What Really Happened at the Start of the COVID-19 Pandemic?

At this two-year point in the pandemic, I’ve been revisiting what happened at the start to get a clearer idea of how things unfolded and whether what happened then can tell us anything useful about the virus. Here’s what we know.

In the last week of December 2019, a doctor in Wuhan noticed unusual pneumonia in six patients who all tested positive for a new coronavirus. We can surmise that to have hospitalised six unconnected people (not all linked to the Huanan market), the new virus must have been circulating in Wuhan for some weeks. Internal reporting of this cluster led to the first public message of a suspected pneumonia outbreak, with precautions advised, from the Wuhan Municipal Health Commission on December 31st. At this point there were just 27 identified cases in hospital, seven of them serious – which wouldn’t seem out of the ordinary for winter.

Despite this inauspicious start, Hubei province went on (according to official data) to have a deadly outbreak, though one that was notably slower burning and milder than later Covid outbreaks elsewhere. It totalled around 4,500 deaths out of a population of 57 million people, making it around a tenth as deadly as the first wave in the U.K., and peaked at 143 reported daily deaths on February 19th (suggesting the infection peak was late January). This doesn’t seem a particularly high death toll for a winter respiratory virus.

Notably, this initial deadly outbreak was very localised. Hubei was locked down on January 23rd, but prior to that the virus circulated freely for weeks, while millions of people left the province ahead of the lockdown. Despite this, no other province in China suffered a deadly outbreak (see below). While we might be inclined to question China’s official data, it fits with what happened in neighbouring countries – no other country in the region suffered a deadly outbreak of the new virus. South Korea’s outbreak peaked at six reported daily deaths on March 30th, Japan’s (which was the worst in the region) at 24 on May 1st.

The lack of deadly outbreak outside Hubei explains why most people were largely ignoring the rumours of another virus in the East at that time. It appeared to be fizzling out, having done little anywhere outside Wuhan, and the WHO refrained from declaring a pandemic. In all likelihood we would have just carried on ignoring it, as we did the earlier bird flu and swine flu scares, were it not for what happened next.

An outbreak in Lombardy in northern Italy led the local government there to order a lockdown – the first after Hubei – in 10 towns covering around 50,000 people on February 21st 2020, seemingly a panic measure in response to the first death in the region. However, that seems to have been an aberration and there was no wider push for restrictions at that point. That came two weeks later, on March 8th, when the Italian Government locked down the northern part of the country. Two days later, it locked the whole country down. The trigger for this appears to have been the extraordinary jump in the death toll. On March 7th Italy reported 36 deaths. On March 8th it reported 133, and on March 10th it reported 168, with full ICUs portending many more to come. The daily toll quickly climbed to a peak of 919 just over two weeks later on March 27th. The WHO declared a pandemic on March 11th, the day after Italy locked down.

It quickly became clear in the first two weeks of March that the outbreaks in Europe were orders of magnitude more deadly than anything seen in the East (see above). A week after Italy, Spain’s death toll climbed at a similarly alarming rate, and other countries soon followed. Italy reached one death per million on March 9th, Spain on March 16th, New York on March 20th, France and Netherlands on March 21st, Sweden on March 23rd and U.K. on March 25th, and all kept climbing to levels far beyond what had been seen in Hubei let alone elsewhere in South East Asia. In this panicked context, the go-to response became lockdown.

Working back from the pattern of deaths, it appears that these deadly outbreaks began in each country at some point in February, many within days of each other, and grew fast. These outbreaks are plausibly linked to one another, and notably follow a similar pattern to the later outbreaks associated with the emergence of variants such as Alpha, Delta and Omicron, where an initial outbreak in one country (in this case Italy) is soon followed by similar outbreaks in other countries as the new variant quickly spreads internationally. This suggests the initial deadly wave in Europe and the U.S. may have been caused by the emergence in northern Italy of a deadlier variant than had been seen in the East. Supporting this, while these outbreaks emerged within days of each other during February and followed a similar pattern of severity, they occurred months after the Hubei outbreak, which (projecting back from the death curve) began some time in November or October 2019, and was a much longer, slower burn. One explanation is the Hubei outbreak was caused by a less transmissible, less deadly variant. If so, it’s unclear why this ‘Hubei variant’ didn’t have the same effect elsewhere in South East Asia or the world – why was the deadly outbreak entirely localised to Hubei? Nonetheless, there is evidence that some form of SARS-CoV-2 was present globally throughout the winter of 2019-20 – though with that winter being in many countries the least deadly in history, this variant or variants was evidently either very mild or not very transmissible or both.

Since the emergence of the (hypothesised) nastier variant of SARS-CoV-2 in Italy in February 2020, similarly virulent variants have emerged such as Alpha and Delta, while Omicron may be a return to an earlier milder form. If so, was it just bad luck that, just as SARS-CoV-2 seemed to be fizzling out in February 2020, a nastier variant emerged in Italy that terrorised the world and shunted us into copying China’s overreaction in Hubei? Or was there something about SARS-CoV-2 that made this occurrence inevitable, or at least odds-on? Does the furin cleavage site that makes it so transmissible make it more likely it would hang around long enough to mutate into the more dangerous forms we have seen?

Mentioning the furin cleavage site brings up the suspicion that the virus originated in a lab, presumably as a leak. A furin cleavage site is not found in viruses of this sort in the wild but is a common change made to them in a lab. If it did leak from the Wuhan Institute of Virology, presumably the leaked virus is the one responsible for the localised deadly outbreak in Hubei (otherwise the connection between the outbreak and the leak would be coincidental). Was it then milder variants of the leaked virus that went round the world that winter, rather than the deadlier one in Hubei? If not, why did only Hubei have a deadly outbreak in December 2019-January 2020, even though the virus appears already to have been present around the world by December?

Seeing the emergence of the pandemic in late 2019 and early 2020 as, in part, a variant story, helps to fill in some gaps, such as the odd delay between the Hubei outbreak in November-January and those in Europe in February-March, as well as the massive difference in severity. The pieces don’t fit together entirely neatly though, and questions – especially about why the Hubei outbreak was so localised and uniquely deadly in South East Asia, why there is evidence of a milder form of SARS-CoV-2 around the world in winter 2019-20, and the role of a laboratory – remain.