SINGAPORE - It has so far taken just over four days for contact tracing efforts to establish links between seemingly unrelated cases and existing clusters, with experts saying it suggests the work put in to contain the spread of community cases may be working.

They note that with ring-fencing strategies and phase two (heightened alert) measures, clusters have been kept small.

There are currently 47 active clusters, with almost half (46.8 per cent) of them having five cases or fewer. Three in 10 clusters (29.8 per cent) consist of three cases, the minimum for it to be defined as a cluster in Singapore.

The largest active cluster is the Changi Airport one with 108 cases.

Checks by The Straits Times showed that from the time a case is first reported as unlinked, it has taken an average of 3.2 days to establish a link to at least one other case. For cases linked to clusters, it has taken an average of 4.5 days to find the link.

However, some have taken a longer time.

The White Sands mall cluster's index case - the earliest known case in a cluster - was first reported as an unlinked case on May 13. It took 23 days before the case was linked to three others to form the cluster on June 5.

Singapore provides daily updates on the number of Covid-19 infections, but they are reported as at noon each day, which means any links found after that time are not reflected until the following day.

Professor Teo Yik Ying, dean of the National University of Singapore's (NUS) Saw Swee Hock School of Public Health, said: "By the cut-off time each day, if no links can be established for the new cases, they will be tentatively classified as unlinked cases, but many of them subsequently become linked to known cases and clusters.

"It is simply a matter of the amount of time and effort required for the contact tracing efforts to be concluded."

In tracing individual cases to known infections, some may initially be linked to smaller clusters before their connections to larger clusters are uncovered.

For instance, the 63 confirmed cases in the Jem and Westgate cluster include 14 who were initially reported as unlinked.

Some of them were later linked to small clusters involving their family members or household contacts. But it was later established that these small clusters were in fact related to the Jem and Westgate cluster.

Experts say it is important to identify connections to other cases, as unlinked cases could suggest hidden reservoirs of infections and trigger large-scale community outbreaks.

"This is why there are all these efforts to perform community surveillance, where entire blocks of residents or patrons to affected malls are invited down for testing," said Prof Teo.

He added that many of the new clusters emerging recently, including those in the north-east region of Singapore such as the Hougang cluster, may actually be linked to one another.

"We have just not discovered the infected individuals between these links," he said.

Prof Teo noted that surveillance efforts are even more important now compared with last year, since there is now a more infectious B16172, or Delta, variant.

"Just a handful of undetected cases can easily trigger a cascading effect to produce many more cases in the community and within a much shorter time span too," he added.

But Associate Professor Alex Cook, vice-dean of research at the NUS Saw Swee Hock School of Public Health, said there is a lower risk now of infections remaining hidden compared with last year.

"If we look at the number of cases in the intensive care unit, they are much lower than they were in the run-up to the circuit breaker last year, even though the number of community cases being found was similar. This points to improved detection of community cases now."

The current outbreak in Singapore started in late April.

Since then, there have been 658 new Covid-19 cases reported in the community. Of those, 147 cases, or 22.3 per cent, were initially reported as unlinked.

More than half - 81 cases - have since been linked to clusters. Another 15 cases were linked to one other case each.

Five cases had positive serology test results, indicating that they had old infections at the time they were tested and were no longer infectious.

Three others were later reclassified as imported cases by the Ministry of Health (MOH).

A further 43 cases, or 29.3 per cent of the 147 cases initially reported as unlinked, still have no known links to date.

The seven-day moving average for new daily community cases has climbed from 5.3 on April 28 to a peak of 28.4 on May 19, before falling to 15 on June 4.

The number of new unlinked cases has also come down, from a peak of 9.1 on May 19 to 2.1 on June 4.

MOH calculates the seven-day moving average as the average of the number of cases on a given day, the three days before and the three days after.

Prof Teo said the fall in unlinked cases could be attributed to heightened alert measures that have been in place since May 16.

The measures, which include making working from home the default and banning dining in, have greatly reduced movements and interactions, thereby minimising opportunities for people to get infected in the community, he said.

Another factor could be that large clusters at TTSH and Changi Airport have been brought under control, Prof Cook said.

On Sunday, the TTSH cluster was closed after no new cases were linked to it for 28 days. It had a total of 48 confirmed cases.

"If the (new) clusters are smaller, that could indicate that contact tracing is identifying the cases quicker before they can spread further, or the impact of phase two (heightened alert) measures on reducing contacts between people," Prof Cook said. "Either way, it's a good sign, but most important is the reduction in the number of unlinked cases."

The cases who remain unlinked after some time may have become infected through contact with contaminated surfaces or poorly ventilated environments rather than direct contact with an infected person.

"The Delta variant that is responsible for the community outbreak now is so transmissible, I suspect contaminated surfaces and poorly ventilated environments may be responsible for more transmissions than previously in 2020," Prof Teo said.

"This makes the process of contact tracing a lot more difficult, even with TraceTogether and SafeEntry. That is why there is a stepped-up regime of large-scale optional testing when specific locations have been frequented by a number of infected cases, instead of simply notifying the public of these locations."