

What Problem Does This Paper Address?

India, after successfully controlling its first wave of COVID-19 in 2020, experienced a devastating resurgence in early 2021. This second wave, characterized by a dramatic surge in infections and deaths, presented a significant challenge to public health systems. The paper aims to understand the drivers of this resurgence, specifically exploring the potential role of emerging SARS-CoV-2 variants, particularly the B.1.617.2 (Delta) variant, and the impact of the timing and effectiveness of public health interventions. It seeks to quantify the potential benefits of earlier and more stringent interventions in averting the scale of the crisis.

Methods

The authors employed a multi-faceted approach:

  • Two-strain, semi-mechanistic model: This model synthesizes mortality and genomic data to assess the epidemiological characteristics of different SARS-CoV-2 strains. It specifically compares the B.1.617.2 variant against previously circulating strains.
  • Extended SIR model: This model accounts for reinfections and waning immunity, allowing for a more nuanced understanding of transmission dynamics.
  • Predictive modeling of intervention effects: The study adapted a previous model to simulate the impact of different lockdown scenarios (strong and moderate effects) implemented at various time points in March and April 2021. This involved using real-world data on the effectiveness of past lockdowns in India and Maharashtra.
  • Sensitivity analyses: Multiple scenarios were run to assess the robustness of the findings to variations in key parameters like under-reporting and infection fatality rates (IFR).

Datasets

  • Reported infected cases and COVID-19-attributed deaths: Data was collected from covid19india.org for the period up to May 31, 2021 (for descriptive analysis) and May 15, 2021 (for predictive modeling).
  • Population data: Obtained from Our World in Data.
  • Genomic data: Acquired from GISAID, focusing on the temporal distribution of variants, particularly B.1.617.2 in Maharashtra.

Key Insights and Ideas

  • Delta variant's significant advantage: The B.1.617.2 (Delta) variant demonstrated a clear transmissibility advantage over previously circulating strains, even under scenarios of immune evasion. This suggests it was a major driver of the second wave.
  • Timing is everything for interventions: The study strongly indicates that early interventions, specifically in mid-to-late March 2021, would have been far more effective in curbing the spread of the virus than later interventions.
  • Healthcare system collapse amplified deaths: While the variant's properties were crucial, the sheer volume of infections, even with a lower case fatality rate, overwhelmed the healthcare system, leading to a significant increase in deaths.
  • Genomic surveillance is critical: The paper underscores the importance of robust genomic surveillance to track emerging variants and their evolving epidemiological properties.

Key Results

  • Wave 2 severity: The second wave was significantly more severe than the first, with nearly four times higher maximum daily cases and deaths, and a higher peak test-positive rate.
  • Delta variant's contribution: The model estimated that the B.1.617.2 variant accounted for a substantial portion of deaths, ranging from 15% in early April to 83% by mid-May 2021.
  • Preventable deaths and cases:
    • A moderate lockdown starting on March 15, 2021, could have prevented approximately 12.9 million infections and 97-109 thousand deaths by May 15, 2021.
  • Even a strong lockdown starting on March 15 could have prevented a similar number of infections (13.0 million) and 94-99% of deaths by May 15.
  • Lockdowns initiated in mid-April showed significantly diminished benefits, with some reductions not being statistically significant.
  • Transmissibility increase: The B.1.617.2 variant showed a transmissibility increase of 49% to 83% depending on the level of cross-immunity from previous infections.

Limitations and Future Work

The authors acknowledge several limitations:

  • Under-reporting: The study acknowledges that under-reporting of cases and deaths is not fully accounted for, particularly in rural areas, posing a challenge for accurate estimations.
  • Age-specific data: The models did not incorporate age-specific fatality rates, which could explain the lower overall infection fatality rate in Wave 2 due to a higher proportion of younger individuals being infected.
  • Vaccine roll-out: The analysis period predates widespread vaccination, and the models do not account for the impact of vaccine roll-out, which could influence transmission dynamics.
  • Intervention complexity: The comparison of intervention effects is against idealized lockdown scenarios, while the actual ground reality involved a mix of mitigation strategies.
  • Genomic data limitations: The exploration of B.1.617.2's properties was restricted to Maharashtra due to limited nationwide genomic data coverage.
  • Homogeneous mixing assumption: The model assumes homogeneous mixing within populations, which may not fully capture nuanced heterogeneities in spread.

Future work suggested includes:

  • More detailed collection of genomic and epidemiological data, including mortality, under-reporting estimates, and serological surveys.
  • Ramping up sequencing efforts to ensure representative samples for genomic and epidemiological analyses.
  • Strengthening surveillance in every district to identify clusters.
  • Exploring wastewater analysis as a complementary method for data validation.
  • Prioritizing sequencing of reinfections and vaccine breakthrough infections to understand vaccine effectiveness.
  • Conducting future analyses with reliable genomic, mortality, and seroprevalence data to capture spatiotemporal heterogeneity.

Summary updated

Abstract

Abstract India has seen a surge of SARS-CoV-2 infections and deaths in early part of 2021, despite having controlled the epidemic during 2020. Building on a two-strain, semi-mechanistic model that synthesizes mortality and genomic data, we find evidence that altered epidemiological properties of B.1.617.2 (Delta) variant play an important role in this resurgence in India. Under all scenarios of immune evasion, we find an increased transmissibility advantage for B.1617.2 against all previously circulating strains. Using an extended SIR model accounting for reinfections and wanning immunity, we produce evidence in support of how early public interventions in March 2021 would have helped to control transmission in the country. We argue that enhanced genomic surveillance along with constant assessment of risk associated with increased transmission is critical for pandemic responsiveness. One Sentence Summary Altered epidemiological characteristics of B.1.617.2 and delayed public health interventions contributed to the resurgence of SARS-CoV-2 in India from February to May 2021.
BibTeX
@misc{salvatore2021resurgence,
  author = {Maxwell Salvatore and Rupam Bhattacharyya and Soumik Purkayastha and Lauren Zimmermann and Debashree Ray and Aditi Hazra and Michael Kleinsasser and Thomas A. Mellan and Charles Whittaker and Seth Flaxman and Samir Bhatt and Swapnil Mishra and Bhramar Mukherjee},
  title = {{Resurgence of SARS-CoV-2 in India: Potential role of the B.1.617.2 (Delta) variant and delayed interventions}},
  year = {2021},
  doi = {10.1101/2021.06.23.21259405},
  howpublished = {medRxiv},
}
Figure