Shelters need wider testing for symptoms, study says

Findings show initial screening didn’t cover enough symptoms, such as headaches, loss of taste

In the throes of COVID-19’s first wave in Toronto in April, Dr. Meb Rashid remembers a single phone call delivering the results of a 60-person testing sweep at a refugee shelter.

Before the tests were done, six residents were flagged by a symptom screen, so some positive test results wouldn’t have been a shock. But when Rashid heard that 25 of the 60 were infected — a whopping 41.6 per cent positivity rate — he was floored.

“It was time to pause,” he said, “and then to scramble.”

Once the outbreak was over, Rashid and a group of medical workers examined what happened. In their new study — shared exclusively with the Star — they revealed that a dozen residents who weren’t caught by the initial symptom screen, but tested positive, had at least some signs of illness that were caught by an assessment the next day.

The findings, the research team said, highlight the need to screen for a diverse slate of symptoms. The study also calls for wider access to testing for shelter residents.

“Where there is an outbreak identified — and that is defined as having one case or more, then we definitely need to make available timely access to testing for all individuals,” lead author Dr. Vanessa Redditt said.

The new study comes on the heels of a pre-print article from several Unity Health researchers, which also supports mass testing of all shelter residents if a single case is found.

Toronto Public Heath (TPH) currently makes shelter testing decisions based on things like layout, an infected person’s close contacts, and how effectively a site has implemented measures such as mask-wearing. In some cases, it recommends everyone get tested, said Vinita Dubey, associate medical officer with TPH. But in others, it might recommend testing only one floor.

As of Monday, there were10 COVID-19 cases reported across Toronto’s shelter system.

In the case probed by the new study, residents were only screened for fevers, coughs and shortness of breath before testing, with six residents flagged. Five of six tested positive, plus 20 others — meaning 80 per cent of their cases may have at first appeared asymptomatic. A more thorough assessment the next day revealed that a dozen more of the infected residents had signs of illness. Within two weeks, all but three of the infected residents contacted for followup had reported at least one symptom.

While Redditt warned their data could be affected by recall bias — someone only noticing a mild symptom after being told they were sick — she described the change that a broader symptom screen could make in determining who had signs of COVID-19 as “striking.”

The most common symptom, both on the first day post-test and in the ensuing two weeks, was a headache — affecting 58.3 per cent of cases. The next most common symptom was loss of taste, with both exceeding the number who reported fevers, coughs and shortness of breath.

The city’s most recent screening tool for shelters asks about a variety of symptoms, including headaches. To Rashid, their directives are now fairly clear: “Where in doubt, particularly people living in congregate living centres … the threshold needs to be very low for testing and isolation.”