– The government of Puerto Rico has quietly acknowledged in a report posted online that in all likelihood more than 1,400 people died in the aftermath of Hurricane Maria — a figure that is more than 20 times the official death toll.

Hurricane Maria cut through the island on Sept. 20, knocking out power and initially killing about a dozen people. The government's official count eventually swelled to 64, as more people died from suicide, lack of access to health care and other factors. The number has not changed despite several academic assessments that official death certificates did not come close to tallying the storm's fatal toll.

But in a draft of a report to Congress requesting $139 billion in recovery funds, the Puerto Rican government acknowledges that 1,427 more people died in the last four months of 2017 compared with the same time frame in the previous year. The figures came from death registry statistics that were released in June, but which were never publicly acknowledged by officials on the island.

"Although the official death count from the Puerto Rico Department of Public Safety was initially 64, the toll appears to be much higher," said the report.

In another section, it said: "According to initial reports, 64 lives were lost. That estimate was later revised to 1,427."

The government was widely criticized for undercounting the number of people who died on the island as the power outage stretched for months.

After a New York Times analysis in December showed that even the preliminary data from the Demographic Registry of Puerto Rico indicated that hurricane-related deaths may have risen to 1,052, Gov. Ricardo Rosselló commissioned a study from George Washington University's school of public health. The report is expected to be released this month.

"We definitely acknowledge this is a realistic estimate," Pedro Cerame, a spokesman for the Puerto Rican government's Federal Affairs Administration, said of the numbers in the upcoming report to Congress. "We don't want to say it out loud or publicize it as an official number. The official number will come, and it could be close. But until we see the study, and have the accuracy, we won't be able to recognize the number as official."