It’s impossible to know what the next few months hold, as it was impossible to know several months ago where we’d be now. In April, it seemed like the pandemic was being stamped out, with new cases plummeting as rapidly as Americans were being vaccinated against the coronavirus. Now it seems that we’re at the front end of a new, dangerous phase powered by the more-contagious delta variant.

       Support our journalism. Subscribe today. arrow-right

       But maybe we’re as wrong now as we were then. After all, the long-term expectation for the virus at this point is that it will become both endemic and less deadly, with both of those outcomes following from broad vaccination against it. We already know the vaccine helps lower rates of infection and, among the infected, makes the disease it causes, covid-19, less dangerous.

       Data from the Centers for Disease Control and Prevention indicate that some 5,600 fully vaccinated people have been hospitalized while coronavirus-positive — though a quarter of them were there for other reasons and learned they had asymptomatic infections once tested in the hospital. When compared with the more than 2 million new hospitalizations this year, that’s a tiny fraction — just as the number of deaths among the vaccinated is a tiny percentage of all covid-related deaths.

       This is the goal: Individuals will get vaccinated so that everyone is less likely to get infected, and in the unlikely event that a vaccinated person is infected, the prognosis is significantly better. So, with 69.4 percent of American adults having received at least one dose of a vaccine, we might expect a greater percentage of new infections to avoid hospitalization or death.

       Since the current surge in cases began at the beginning of this month, the number of hospitalizations and the number of patients sent to the intensive care unit has tracked with that surge. The number of deaths, though, hasn’t.

       It’s tricky to compare these values, because there is necessarily a lag between them. A person is infected. That infection is detected. The person becomes sick and goes to the hospital. Then the illness worsens and the patient dies. (At each step, the percentage affected declines; very few of the infected die of covid-19.) The periods between those stages vary. In some cases, a person’s illness is detected at admission to the ICU, for example. While there is a pattern to the emergence of symptoms after an infection, it can vary by days. So, figuring out when we’d expect deaths to increase after cases are detected is challenging.

       One thing we can look at is the rate of change. Compared with two weeks earlier, the number of new cases shot up this month, with the rate of increase beginning to decline only this week. Hospitalizations are still rising. But, again, the number of deaths each day (measured as a seven-day average, as with the new case data) hasn’t changed much.

       (Notice that the rate of change for ICU visits — the lighter orange line — tracks closely with hospitalizations. We’ll ignore it below for that reason.)

       Advertisement

       Story continues below advertisement

       Since last August, before the massive third surge in cases that carried through the winter, the rates of increase between infections, hospitalizations and deaths have moved in sync when offset slightly. Hospitalizations generally track with new cases of infection on a one-week delay; deaths generally track when offset by about two weeks.

       When overlaid, we see that the pattern for deaths is, so far, not tracking with cases and hospitalizations.

       There are caveats. The death changes are noisy, a function of the total being lower than the number of cases of infection. It’s possible a surge in deaths will start soon and keep pace with the growth in the other metrics this month. Given that the vast majority of new cases are among the unvaccinated, according to government reports, we should see a sharp increase in deaths.

       Advertisement

       Story continues below advertisement

       Using a different measure, the patterns are trickier to parse. From August through November of last year, there were an average of 86 hospitalizations each day for every 100 new cases a week earlier. From December through March, there were an average of 64. Since April 1, there have been 99.

       In that first period, there were an average of 1.74 deaths each day for every 100 cases of infection reported 24 days prior. From December to March, the period in which vaccines were first being rolled out, there were 1.45 deaths. Since April 1, the average has been 1.41 deaths — but since May the ratio has steadily climbed higher.

       When we’re dealing with ratios like these, small shifts in small numbers can lead to larger perceived shifts. When we talk about the surge in cases, for example, the percentage-point increases are massive in part because infection numbers were so low — but also, in part, because the number of new infections increased so rapidly.

       Advertisement

       Story continues below advertisement

       But this is perhaps the most important question from both a medical and a political standpoint: Will the rate of deaths increase as rapidly as the rate of new infections? If it doesn’t, if that figure stays flat, it means a very different pandemic. Still one with a lot of deaths, certainly, but with a virus that’s less deadly to a better-protected public. That means potentially less urgency in keeping infection totals low and, therefore, less-restrictive measures aimed at preventing the virus’s spread.

       Again, though, it’s hard to predict. So once again, we have to wait and see, watching the numbers and crossing our fingers.