Thread by @stukhan

Wastewater-based epidemiology (“sewage testing”) for SARS-CoV-2 is exciting and some of the best work in the world is coming from Australia. But it's important to communicate accurate conclusions from the science (1/8). abc.net.au/news/2020-12-1…

This story, like many others before it, claims breathlessly that this study “shows wastewater testing can detect coronavirus weeks before people show symptoms”. It really does not do that (2/8).

The key point is that this study reports the analysis of archived sewage samples, collected in Brisbane in Feb 2020, but analysed months later. Its true that there were no clinical cases reported in Feb, but there was practically no clinical testing occuring to do that (3/8).

If people were widely encouraged to get tested in February (and clinical samples were collected and archived), subsequent analysis of those samples would also likely reveal cases (4/8).

And we have no basis to claim that nobody in Brisbane had symptoms of COVID-19 in Feb, since the symptoms are often mild and could easily have been dismissed for other illnesses. It would be a differnt story now since even those with mild symptoms are encouraged to test (5/8).

So, there is no level playing field on which to compare the rapidness of sewage testing to that of detecting COVID-19 by symptoms and widespread available clinical testing. In other words, the conclusion is drawn from a highly biased retrospective assessment of the data (6/8).

Why does it matter? Because this pandemic is affecting all our lives and in serious ways. We need a good handle on what approaches are most effective for managing it. Misleading reporting risks focusing attention in the wrong places and that will have negative consequences (7/8).

Great science requires great and accurate communication. We’re kicking lots of goals in the science department, but occasionally, we’re all let down by overzealous science communication (8/8).

One further comment: there’s a perfectly important and exciting headline that should have come from this work: “archived sewage samples reveal COVID-19 was more prevalent in Australia earlier than was known at the time”. But that gets lost in the quest for something more sexy.