No one knows -- or may ever know -- exactly how many people died in the aftermath of Hurricane Maria in Puerto Rico. But an increasingly tall stack of evidence indicates the official government estimate -- 64 deaths -- misses the mark. By a long shot.
A research letter published Thursday in the medical journal JAMA estimated that between 1,006 and 1,272 people died in relation to the storm.
Accidents, disasters and safety
Continents and regions
Deaths and fatalities
That's about 15 to 20 times the Puerto Rican government's figure.
"They clearly were not counting all of (the deaths)," Jeffrey T. Howard, assistant professor of health at the University of Texas at San Antonio, and one of the study's authors, said of the Puerto Rican government. "They can't be. There were just too many deaths."
To get their own estimate, Howard and Alexis Santos-Lozada, from Penn State University, compared the number of deaths in the months following the hurricane to those in previous years. This measure of "excess death" is considered by academics to be a sound way of estimating a death toll in aggregate; meaning it does not account for individual cases.
The report from Santos and Howard follows a number of others like it. In May, a team that included researchers from Harvard University published a study in the New England Journal of Medicine estimating 793 to 8,498 people died in Maria's wake, a range that some academics have criticized as overly broad. The study's midpoint estimate -- 4,645 deaths -- became a rallying cry for activists upset by what they see as a lack of accountability for the scale of the catastrophe by officials in Puerto Rico and the United States.
The Harvard estimate was based on surveys of 3,299 households in Puerto Rico, in which residents were asked about deaths in their homes after Hurricane Maria. The differences in methodology -- interviews vs. a statistical accounting of "excess deaths" -- helps to explain the disparity in the two mortality estimates, Santos and Howard write in their report.
Puerto Rico officials say they do not plan to update the death toll until additional research is completed. In February, Gov. Ricardo Rosselló announced Puerto Rico had hired George Washington University to review the death toll from the September 20 storm, which left more than a million people without power for months and led to widespread criticism of the federal response to the chaos. That still-unreleased report was said to be coming in May. The university now says the initial report will be out this summer.
"Once we commissioned GW to do this study we stopped counting," said Pedro Cerame, spokesman for the Puerto Rican governor's office in Washington, D.C. "The number hasn't changed not because we believe there were only 64 deaths but because we're waiting for this study."
To calculate the estimate published in the newly released study, Santos and Howard looked at deaths from September to December, 2017. They compared those deaths to those that occurred each month from January 2010 through December 2017. The researchers estimate 459 excess deaths occurred in September, the month of the storm; followed by 564 in October and 116 in November. In December, mortality "had returned to a level within historical variation," they wrote.
Their midpoint estimate is that 1,139 "excess deaths" occurred in apparent relation to the storm.
CNN and other news organizations have been raising questions about the official Hurricane Maria death toll for months. In November, CNN reporters surveyed 112 funeral homes across the island, about half the total. Reporters found that funeral home directors identified 499 deaths they considered to be hurricane-related. In December, the New York Times estimated 1,052 "excess deaths" occurred after Maria. Others produced similar estimates.
This year, CNN and the Centro de Periodismo Investigativo (CPI) in Puerto Rico sued the island's Demographic Registry to make public a database with information about everyone who died in the months after the storm. The study released Thursday was possible because of that lawsuit and the data it made public, Howard said.
Using the same database, CNN reported on deaths labeled in government records as hurricane-related that were not counted by officials; and, in partnership with CPI, reported on an apparent leptospirosis "outbreak" that was not identified that way by authorities.
The network also created an online database where the public can search for names of all the people who died in the months after the storm -- and tell reporters about deaths that may have been related to Maria.
Most of the deaths in question are indirectly related to the hurricane, Howard said, meaning they occurred because of the chaotic circumstances that followed the storm -- no power or water and limited medical services -- rather than blunt trauma on the day the hurricane crossed the Caribbean.
Experts say understanding how many people died and why can be used to prevent similar deaths in the future. Family members of the deceased also are eligible for certain federal aid, such as having some of their funeral expenses paid, if their relatives' deaths are counted as hurricane-related.
"Having an accurate accounting of the death toll ... is really important to understanding how to best prepare for the future and how to respond appropriately," Howard said by phone. "When you have a massive undercounting of casualties, that can derail response times, (and) it can lead officials to underestimate what the true need is."