Hurricane Maria Death Tally: A Matter of Credibility
“We have not talked about a preliminary report; we spoke about the final report, which corresponded to a May 22 delivery. That’s the only date talked about,” said the spokesperson for the Puerto Rico Public Safety Department, which commissioned George Washington University (GW) to produce a study on the actual number of deaths corresponding to Hurricane Maria’s devastating path across Puerto Rico on Sept. 20, 2017.
Contrary to what was expressed by the umbrella agency headed by Héctor Pesquera, Caribbean Business confirmed that the agreement signed not only included delivery of that report, but also the payment of $125,000.
Since May 4—long before the discrepancy arose between the official data and the contract—Caribbean Business submitted multiple requests for information from the communications office of GW’s Milken Institute School of Public Health. The petitions have included specific questions about the status of the reports, which cost the island’s Treasury $305,368; to interview the leader of the investigation, epidemiologist Carlos Santos Burgoa; and the amount the administration of Gov. Ricardo Rosselló must pay to receive the final answers to its questions.
“I’m connecting you with Yennifer Álvarez, the governor’s press secretary. She can answer your question,” replied the institution’s communications manager 80 days after CB’s first request. Immediately after, Álvarez contacted this publication, confirmed the existence of the preliminary report and coordinated its delivery through the Public Safety Department.
Preliminary report or sales pitch?
Caribbean Business found the GW document, “Ascertainment of the Estimated Excess Mortality from Hurricane Maria in Puerto Rico,” hardly qualifies as a preliminary report. Beyond raising concerns about the work that has been conducted since Feb. 22, it reflects poorly on the administration’s unusual insistence on its commitment to transparency.
“This preliminary report provides an update on the activities for the project, from its launching to date. It is a narrative of performance more than products,” reads the introduction of the 24-page document, which suggests the report is solely a document on contemplated methodologies and a proposal for a second phase, an “in-depth evaluation of the mortality,” to close a lucrative deal with the government.
“I recommend watching the press conference where it is explained that the first phase has a cost of $305,000 and, to complete the next phase, could reach $1 million; after completion of the first phase, a search for grants and/or donations to subsidize the rest will be worked on,” Rosselló Nevares’ press secretary said in May when asked about the total cost to commission the study. Álvarez resigned from her post on July 25.
The Milken Institute says in its report that it has contacted or received contributions from at least 24 entities, most of which are state and federal. However, private universities such as Penn State (No. 12 in QS World University Rankings), Harvard (ranked No. 2 worldwide by QS) and Ponce Health Sciences University do not appear on the list. These independent institutions have valuable data that could accelerate the investigation carried out by GW (ranked 345 worldwide by QS), a congressionally chartered nonprofit, which incorporates the institution at the federal level.
However, GW assured it will use Penn State, Harvard and Center for Investigative Journalism studies “to provide a way for the public to assess the relative differences between them and those produced by the current project.” “In the meantime, the work continues to develop the proposal for the in-depth analysis of the cause-specific mortality attributable to the hurricane, including more complex statistical modeling as well as an in-depth survey on death circumstances. Funding to complete this part of the project is actively being sought from different sources and institutional synergies,” reads the last paragraph of the report, which adds that a final report is expected to be produced “within the timeframe of the contract.” As recently as June 30, the contractual term was extended until Sept. 18 through an amendment to the original agreement.
“Since I’m not involved in this project, it is difficult for me to determine if it is a preliminary report. Certainly, this study is expected to certify and/or approximate a number of deaths associated with Hurricane Maria; this is according to the public expressions of the secretary of Public Safety himself. Since that data isn’t available, I believe there is still work to be done or information to be included,” said the director of a specialized program in estimates and projections at Pennsylvania State University, Dr. Alexis Santos Lozada. His comments directly addressed a controversial statement Pesquera made after publication by the New England Journal of Medicine of the Harvard study, whose researchers estimated there were 4,645 (95 percent confidence interval, 793 to 8,498) additional deaths on the island in the three-month period following Maria.
“Although we have not seen Harvard’s study, it is based on a survey; it’s not based on scientific data. George Washington’s comes with a different methodology. At some point, I’m sure GW is going to get up to speed with them and they will try to validate it, but a survey is not the same as a scientific basis, which is what George Washington is carrying out,” Pesquera said May 31.
For the expert in applied demography and population health, it is important for GW to establish what is meant by “rapid response,” which the school cites in the preliminary study to differentiate its research from other “more conventional research projects,” according to the report.
“We would have to ask the GW researchers what they consider is ‘rapid response.’ Certainly, this study comes after reports that have been available since November (for example, the study by Jeffrey Howard and I) and other studies by reporters. Our report was published two months [faster] after the phenomenon and when a large part of the island was still in the dark and without basic services. This report comes 10 months later, which is a slower response than all other analyses,” added Santos Lozada, who led an investigation that yielded an estimated 1,085 deaths from the hurricane, and was published in April by the journal Health Affairs.
“After analyzing the successes and failures experienced as part of the emergency after the passage of hurricanes Irma and Maria, our government is reviewing all protocols and processes to ensure we are ready for a new hurricane season,” Rosselló said March 20. At the time, the governor announced he would hold a summit May 29 with central government agencies and the Federal Emergency Management Agency, and a Municipal Emergency Summit the next day, the conclusions of which are unknown more than a month after the hurricane season began.
The contract between George Washington University and Puerto Rico’s Forensic Sciences Bureau: