key: cord-0254091-lb6uvpnu
authors: Hart, W. S.; Miller, L.; Andrews, N. J.; Waight, P.; Maini, P. K.; Funk, S.; Thompson, R. N.
title: Generation time of the Alpha and Delta SARS-CoV-2 variants
date: 2021-10-23
journal: nan
DOI: 10.1101/2021.10.21.21265216
sha: 33907aee19206b681b2e34ce5b3e531341a2f4c5
doc_id: 254091
cord_uid: lb6uvpnu

Background: In May 2021, the Delta SARS-CoV-2 variant became dominant in the UK. This variant is associated with increased transmissibility compared to the Alpha variant that was previously dominant. To understand ongoing transmission and interventions, a key question is whether the Delta variant generation time (the time between infections in infector-infectee pairs) is typically shorter-i.e., transmissions are happening more quickly-or whether infected individuals simply generate more infections. Methods: We analysed transmission data from a UK Health Security Agency household study. By fitting a mathematical transmission model to the data, we estimated the generation times for the Alpha and Delta variants. Results: The mean intrinsic generation time (the generation time if there had been a constant supply of susceptibles throughout infection) was shorter for the Delta variant (4.6 days, 95% CrI 4.0-5.4 days) than the Alpha variant (5.5 days, 95% CrI 4.6-6.4 days), although within uncertainty ranges. However, there was a larger difference in the realised mean household generation time between the Delta (3.2 days, 95% CrI 2.4-4.2 days) and Alpha (4.5 days, 95% CrI 3.7-5.4 days) variants. This is because higher transmissibility led to faster susceptible depletion in households, in addition to the reduced intrinsic generation time. Conclusions: The Delta variant transmits more quickly than previously circulating variants. This has implications for interventions such as contact tracing, testing and isolation, which are less effective if the virus is transmitted quickly. Epidemiological models of interventions should be updated to include the shorter generation time of the Delta variant.

In mid-2021, the Delta SARS-CoV-2 variant became the dominant variant in the UK 1 and 36 globally. 2 This variant presents a higher risk of severe disease compared to previously 37 circulating variants, 2,3 although vaccination still provides significant protection. 4 38 Concerningly, the Delta variant led to an increase in the growth rate of COVID-19 cases in 39 the UK, swiftly outcompeting other SARS-CoV-2 variants. 1, 5 This was attributed to the 40 increased transmissibility of the Delta variant, with an epidemiological study 5 suggesting that 41 it is 43-68% more transmissible than the Alpha variant that was identified in the UK in late which is measured by the exponential growth rate and can be inferred from disease incidence 46 data for that variant. 8 However, understanding transmission of a novel variant also requires its 47 strength to be estimated 8 -this is the variant's transmissibility, which is typically measured 48 by the time-dependent reproduction number, or "R number" (the number of people that each 49 infected person is expected to infect). The generation time (the time between infection events 50 in infector-infectee pairs) determines the relationship between the speed and strength of a 51 variant, [8] [9] [10] [11] and is a crucial input to many models used to estimate the reproduction number 52 from case notification data. 12, 13 In principle, an increase in the growth rate of COVID-19 53 cases as observed for the Delta variant can be attributed to either an increase in 54 transmissibility or a shortening of the generation time, or both of these factors. 8,14 55 Previous studies have estimated the generation time for SARS-CoV-2 (for example 15-18 ), with 56 most of these estimates using data collected during the very early months of the COVID-19 57 pandemic. A household study conducted by Public Health England (PHE) between February 58 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 23, 2021. ; https://doi.org/10.1101/2021.10.21.21265216 doi: medRxiv preprint and November 2020 in the UK indicated that the generation time became shorter in the 59 autumn of 2020 compared to earlier months. 18 Since the SARS-CoV-2 generation time is 60 changing, up-to-date generation time estimates are crucial to inform estimates of the 61 reproduction number and to understand the relative transmissibility of different variants.

However, the Alpha variant was responsible for infections in only two households in that 63 study, while the Delta variant had not yet emerged by the end of the study. 18 The effect of 64 different variants on the SARS-CoV-2 generation time has therefore not been properly 65 assessed.

Here, we report analyses of data from an ongoing household study conducted by the UK In addition to estimating the household generation time for the Alpha and Delta variants, we 81 estimated this quantity for individuals of different ages and vaccination statuses, and at 82 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 

• The number of vaccine doses received, the vaccine type, and the date(s) of vaccination. 99 We assumed individuals were infected during the household transmission cluster if they 100 returned a positive PCR test (including tests taken up to 28 days before the household was 101 recruited to the study) and/or developed symptoms. Otherwise, we assumed individuals 102 remained uninfected if they returned only negative tests. Household members who did not 103 participate in the study, or withdrew before either taking a test or developing symptoms, were 104 excluded from our analyses (but were included in the household size).

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 23, 2021. ; https://doi.org/10.1101/2021.10.21.21265216 doi: medRxiv preprint

Genomic sequencing was carried out in 60% of study households. Where sequencing data 106 were not available, we assumed that the Alpha variant was responsible for infections in 107 households that were recruited to the study before May 2021, and the Delta variant for 108 infections in households recruited after May 2021 (data from three households recruited 109 during May 2021 in which sequencing was not carried out were excluded from our 110 analyses). 1 Overall, the Alpha and Delta variants were determined to be responsible for (Table S1) 

In Figure 1A , we show posterior estimates of the mean intrinsic generation time (which 149 involves an assumption that infected individuals have contacts with susceptible hosts at a 150 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 23, 2021. ; https://doi.org/10.1101/2021.10.21.21265216 doi: medRxiv preprint constant rate throughout their infectious period, and therefore provides a generalised 151 generation time estimate that is independent of the household size) for the Alpha and Delta 152 variants. We estimated this quantity to be slightly shorter for the Delta variant (4·6 days, 95% 153 CrI 4·0-5·4 days) compared to the Alpha variant (5·5 days, 95% CrI 4·6-6·4 days), although 154 there was overlap between the credible intervals ( Figure 1A ). On the other hand, our results is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 23, 2021. ; https://doi.org/10.1101/2021.10.21.21265216 doi: medRxiv preprint households); D. The standard deviation of household generation times. Posterior means and 95% credible 165 intervals for these quantities are given in Table S2 and Table S3 in the Supplementary Material.

Estimates of the mean household generation time for the two variants ( Figure 1C ) are shorter 167 than the corresponding intrinsic generation time estimates (Figure 1A) , since the household Figure 1A . We also 174 estimated the standard deviation of generation times within households to be smaller for the 175 Delta variant than for the Alpha variant ( Figure 1D ). In Figure S2 Figure S2C ). 180 We note that the estimates in Figure 1B infector. 186 We also explored the effects of vaccination (Figure 2A 

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 23, 2021. The data collection for the study was undertaken by the UKHSA (an executive agency of the Reference EP/V053507/1). Thanks to the household members who took part in this study, the 365 nursing staff at the UKHSA who recruited and followed up the households, the laboratory 366 staff at the UKHSA who tested the swabs and the UKHSA administrative staff who arranged 367 for the delivery and collection of testing kits from households. The household surveillance CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 23, 2021. ; https://doi.org/10.1101/2021.10.21.21265216 doi: medRxiv preprint informed consent for sampling and follow up was obtained by the nurses from household 371 members who were free to decline to participate in the surveillance at any time. Consent for 372 children was obtained by a parent or legal guardian. Only anonymised data were provided to 373 non-UKHSA authors. The authors declare no conflicts of interest.

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 23, 2021. ; https://doi.org/10.1101/2021.10.21.21265216 doi: medRxiv preprint

SARS-CoV-2 variants of concern and variants under investigation in 281 England (Technical briefing 23)

World Health Organization. Weekly epidemiological update on COVID-19 (Edition 58)

Hospital admission and emergency care attendance risk for 284 SARS-CoV-2 delta (B.1.617.2) compared with alpha (B.1.1.7) variants of concern: a cohort study

Effectiveness of Covid-19 vaccines against the 287 B.1.617.2 (Delta) variant

Increased transmissibility and global spread of 289 SARS-CoV-2 variants of concern as at

SARS-CoV-2 variants of concern and variants under investigation in 291 England

Speed and strength of an epidemic intervention

Roles of generation-interval distributions in shaping relative 295 epidemic strength, speed, and control of new SARS-CoV-2 variants

Estimating the time-varying reproduction number of 306 SARS-CoV-2 using national and subnational case counts

Estimated transmissibility and impact of SARS-CoV-2 308 lineage B.1.1.7 in England

Quantifying SARS-CoV-2 transmission suggests epidemic 310 control with digital contact tracing

Estimating the generation interval for coronavirus disease 312 (COVID-19) based on symptom onset data

High infectiousness immediately before COVID-19 symptom 314 onset highlights the importance of continued contact tracing

Inference of SARS-CoV-2 generation times using UK 316 household data

Inferring generation-interval distributions from contact-318 tracing data

Strategies for containing an emerging 320 influenza pandemic in Southeast Asia

Incubation period of COVID-19: a rapid systematic review 322 and meta-analysis of observational research

Occurrence and transmission potential of 324 asymptomatic and presymptomatic SARS-CoV-2 infections: A living systematic review and meta-325 analysis

Impact of Delta on viral burden and vaccine 327 effectiveness against new SARS-CoV-2 infections in the UK

On the relationship between serial interval, infectiousness 330 profile and generation time

Transmission dynamics of an outbreak of the COVID-19 Delta 332 variant B.1.617.2 -Guangdong Province, China

Serial intervals in SARS-CoV-2 B.1.617.2 variant cases. 335 The

Transmission dynamics and epidemiological characteristics of Delta 337 variant infections in China

Changes in the serial interval and transmission 339 dynamics associated with the SARS-CoV-2 Delta variant in South Korea

Viral dynamics of SARS-CoV-2 variants in vaccinated and 342 unvaccinated individuals