Epidemiological drivers of transmissibility and severity of SARS-CoV-2 in England

Abstract

As the SARS-CoV-2 pandemic progressed, distinct variants emerged and dominated in England. These variants, Wildtype, Alpha, Delta, and Omicron were characterized by variations in transmissibility and severity. We used a robust mathematical model and Bayesian inference framework to analyse epidemiological surveillance data from England. We quantified the impact of non-pharmaceutical interventions (NPIs), therapeutics, and vaccination on virus transmission and severity. Each successive variant had a higher intrinsic transmissibility. Omicron (BA.1) had the highest basic reproduction number at 8.3 (95% credible interval (CrI) 7.7-8.8). Varying levels of NPIs were crucial in controlling virus transmission until population immunity accumulated. Immune escape properties of Omicron decreased effective levels of immunity in the population by a third. Furthermore, in contrast to previous studies, we found Alpha had the highest basic infection fatality ratio (2.9%, 95% CrI 2.7-3.2), followed by Delta (2.2%, 95% CrI 2.0-2.4), Wildtype (1.2%, 95% CrI 1.1-1.2), and Omicron (0.7%, 95% CrI 0.6-0.8). Our findings highlight the importance of continued surveillance. Long-term strategies for monitoring and maintaining effective immunity against SARS-CoV-2 are critical to inform the role of NPIs to effectively manage future variants with potentially higher intrinsic transmissibility and severe outcomes.

Fig. 1. Population-level transmission of SARS-CoV-2 between March 2020 and February 2022 in England.

a Infection positivity amongst individuals 15 years and older in the community through the national PCR testing programme (Pillar 2). b Infection prevalence (PCR) in representative samples of households from the REACT-1 study. Panels a and b show data in green (lines and point, with binomial 95% confidence intervals in error bars) and model trajectory in red (average and 95% credible interval (95% CrI)). Grey shading indicates periods of non-pharmaceutical interventions (NPIs) of interest for the analysis; see in f. For a complete list of modelled change points in contact rates see Supplement Table S11 and Figs. S28 and S29. c Model-inferred average frequency of daily infections by variant and type of infection (either primary or re-infection following any prior infection, “reinf”). d Intrinsic basic reproduction number (R0) estimates by variant (mean and 95% CrI). e Model trajectory of vaccine status of the national population (all ages as denominator), as informed by official data of daily doses administered (see sources in Table S1); transition between vaccination classes was modelled stochastically (Supplement section 3.2) to capture smooth changes in population-level immunity over time. f Model trajectories of the instantaneous reproduction number in the absence of the effect of immunity (Rt) or accounting for immunity (effective Rt). Legends and grey areas specify date and duration of official NPIs in England over the study period. g Inferred effective levels of immunity in the population (all ages as denominator). Lines correspond to the effective immunity against specific variants, with colour scheme as in c and d, whereas areas indicate overall effective immunity by type of immunity (vacc: vaccine-induced, inf: from prior infection, or hybrid). Note that the coloured areas corresponding to the different types of immunity cover different periods of variant dominance and should be interpreted in the context of the circulating variants (see Supplement section 4.8). During periods of variant replacement (e.g. Alpha to Delta) the effective immunity transitions from the levels associated with the variant being replaced to the level of the variant that becomes dominant.

Fig. 2. Inferred severity of SARS-CoV-2 variants in England between March 2020 and February 2022.

a–c Inferred basic and effective infection hospitalisation ratio (IHR, a), hospital fatality ratio (HFR, b) and infection fatality ratio (IFR, c). Boxplots to the left show the basic severity measure and trajectories to the right show severity measure over time (mean and 95% credible interval, 95% CrI) for each variant, with black showing the weighted average across co-circulating variants at any time. Basic severity is measured for each variant assuming healthcare characteristics of the early epidemic, to allow a like-for-like comparison. Effective severity trajectories account for changing vaccine- and infection-induced protection against severe disease, as well as underlying healthcare variations (see Table S8 for model time-varying severity parameters). d–f Age-specific (selected age groups) effective IHR (d), HFR (e) and IFR (f), assessed at the date of key milestones of the national vaccination programme, “rollout” refers to the start of the vaccination programme on 8th December 2020.