That was yesterday’s reminder of how brutal and swift Mother Nature’s fury can be. This morning’s reminder comes in the form of a New England Journal of Medicine article by Nishant Kishore and colleagues at the Harvard T.H. Chan School of Public Health. The NEJM study offers one thoughtful, evidenced-based answer to a question that has lingered like subtext since last September: How many people in Puerto Rico died as the result of Hurricane Maria?

The answer, by their count: 4,645 people—or more than seventy times the official death toll of 64.

Subscribe to Brainstorm Health Daily, our newsletter about the most exciting health innovations.

That government figure—64—had been thought by many experts to be unrealistically low. That’s because, in order to be counted in the official toll from the hurricane, bodies had to be examined by a medical examiner. If a medical examiner couldn’t travel to the scene, the body had to be transported to an examiner’s office (typically, in one of the larger cities) in the immediate aftermath of the storm—something that was, in many cases, impossible given the washed-out roads and other hazards. So some of the storm-related dead simply weren’t counted.

But a still bigger issue was delay. Among the official tally were those who had been killed by debris swirling in the high winds, or by fallen trees or collapsed buildings. Left out, however, were potentially thousands of individuals who died from the longer-term effects of the storm. “According to the Centers for Disease Control and Prevention,” note the study authors, “deaths can be directly attributed to a tropical cyclone if they are caused by forces related to the event, such as flying debris, or if they are caused by unsafe or unhealthy conditions resulting in injury, illness, or loss of necessary medical services.”

As power went out across the island, and normal life and basic services were interrupted for months, it was inevitable that thousands of residents would perish.

The question was how many?

To find out, Kishore, a computational epidemiologist, and his research colleagues surveyed 3,299 households across the island—a sample that ranged across each of Puerto Rico’s 900 barrios, from the densest districts to the most remote of outposts—and they counted up the deaths in the first months after the storm. The team calculated an average rate of 14.3 deaths per every 1,000 residents from September 20 through December 31, 2017—a mortality rate that was an astounding 62% higher than in the same period in 2016. That works out, they said, to an estimated 4,645 excess deaths—one-third of which “were attributed to delayed or interrupted health care.”

I bring this up not merely to talk about the ferocity of Mother Nature. I bring this up to point out, once again, that delayed care, and the lack of accessibility to care, and the inability to afford care, leads to death. Plain and simple. Faithful readers of Brainstorm Health Daily will know that I wrote about this in January (“When Delay is Deadly”), after seeing a frightening study by a team at JPMorgan Chase. Researchers there examined the spending habits of 2.3 million families across the U.S., and found overwhelming evidence to show that Americans, across nearly every economic stratum, routinely defer or forego needed medical care when cash flow is short.

Yes, it’s natural to measure nature’s horrific power through the immediacy of a death toll—the way we do airplane crashes and the like. But worth remembering is that the more brutal disasters, often, are the ones that linger after the sirens have stopped and the ambulances are gone. Those, presumably, are the disasters we mere mortals have the power to prevent.