How Singapore Found and Contained Its First 100 COVID-19 Cases, January–February 2020

COVID-19 was first reported in Wuhan, China, in December 2019, and by March 5, 2020, more than 95,000 cases had been confirmed worldwide. Singapore, a small island city-state in Southeast Asia of about 5.7 million people, had identified 117 of them by that date. Its Ministry of Health analyzed the first 100 to see how well its surveillance and containment measures were working; the report was posted as an MMWR Early Release on March 13, 2020.

A layered net

On January 2, 2020, days after China first reported the disease, the ministry drew up a local case definition and told every medical practitioner to watch for suspected cases. The definition was revised five times as the global picture changed. Doctors were required to report every suspected and confirmed case to a central notification system. In hospital, suspected patients had a chest X-ray and RT-PCR tests on at least two nose-and-throat swabs taken 24 hours apart.

Several methods then worked side by side:

• Contact tracing. Close contacts — within 2 meters (6.6 feet) of a patient, generally for 30 minutes or more — who had no symptoms were placed in compulsory quarantine for 14 days; lower-risk contacts were actively monitored. Officials telephoned close contacts three times a day and others once a day to ask about fever or respiratory symptoms, and anyone who became ill was taken straight to hospital.
• Enhanced surveillance, from late January: testing every hospitalized patient with pneumonia (later also pneumonia patients in primary care), intensive care patients with a possible infectious cause, patients with influenza-like illness at sentinel clinics, and deaths from possible infectious causes.
• Clinicians’ discretion: doctors could order a test on clinical suspicion — a prolonged unexplained respiratory illness, say, or a link to a known cluster — even if the case definition was not met.

Alongside them came border controls, community education and precautions.

The first 100 patients

Their average age was 42.5 (median 41); 72% were aged 30–59 and 60% were men. Twenty-four cases were imported and the rest were locally transmitted. Fifteen patients had been in intensive care at some point; none had died at the time of the report.

The methods caught different kinds of patients. Among cases linked to another patient or to travel from China, contact tracing found the most (62.7%). Among cases with no known source, enhanced surveillance found the most (58.8%) — patients who, if missed, could have started chains of transmission that are hard to contain. None of the methods alone would have found everyone. Of the 53 patients found by contact tracing, 13 had been reached on or before the day their symptoms began.

Cases by date of symptom onset (top) and date of report (bottom). Image from CDC’s page.

The earliest onset was January 14. Cases peaked on January 30, with nine patients, then fell to between two and five a day from February 11 onward. Imported cases made up most of the early ones before local cases took over.

Faster isolation

The key measure was the time from symptom onset to isolation in hospital or quarantine — roughly how long an infectious person spent in the community. Overall it averaged 5.6 days (median 5). Measured as a 7-day moving average, it fell significantly over the period: from 9.0 to 0.9 days for imported cases, and from 18.0 to 3.1 days for local ones.

Days from symptom onset to isolation, 7-day moving average, local and imported cases. Image from CDC’s page.

Other measures

Singapore screened temperatures at its borders, first for flights from Wuhan and then for everyone arriving by air, sea or land. Short-term visitors who had been in certain places in the past 14 days — first mainland China, later South Korea, northern Italy and Iran — were refused entry, and returning residents went into 14-day home quarantine. Public messages stressed hygiene, seeing a doctor early and isolating when ill. Schools stayed open, since there was no widespread community transmission and few cases in children, but classes and schools mixed less.

What it showed

The authors conclude that catching cases early, while numbers were small and each could be contained individually, likely slowed the outbreak. A separate study had rated Singapore’s surveillance capacity the highest of any country, estimating that if every country detected cases as well, 2.8 times as many imported cases would have been found worldwide. Countries unlike a small island city-state may need to adapt and add to its approach.

Limitations

• The moving average can shift for recent dates as more patients are found, and one measure alone may not capture how well containment works — though the downward trend was significant through early February.
• Detection focused on people with symptoms; how many infections were asymptomatic, and how much they spread, is unknown.
• The sample was small, with no cases in settings such as nursing homes or health care facilities.

Sources

Based on Ng Y, Li Z, Chua YX, et al., "Evaluation of the Effectiveness of Surveillance and Containment Measures for the First 100 Patients with COVID-19 in Singapore — January 2–February 29, 2020," Morbidity and Mortality Weekly Report, Centers for Disease Control and Prevention; rewritten in hubnx’s own words. The imported copy was cut off in one place; the missing results were taken from the same report’s full text in PubMed Central (PMC7739977).