Propagation of a hospital-associated cluster of COVID-19 in Malaysia

Abstract

Background: Hospitals are vulnerable to COVID-19 outbreaks. Intrahospital transmission of the disease is a threat to the healthcare systems as it increases morbidity and mortality among patients. It is imperative to deepen our understanding of transmission events in hospital-associated cases of COVID-19 for timely implementation of infection prevention and control measures in the hospital in avoiding future outbreaks. We examined the use of epidemiological case investigation combined with whole genome sequencing of cases to investigate and manage a hospital-associated cluster of COVID-19 cases.

Methods: An epidemiological investigation was conducted in a University Hospital in Malaysia from 23 March to 22 April 2020. Contact tracing, risk assessment, testing, symptom surveillance, and outbreak management were conducted following the diagnosis of a healthcare worker with SARS-CoV-2 by real-time PCR. These findings were complemented by whole genome sequencing analysis of a subset of positive cases.

Results: The index case was symptomatic but did not fulfill the initial epidemiological criteria for routine screening. Contact tracing suggested epidemiological linkages of 38 cases with COVID-19. Phylogenetic analysis excluded four of these cases. This cluster included 34 cases comprising ten healthcare worker-cases, nine patient-cases, and 15 community-cases. The epidemic curve demonstrated initial intrahospital transmission that propagated into the community. The estimated median incubation period was 4.7 days (95% CI: 3.5-6.4), and the serial interval was 5.3 days (95% CI: 4.3-6.5).

Conclusion: The study demonstrated the contribution of integrating epidemiological investigation and whole genome sequencing in understanding disease transmission in the hospital setting. Contact tracing, risk assessment, testing, and symptom surveillance remain imperative in resource-limited settings to identify and isolate cases, thereby controlling COVID-19 outbreaks. The use of whole genome sequencing complements field investigation findings in clarifying transmission networks. The safety of a hospital population during this COVID-19 pandemic may be secured with a multidisciplinary approach, good infection control measures, effective preparedness and response plan, and individual-level compliance among the hospital population.

Keywords: Epidemiological investigation; Hospital; Infection control; Multidisciplinary; SARS-CoV-2; Whole genome sequencing.

Fig. 1

Epidemic curve of 34 COVID-19 cases based on symptom onset. Day 1 represents the day of symptom onset for the index case (17 March 2020). The dates of diagnosis based on real time-PCR are presented for the six cases who remained asymptomatic throughout the infection. Cases are stratified by location and category of case

Fig. 2

Timeline of events based on first exposure to a COVID-19 case, incubation period, diagnosis, and admission to the hospital from 15 March to 9 April 2020. Data of 34 cases are presented here. Top panel: Each row represents a case, and the rows are connected by lines to demonstrate the likely transmission of disease between case and contact. Each individual (case) is labeled based on transmission location (AH healthcare worker; AP- hospital patient; AF and AG- household members; AN- nursing home older adults; AE- nursing home employees) and number, age (years), and sex (F = female; M = male). AF1 was a one-month-old infant. AP3, AP7, AP8, and AP9 were patient-cases who were tested positive for SARS-CoV-2. Deaths occurred among AP1, AP3, AP4, AP7, AN4. The bottom left panel demonstrates multiple encounters between the cases and contacts

Fig. 3

Phylogenetic tree and nucleotide and amino acid changes of 75 Malaysian SARS-CoV-2 sequences from the B.6 lineage. Sequences available in GISAID up to 20 September 2020 are included and compared to the reference strain Wuhan-Hu-1 (GenBank accession number MN908947). Sequences are named as “GISAID reference number|date of sample [case code].” Amino acids changes are shown in brackets. Sequences from 17 cases associated with the outbreak are shown in blue, with unique substitutions highlighted in orange. Sequences from 4 cases that were investigated but not considered part of this outbreak are indicated by their case numbers (AX1 to AX4). Only non-synonymous substitutions present in > 1 sequence are shown