It will probably end up being the case that no category of illustration in human history has saved more lives than those showing the need to “flatten the curve."

You’re almost certainly familiar with the term by now, weeks after a new strain of coronavirus escaped containment and began infecting thousands of Americans. The thrust is that slowing the rate at which the virus spreads will similarly slow the number of people needing to go to the hospital — and, therefore, give doctors more space to treat not only those with covid-19, the disease caused by the virus, but to treat everything else, as well.

The illustration generally looks something like this. An uncontrolled spread of the virus (red) could see a rapid surge in demand for hospital facilities. A slower spread (yellow) offers a better chance to care for all of the patients coming into the system. Push down the peak number of cases and stay below capacity.

It makes sense in the abstract. In reality, though, it depends on a number of factors, including:

  1. How many people end up getting sick (the area of the curves above).
  2. How quickly they get sick (the slope of the curves).
  3. The actual capacity of the hospitals.

Changes in any of those factors can dramatically affect a community’s ability to handle the coronavirus outbreak. A spike in new cases can overwhelm a health-care system. If you hold the rate of infections steady but a higher percentage of the population gets sick, you might still be in trouble. If you’re in a region with fewer available hospital beds, your hospitals may be overwhelmed more quickly than another place, even if your rate of infection is slower.

This week, researchers from the Harvard Global Health Institute released estimates looking at how the three factors above vary in hospital systems around the country. The data, first reported by the New York Times and ProPublica, show how individual communities might fare under rates of infection and overall spread. It includes estimates relative to the severity of illness — is intensive care required? — and considering available bed counts (recognizing that hospitals have patients being treated for other conditions) and total bed counts, should every bad in every hospital be converted to use solely for covid-19.

Inspired by the Times’s visuals, we took those data and made regionalized curves showing the percentage of capacity in any hospital system that would be required under different infection scenarios, including the number of people infected and the length of time over which those infections will occur.

What percentage of adults will be infected? 20%

How long will that take? 12 months

What region are you concerned about?

Your browser can't display this graph.

You can see how limiting the number of infections or expanding the timeline over which those infections occur can make the curve flatter. But you also probably noticed that in most scenarios, the number of cases quickly exceeds the system’s available beds. Those red curves that extend above 100 percent capacity indicate scenarios in which the curve wasn’t flattened enough.

You probably noticed that even when considering situations in which every single hospital or ICU bed is dedicated to handling covid-19, capacity was still exceeded. While it’s incumbent upon Americans to work to flatten the curve by limiting the spread of the virus, isolating and maintaining distance from other people, our ability to affect the conditions that will protect the health-care system is limited to the first two of the three factors we listed above. Another way to prevent hospital systems from being overwhelmed is to increase that capacity, to add beds as rapidly as possible. That’s not something you or I can do; it’s up to the government.

We need to work to flatten the curve. But almost everywhere, the government also needs to rapidly add capacity so that the curve doesn’t need to be quite as flat.