What’s in the censored Florida Medical Examiners database of COVID-19 deaths?
FLORIDA TODAY obtains data revealing patterns in COVID-19 deaths that the state doesn't want you to see. Florida Today
EDITORS NOTE: With this story we are providing information about the COVID-19 crisis that state leaders don’t want you to see. Normally, this is the kind of exclusive reporting we would limit to subscribers only. We chose instead to make this story more broadly accessible because of the importance of its content. But quality, fact-checked reporting isn’t free. Collecting and verifying news is a costly and time-consuming endeavor and perhaps has never mattered more than it does now. Please show your support for quality reporting like this by subscribing to FLORIDA TODAY. Journalism matters. Subscribe here.
When a 66-year-old man was found dying on an Amtrak train passing through Okeechobee County on April 5, there was nothing to indicate that he had COVID-19. It was the local medical examiner's office that pieced it together.
The examiner discovered the man had recently arrived with a fever at New Jersey's Newark Liberty International Airport from the United Kingdom. The Centers for Disease Control stopped him from boarding a flight to Florida and sent him to a local hospital for a coronavirus test. Released before the results came back, he got on a southbound train, went into cardiac arrest while traversing the Sunshine State, and was pronounced dead at a Florida hospital.
The district medical examiner collected the facts, swabbed the body, confirmed coronavirus and entered the man's death into the Medical Examiners Commission record as a COVID-19 fatality on April 13. The detective work that pulled this story together and exposed the missteps of the haphazard response to the COVID-19 pandemic well over a month into the crisis is part of the record that the State of Florida doesn’t want you to see.
Throughout the pandemic many of the state's 25 medical examiner district offices have been doing this work, producing critical information on COVID-19 deaths that could be used to help policy makers better plan their response for the crisis going forward.
But since at least April 20, the Florida Department of Health has blocked the Medical Examiners Commission from releasing their own detailed spreadsheet of the COVID-19 dead. On Wednesday, the state released the medical examiners' spreadsheet but redacted the narratives and cause of death entries.
Before the DOH clamped down, FLORIDA TODAY obtained an unredacted version of the spreadsheet from April 15, which recorded the first 601 Florida deaths attributed to COVID-19.
Though not fully standardized, the data provide grim insight into the early failures of state officials and the medical system to contain and respond to the virus. It tells a story of patients denied testing until their second or third hospital visit, and reveals deadly clusters of infection at nursing homes and cruise ships.
The data also includes details on the underlying conditions of patients and how long they were in hospital before they died — vital information needed to better understand the disease.
At times, the Medical Examiners data tallied more deaths than the FDOH, recording as much as a 10% difference. But DOH officials said it was not a dispute over numbers. According to a spokesperson, the DOH only wanted to block the data release out of concern for privacy.
“I think there was a lot of uncertainty, a lot of concern about how the data might be interpreted or misinterpreted," said Jay Wolfson, a public health expert and Senior Associate Dean at University of South Florida's Morsani College of Medicine.
"I know that at an executive level in this state, there were grave concerns about the data being misused or creating a sense of panic," Wolfson said, noting that the medical examiners' descriptive information was not in a neatly standardized format.
A major point of contention between the Department of Health and medical examiners was the fact that DOH had been excluding non-resident deaths from its count. Whereas medical examiners count deaths where the person died, DOH's practice has been to count deaths based on the deceased's state of residence. This means that losses among seasonal residents — who account for over 1 million people — have been excluded from the state's death toll.
Whether that is still the case, however, is unclear; the total counts from the medical examiners and DOH now are separated by just a handful of cases.
A FLORIDA TODAY analysis of the first 601 COVID-19 deaths recorded by medical examiners — the state's total death count now tops 1,600 — found several patterns.
Details from the medical examiners’ narratives paint harrowing tales of patients' final days. They show a chaotic handling of patients in early days of the pandemic, with healthcare systems scrambling to keep pace. Patients were often turned away for coronavirus tests, or sent home, only to die days later. Some patients were bounced around repeatedly before finally getting tested, some only after they had died.
Among the findings:
- Care facilities, conferences, nursing homes and cruise ships drove deadly clusters early on in the pandemic. At least 24 cases had links to cruise ships.
- Many early deaths were from untraceable cases, meaning contact tracing either didn't happen or fell apart immediately.
- The first hospitalizations and first confirmed case through testing came a week before the first publicly confirmed cases.
- Hospital stays varied significantly, from patients dying the same day they went to the hospital to languishing 42 days before expiring. But on average Florida hospitalizations trended less than a week.
- 87% of the deaths were people over age 60. The youngest was a 27-year-old cruise ship worker from Broward County, who also had the flu. The median age was 77.
- Diabetes is among the deadliest underlying conditions, noted in about a third of the COVID-19 deaths. Heart disease, hypertension and problems associated with obesity also were common co-factors.
Wolfson believes the data should be public, even if preferably more standardized and its limitations clearly noted. Such data, he said, would allow the public and private sectors to better plan, and move forward to ensure the health of communities in the long term.
"Reporting should not mask potentially sensitive areas of economic or political cause (and) effect," he said.
The ME data crucially identified clusters of infection well before the state published lists of nursing homes. It also pointed to specific cruises, conferences and other institutions that deaths were linked to.
Medical examiners noted links to cruise ships — passengers, employees and port workers — in 24 cases. Among those, seven were explicitly linked to Carnival's Holland America Zandaam cruise ship. Royal Caribbean is also mentioned.
"If there's one or two cases that's one thing, but when you start getting clusters there's an imperative, public health policy reason to report and use those data effectively," Wolfson said.
Nursing homes were an early and enduring epicenter of deaths from the novel coronavirus.
As of May 7, the Florida Department of Health listed 622 COVID-19 deaths of residents and staff in nursing homes, a third of them in Dade and Broward counties. That's 39% of the 1,600 Florida residents who have died from the disease. But before the DOH succumbed to public pressure to release names of nursing homes, medical examiners were identifying facilities in their notes.
Prisons and jails also were identified early on as potential cluster areas.
At least three inmates figured among the first 601 COVID-19 dead, according to the medical examiners' database. Two who died on April 9 and 12 were housed at the Blackwater Correctional Facility in Santa Rosa County.
One, a 69-year-old male went to the infirmary with shortness of breath, a cough and diarrhea. Four days later, when he was taken back to his cell, "he laid down in front of the cell door, went unresponsive and was pronounced deceased shortly after at the facility."
That facility has since recorded four more deaths.
Jails and prisons continue to remain at risk but the officials in charge of those facilities often offer less-than-transparent assurances.
At least three corrections deputies at the Brevard County Jail have tested positive for COVID-19 but no inmates have been reported to have contracted the disease, according to Sheriff Wayne Ivey. Yet neither the sheriff's office nor the health services contractor Corizon Health have provided answers on how many inmates have been tested.
Conferences also may have been early incubators of the disease.
On March 23, an 84-year-old Lake County woman had a fever and vomiting and was admitted to a hospital, diagnosed with pneumonia. She had attended a “State Trooper” conference weeks prior to her death. She tested positive for COVID-19, and the medical examiner noted other conference attendees also had tested positive.
Through April 15, the youngest to die from COVID-19 was a 27-year-old housekeeper on Royal Caribbean's Symphony of the Seas cruise ship.
According to the medical examiner's data, on March 29, the housekeeper had flu-like symptoms with a fever, cough, and shortness of breath. He was eventually airlifted to a local hospital where he was put on a ventilator. He tested positive for COVID-19. By April 12, he was dead, with influenza A listed as a probable cause of death along with COVID-19.
Global travel and untraceable first deaths
The first confirmed death in Florida hit March 5. It was a Lee County woman, 77, who'd traveled to the Dominican Republic with her husband. The couple returned to Fort Myers on Feb. 15, not feeling well.
They visited family in Fort Lauderdale on Feb. 23 and went to the hospital two days later. They returned to Fort Myers, then, on March 4, a friend found them with fevers. They were taken to a hospital. Two days later she died. Complications in her death included chronic obstructive pulmonary disease (COPD) and a heart attack.
Florida's third person to die, a 77-year-old man from Lee County, had a medical history of heart disease, diabetes and COPD and had been ill for two months. He’d visited his pulmonologist multiple times for antibiotics and steroids.
On March 10, he went to the hospital with shortness of breath. He had a fever and chills, so was swabbed for COVID-19. He went into shock and died on March 14, four days after entering the hospital. He had no known recent travel history or exposure to anyone with COVID-19, so "the exposure was assumed to be community based," the medical examiner noted.
Even before March, it appears the virus was spreading undetected.
Reporting by The Palm Beach Post points to confirmed patients reporting symptoms as early as January. The Medical Examiner data reveals that the first COVID-19 patient confirmed through testing was at least as early as February.
On February 24, an 87-year-old Duval County woman was admitted to the hospital for lethargy. A COVID-19 test came back positive, making her the first confirmed case through testing, a week before the state announced presumptive positives on March 1.
The woman also tested positive when a second test was performed on April 1. She died on April 4 after battling the disease for a near-record 40 days in hospital.
The longest hospitalization — recorded at 42 days — was a 55-year-old Dade County man. He was not tested until after he died on April 10. These two cases, however, are outliers. The overall trend in Florida was shorter, not longer hospital stays.
Hospital stays trend shorter in Florida
Florida's hospitalization lengths were days shorter on average than in many countries around the world. Florida's median hospital length for the deceased was seven days, with 45% of the decedents dying within 6 days.
The numbers are closer to those observed in the much more overwhelmed medical system in Italy, where hospital stays before death averaged eight days.
Average data from cases across China, Italy and South Korea used by the University of Washington modelers is longer, with hospital stays prior to death being estimated up to 10 days.
In the UK, the country's Chief Medical Advisor Chris Whitty reported average hospital stays for COVID-19 patients of 8 days for those not requiring ventilation, and 16 days for those requiring more advanced care.
Several factors may be causing Florida patients to go to the hospitals relatively late, according to experts, including their income bracket, race, as well as a pervasive fear that going to a hospital could result in them catching the virus.
"There's good evidence that African Americans and Hispanics have tended to avoid the system because they don't trust it," USF's Wolfson said.
A West Health and Gallup poll last month said that as many as one in seven Americans would not seek treatment if they thought they had the coronavirus over fear of the medical costs.
Wolfson said another possible reason people did not want to go to the hospital is that public health communication faltered and was late in the crucial early days of the pandemic.
"I think part of it is we did get a late start in this country, and we got a later start in Florida. We weren't quite sure how to mobilize our communications resources, and there are lots of communities in this state that don't get access to information about the health of their communities and what they should do."
Diabetes and coronavirus, a deadly mix
Of the first 601 deaths in Florida, 191 of them, almost a third, the medical examiner makes some mention of diabetes as a contributing factor.
That's because infections like COVID-19 raise blood sugars more complications, according to Tracey D. Brown, CEO of the American Diabetes Association.
Some states fare even worse than Florida, Brown added.
"There are states reporting that over 40% of the people who have died of COVID-19 had diabetes," Brown said. "That’s a disproportional and alerting number.”
The medical examiner data falls in line with what the Centers for Disease Control and Prevention is finding nationwide: the clear majority of hospitalizations also have some kind of underlying condition, most commonly hypertension (49.7%), obesity (48.3%), chronic lung disease (34.6%), diabetes mellitus (28.3%), and cardiovascular disease (27.8%). Other CDC research has found that in about 90% of COVID-19 deaths, the victim had at least one underlying condition.
"All these chronic conditions compromise your immune system," said Ken Thorpe, chairman of the Partnership to Fight Chronic Disease and a professor of health policy at Emory University in Atlanta.
Compared to COVID-19 patients with no underlying conditions, those with diabetes and other chronic disease are almost twice as likely to suffer a poor outcome: being hospitalized, put on a ventilator or dying.
The pandemic proves the nation needs more investment in public health infrastructure, such as increased funding for public health departments and better access to primary care, Thorpe said.
"I think the lesson going forward, is you can't be complacent," Thorpe added. "We need to start thinking about the next version of this ... We have to identify people and isolate them."
Turned away, then died
But the actions of hospitals and doctors may be the most sensitive data noted by medical examiners.
One pattern that emerges from the medical examiners' accounts is patients sent back home without a COVID-19 test or with results pending, only to return to the hospital and die days later, or sometimes to never return and die at home.
A 71-year-old Hillsborough County woman went to the emergency room complaining of nausea and was put on a breathing treatment. A flu test came back negative.
"At that point the ER doctor wanted to admit her and run the COVID-19 swab test, however, according to the nurse's notes, the admitting doctor advised against this and discharged the decedent home and a COVID-19 test was never performed," the medical examiner noted.
On March 24, her family grew concerned since they had not heard from her. She was found unresponsive and not breathing, rushed to the hospital, and pronounced dead in the emergency room.
Others were sent home on doctor's orders.
A 79-year-old woman in Palm Beach County suffering from dementia was given antibiotics and sent home. On March 24, she complained of a fever and cough. She was transported to the hospital and given a Z-Pak course of antibiotics and discharged.
She returned to the hospital two days later, due to diarrhea and respiratory distress. She was admitted and given broad-spectrum antibiotics. Then on April 2, her lungs showed "severe progression of probable viral pneumonia" and she was given oxygen but refused to be intubated. Two days later, her COVID-19 result was listed as "DETECTED." On April 5 she was transferred to hospice where she died six days later.
Some were simply turned away for testing, because they lacked certain symptoms.
On March 18, a Hillsborough man, 63, went to get tested for COVID-19 but was denied because he had no cough, the medical examiner noted. Four days later, he went to the emergency room, with shortness of breath, aches, fever and chills. He was admitted with diagnosis of diarrhea, and tested positive for COVID-19, the examiner noted.
"There was, again, a lack of consistency, a lack of testing kits, a lack of resources, a lack of trained personnel. We weren't ready for this, and certainly in the early days, hospitals weren't sure what to do," Wolfson said.
The data shows this.
“It's fair to say that we weren't prepared. We didn't have a plan. And we were late coming in, and we were not very good at coordinating things in great part because the Department of Health in this state over the last decade, had been gutted," Wolfson said.
Alessandro Marazzi Sassoon is a watchdog reporter for FLORIDA TODAY. Contact him at 321-355-8144, or firstname.lastname@example.org. Twitter: @alemzs
Jim Waymer is environment reporter at FLORIDA TODAY. Contact Waymer at 321-242-3663 or email@example.com. Twitter: @JWayEnviro Facebook: www.facebook.com/jim.waymer