The recent spike in U.S. COVID-19 infections has mercifully been accompanied by a declining death count. There were days in the spring when the country had half the number of cases but twice as many deaths. Now, at least, the U.S. is testing more widely.

And even though death is a lagging indicator, and the numbers are likely to catch up to some degree, there is reason to hope that the lag could now be longer and slower than it was in the spring. After all, much has been learned about how to treat COVID-19 patients.

At the same time, however, a lower death count is no justification for states to reopen their economies incautiously or to suggest, as the White House appears to be doing, that Americans should just get used to living with COVID-19.

Fatalities lag behind cases because COVID-19 is a rather slow-going illness. It takes time to develop an infection severe enough to require hospitalization. It can take longer still for an acute infection to result in death. Add in that deaths aren’t always reported in a timely manner, and you see why fatalities are slow to rise and fall. Note that at the moment, the U.S. is less than a month into a sustained increase in cases.

Rising death tolls in hot-spot states such as Arizona, Florida and Texas have been balanced so far by declining figures in initially hard-hit states that now have the virus better under control.

California is an unusual case. Its case numbers are rising, yet its death count is still flat. This may be evidence that fewer of those people most likely to die from COVID-19 are now contracting the virus. This possibility is reinforced by data from several other states showing that COVID cases are skewing younger. Such an age shift means that more people arriving at the hospital have a better chance of surviving.

Once there, patients of all ages are getting better treatment. Remdesivir is helping some, and the steroid dexamethasone is preventing deaths among the severely ill. Having patients lie on their stomachs, a practice known as proning, may also improve outcomes. COVID patients do better in hospitals that are less crowded and overwhelmed, which many have been, until recently.

It must be kept in mind, however, that a lower average death count — say, 500 a day — is still tragic. And if the U.S. continues to add 50,000-plus new cases a day, the number of deaths will rise. Rampant case growth will inevitably breach measures that lead to better outcomes.

If current trends continue, the virus will inevitably spread to more older Americans and other vulnerable people in households with those who get out into the community. Few states provide housing for infected people to wait out the virus.

And as hospitals in some hot spots fill up, the crowding will lead to more deaths — from COVID-19 as well as other illnesses for which people cannot get care.

Nor is death is the only concern raised by rising infection rates. COVID-19 may also cause significant harm to those who survive. Organ damage might be lasting, and extended time in a hospital or on a ventilator can weaken people for months.

The long-term consequences of infection, still poorly understood, are not included in the data that appear on state and national dashboards. But it’s obvious that with the numbers the U.S. has seen in the past month — more than a million new infections and more than 30,000 people hospitalized — many more people, and the health system, will feel the effects for years to come.

COVID patients probably have a better chance of survival now than they did in March. They’ll almost certainly have even better odds a few months hence. But that’s no reason to allow infections to keep soaring.

Max Nisen is a Bloomberg columnist covering biotech, pharma and health care.