Case clustering, contact stratification, and transmission heterogeneity of SARS-CoV-2 Omicron BA.5 variants in Urumqi, China: An observational study

Abstract

Background: From August to September 2022, Urumqi, the capital of the Xinjiang Uygur Autonomous Region in China, faced its largest COVID-19 outbreak caused by the emergence of the SARS-CoV-2 Omicron BA.5.2 variants. Although the superspreading of COVID-19 played an important role in triggering large-scale outbreaks, little was known about the superspreading potential and heterogeneity in the transmission of Omicron BA.5 variants.

Methods: In this retrospective observational, contact tracing study, we identified 1139 laboratory-confirmed COVID-19 cases of Omicron BA.5.2 variants, and 51 323 test-negative close contacts in Urumqi from 7 August to 7 September 2022. By using detailed contact tracing information and exposure history of linked case-contact pairs, we described stratification in contact and heterogeneity in transmission across different demographic strata, vaccine statuses, and contact settings. We adopted beta-binomial models to characterise the secondary attack rate (SAR) distribution among close contacts and modelled COVID-19 transmission as a branching process with heterogeneity in transmission governed by negative binomial models.

Results: After the city lockdown, the mean case cluster size decreased from 2.0 (before lockdown) to 1.6, with decreased proportions of contacts in workplace and community settings compared with household settings. We estimated that 14% of the most infectious index cases generated 80% transmission, whereas transmission in the community setting presented the highest heterogeneity, with 5% index cases seeding 80% transmission. Compared with zero, one, and two doses of inactivated vaccine (Sinopharm), index cases with three doses of vaccine had a lower risk of generating secondary cases in terms of the reproduction number. Contacts of female cases, cases with ages 0-17 years, and household settings had relatively higher SAR.

Conclusions: In the context of intensive control measures, active case detection, and relatively high vaccine coverage, but with an infection-naive population, our findings suggested high heterogeneity in the contact and transmission risks of Omicron BA.5 variants across different demographic strata, vaccine statuses, and contact settings. Given the rapid evolution of SARS-CoV-2, investigating the distribution of transmission not only helped promote public awareness and preparedness among high-risk groups, but also highlighted the importance of continuously monitoring the transmission characteristics of genetic variants of SARS-CoV-2.

Figure 1

The epidemic curve, the trend of number of close contacts, and the case cluster size from 7 August to 7 September 2022 in Urumqi, China. Panel A. Epidemic curve of the daily number of laboratory-confirmed COVID-19 cases by the test-positive date. Panel B. Daily number of close contacts by the last date of contact with their associated index cases. Panel C. Number of close contacts of each index case by the last date of contact, stratified by contact settings. Panel D. The case cluster size by the test-positive date of the identified seed case of the cluster, stratified by epidemic periods before vs after city lockdown (the vertical bold dashed line).

Figure 2

Visualization of the transmission network of all COVID-19 case clusters in Urumqi between August 7 and September 10, 2022. Epidemiologically linked cases are connected by grey edges, and the arrows denote the direction of transmission, stratified by the epidemic period. Panel A. Before city lockdown. Panel B. After city lockdown.

Figure 3

Complementary cumulative distribution function (i.e. tail distribution) of the fitted beta distributions of secondary attack rate (SAR). Each curve represents the proportion of seed cases (i.e. infectors) that had SAR larger than the given number on the horizontal axis. Panel A. Stratified by age groups. Panel B. Stratified by vaccine doses. Panel C. Stratified by contact settings. Panel D. Stratified by epidemic period.

Figure 4

The estimated effective reproduction number (R) and dispersion parameter (k). The solid circles denote mean estimates, and the horizontal and vertical bars denote 95% CrIs of R and k, respectively. The gradient colour of middle dots denoted the proportion (%) of the most infectious cases that seeded 80% of transmissions. Panel A. Stratified by contact settings. Panel B. Stratified by epidemic period (right panel).