Uncovering Construction Site–Specific Transmission Patterns of COVID-19: A Spatiotemporal Connectivity Analysis in Hong Kong

Ziyue Yuan, Shi Zhao, Clara Cheung, Shu Chien Hsu

Research output: Contribution to journalArticlepeer-review

Abstract

To adapt to the prolonged pandemic, the construction industry, which has a high vulnerability to coronavirus disease 2019(COVID-19) infection, has sought more sector-specific and individual-level nonpharmaceutical interventions (NPIs). Understanding infec-tion transmission patterns can determine what, when, and how NPIs should be implemented. This study examined infection transmissionproceeding from construction sites using spatiotemporal analysis with COVID-19 case cluster data from construction sites in Hong Kong. Thestudy revealed that COVID-19 transmission diffuses from the workplace to residential neighborhoods where infected construction workerslive but not to the surroundings of infected construction sites. The average number of offspring cases infected by each seed case in the first tofifth transmission generations were 7.8, 26.1, 10.6, 3.6, and 1.3, respectively. Around 18% of cases were responsible for 79.6% of all COVID-19 transmission, driven mainly by workplace and household settings. The study found that closing a workplace within two working days aftera primary case is identified can help reduce the attack rate by 5.33%. Encouraging household members of infected construction workers tofollow quarantines can reduce offspring cases by 15.84% on average. A priori identification of superspreaders can help remove half ofCOVID-19 cases.

Original languageEnglish
JournalJournal of Management in Engineering
Volume39
Issue number1
Early online date20 Oct 2022
DOIs
Publication statusPublished - 1 Jan 2023

Keywords

  • Transmission pattern
  • Covid-19
  • Construction
  • Hong Kong

Fingerprint

Dive into the research topics of 'Uncovering Construction Site–Specific Transmission Patterns of COVID-19: A Spatiotemporal Connectivity Analysis in Hong Kong'. Together they form a unique fingerprint.

Cite this