Why Daily Virus Numbers Mean Little in Post-Lockdown Stakes

(Bloomberg) -- Wondering whether it’s safe to emerge from lockdown? Watching the daily coronavirus case trackers won’t help much. A better indication might come from the length of the lines outside the doctor’s office.

To strike the right balance between restarting the economy and guarding against new waves of the pandemic, countries must focus on more than daily infection numbers and death tolls, according to scientists. That’s because these are lagging indicators, showing the virus’s trajectory days or even weeks ago. The actual danger level in real time is better seen in more direct indicators.

In France, for instance, the country is monitoring increases in the number of calls to doctors’ hotlines and visits to general practitioners, alongside data like admissions to intensive care. In Germany, there’s more focus on the virus’s basic reproduction number, or R0, thought to be a leading indicator of sorts. In the U.K., the government has commissioned surveys of people selected at random to gauge more precisely how entrenched the virus might be.

Countries “need to build up a picture from a range of sources, and some of that is the death data, some of that is the infection data, and then some of that perhaps is survey data,” said Graham Cooke, a professor of infectious diseases at Imperial College London, which is working on a government survey of 100,000 people this week. “Looking at the number of infections diagnosed as such is quite difficult to interpret on its own.”

Reported cases are a lagging indicator of the impact of easing restrictions because it can take up to two weeks for symptoms to show. Then, when a patient seeks a test, it can take a few days or more for results to come back, if diagnostics are even available.

Confirmed Cases

As the U.S. and a number of European countries relax lockdown rules in an attempt to restart their economies, the big question is whether those steps will lead to an unmanageable second wave of Covid-19. The roughly 4.4 million confirmed cases globally -- almost certainly an undercount of actual infections -- suggest only a tiny fraction of the world’s population have so far contracted the disease.

The slow rollout of reliable antibody tests, which show who has already had the virus and possibly developed immunity, further clouds the picture. With economies at risk, governments can’t afford to wait the many months, or even years, it might take to get a vaccine before allowing more normal life to resume.

Because testing is inconsistent and the new coronavirus causes mild symptoms in many people, data for new infections can be misleading, missing places where the virus can keep spreading undetected for weeks. Relying on deaths from Covid-19 is even more problematic since it’s a lagging indicator: by the time a patient dies, his contact with an infected person goes back several weeks.

When it comes to studying mortality, a better option may be to look at excess deaths -- the total number of deaths from all causes and whether it exceeds those reported in prior years.

The Independent SAGE group, a panel set up as an alternative to the U.K. government’s official Covid-19 advisory group, said this week that the country should zero in on these weekly numbers. They may give a more a meaningful view of the pandemic’s impact by offering a reliable trend, while coronavirus death statistics can swing from changes in the underlying counting strategies, like the addition of nursing-home deaths in countries that previously didn’t include them.

Moving Averages

Scientists also suggest focusing on multiday moving averages to avoid overreacting to tiny changes in daily tallies. In Germany, for instance, the so-called reproduction rate started creeping up over one again in recent days, just as the country eased its lockdown -- meaning that every infected person was statistically passing it to more than one other person. But scientists advising the German government said individual days couldn’t be relied on because of the lag in data, and the rate has since fallen.

Germany reported a seven-day R0 -- or R-naught -- estimate of 0.88 Thursday, which means 100 people would only infect 88 people currently. Countries are trying to keep this transmission number below 1 so that the disease eventually peters out.

“We just need to look at the graph of daily figures to see the problems” with only relying on that data, said Martin McKee, professor of European public health at the London School of Hygiene & Tropical Medicine. Moving averages can help “smooth out the fluctuations. This is the only way to make sense of what we are seeing.”

The World Health Organization said Wednesday there are six indications of whether countries have got measures right when easing quarantines. They include the ability to detect, isolate and care for cases, as well as protective measures in workplaces. But even if all the criteria on its list are met, there are no guarantees of success.

It’s “no time for celebration, but time for preparation,” WHO Regional Director for Europe Hans Kluge said Thursday. “We have seen that the strongest health systems can be overwhelmed in a couple of weeks.”

©2020 Bloomberg L.P.