What We Don’t Know About COVID-19 Can Hurt Us

Countries around the world have introduced stringent control measures to stop COVID-19 outbreaks growing, but now many find themselves facing the same situation again. From Melbourne to Miami, the relaxation of measures had led to increasing flare-ups, which in some places has already meant reclosing schools, businesses or travel routes. Within the U.S. and among different countries, places with wildly varying public-health policies have experienced wildly diverse outcomes. Most ominously, infections are rising rapidly in many places where they once were falling.

So how do countries avoid an indefinite, unsustainable, cycle of opening and closing society? What is needed to prevent a future of strict social distancing and closed borders? To escape this limbo, we need to know more about each step in the chain of infection: why some people are more susceptible or have more symptoms, how our interactions and surroundings influence risk, and how we can curb the impact of the resulting disease. Research around the globe has yielded some promising insights into these questions, but also some contradictory findings. Some studies suggest children are less susceptible to disease, while others suggest they are less likely to spread infection too. There is evidence that some aspects of immunity against the virus may wane quickly, while others persist. Based on certain datasets, few individuals are truly asymptomatic; according to other studies, a larger proportion may be.

Often when a new disease outbreak declines, we only later discover precisely why it took the shape it did. In 1854, John Snow famously took the handle off a water pump in the London neighborhood of Soho to tackle a cholera outbreak. What happened next is less well known: by the time the handle was removed, the outbreak was already in decline. From SARS to the 1918 influenza pandemic, history shows that many crucial insights only come long after the battle is over. Unfortunately, we don’t have that luxury for COVID-19. To effectively and sustainably control the disease, we will need a better understanding of the infection that causes it.

As the COVID-19 pandemic has grown, so has the apparent diversity of the disease’s impact on individuals, with susceptibility and severity varying enormously with age and medical history. New medical insights could help shrink this diversity for the better. An effective vaccine would reduce susceptibility, stopping the infection taking hold; an effective treatment would reduce the impact of infection in its final stages in hospital, preventing severe or fatal outcomes. But to know what is truly effective, we need large clinical trials that can roll out before cases decline, as studies have for hydroxychloroquine and dexamethasone in the U.K.

If we can’t change susceptibility or severity, then we need to look at what happens in between: how people are exposed to the virus. That means understanding the range of human surroundings that drive transmission. However, lockdowns were the equivalent of treating a patient with everything at once; something worked, but it’s still not entirely clear what. From schools and shops and bars and workplaces, each type of closure changed our interactions and the transmission opportunities they create—but we still don’t have the data to say for certain which of these changes mattered the most.

To successfully avoid ongoing disruption, we need to harness the variability of COVID-19 outbreaks—and the control measures countries have introduced—to make more sense of how the infection starts, spreads and ends. This variability was damaging and unpredictable in the spring, but going into the fall, this same diversity could help us target the weak links in the infection process.

As with any noisy data source, researchers detected the strongest signals about risk first: the meatpacking factories, the cruise ships, the nursing homes. Countries like Taiwan, Japan and South Korea have avoided untargeted lockdowns in part by refining this knowledge, collecting the data required to know where and how outbreaks are happening. There’s evidence that most people infected with COVID-19 don’t pass the infection on to anyone; even if risk can’t be avoided entirely, smarter structuring of interactions and lifestyles could limit superspreading events. Venues in several countries have introduced check-ins to help flag emerging clusters, while routine, rapid testing in workplaces could help get ahead of flare-ups before they happen.

Uncertainty about risk can translate into uncertainty in behavior, as we saw on the way into lockdown and are seeing on the way out. The virus seems to thrive indoors, particularly at crowded, loud, and lengthy gatherings. It’s becoming clear that a day at the park is very different to an hour in a bar. But the multiple dimensions of transmission are often ignored in favor of a focus on single measures. Is six feet enough distance? Should face masks be worn everywhere? Media reports show pictures of crowded beaches and parks, which are likely seeing little transmission, while superspreading events are happening in bars and workplaces. As measures are relaxed, this uncertainty means some are unwittingly heading once again into risky environments, while others, remembering stay-at-home messaging, are avoiding surroundings that are now known to be relatively safe.

Although local variation in control measures may appear to create uncertainty, it can actually help reduce it, by providing faster insights into what behavioral changes are effective. Take the introduction of face mask requirements in Germany, where different states rolled them out over a 10-day period. One concern with introducing mandated face mask usage was that it might lead people to become complacent in their behavior. By staggering the timing of when masks were introduced, Germany created a ‘natural experiment’ that could reveal what happened across multiple locations. The results suggested that when states introduced compulsory masks, it didn’t lead to a subsequent increase in visits to public spaces.

There won’t be a single, perfect dataset that will reveal how to sustainably control COVID-19. Whether we want to reduce susceptibility and severity or understand infection and immunity, we will have to learn from the diversity of outcomes and dynamics we are seeing globally. That means having surveillance systems tracking infections from multiple directions, research studies set up to monitor potential cases, and disparate data streams linked together. As the second epidemic wave hits several countries, it’s crucial to collect the information needed to prevent a third.