Add To Favorites

Did coronavirus kill them? A Courant investigation of death certificates in three Connecticut towns uncovers a confusing patchwork

Hartford Courant - 6/28/2020

When Delphine Green died April 4 at the Riverside Rehabilitation and Health Center in East Hartford, the cause of death was originally listed as dementia.

But Green’s sister, Linda Mickels, was suspicious and asked staff whether it could have been the coronavirus that killed her. She said she was told the coronavirus wasn’t in the building then. The first official COVID-19 death at Riverside wasn’t recorded until four days later.

“I knew that something didn’t seem right; all of a sudden she was on oxygen and then they called me and said they were putting her on morphine,” Mickels said.

Chief Medical Examiner James Gill confirmed her suspicions. After being notified of Green’s death, he investigated and determined that she did, in fact, have COVID-19; it was one of five death certificates Gill said he amended at Riverside.

The Courant analyzed death certificate data for the month of April, when COVID-19 deaths soared, from three towns hit hard by the coronavirus pandemic -- East Hartford, Torrington and Milford -- in an effort to look beyond the official numbers and get a more complete picture of how the virus was racing through the state.

The picture that emerges is a confusing one, raising questions about whether we yet have a full picture of how insidious and deadly the coronavirus actually is.

A number of deaths -- especially in the early days of the pandemic -- were either not reported to the state medical examiner or not attributed to COVID-19. Despite continuing efforts to capture the true extent of the death toll, the actual number of coronavirus deaths may well exceed the 4,311 confirmed deaths that have already made Connecticut one of the hardest hit states in the nation.

The data also raises two critical questions that may only be answered when the full picture is analyzed: Was the deadly virus far more prevalent in Connecticut early than many have assumed? And, more pointedly, could many of those who perished have been spared with more decisive action earlier in the course of the pandemic?

“In the future, perhaps when this wave of the pandemic is over, epidemiologists will use mathematical modeling to better estimate the numbers of COVID-19 illnesses, medical visits, hospitalizations and deaths,” said Dr. Matthew Cartter, the state’s epidemiologist.

The three small cities analyzed by The Courant, while in different areas of the state, have several similarities. All had nursing homes with large numbers of deaths. Two of them, Milford and Torrington, have hospitals that serve their communities, and all three have state-run group homes or assisted living facilities.

Death certificates offer insights

Even with the dramatic numbers being issued daily by the governor’s office in April and May, measuring the extent of the death toll has been an inexact science.

Analyzing death certificates is the only way that experts truly can determine how a pandemic has impacted a community or a state. They include not only what caused a person to die, but where they died and specific information about them such as whether they had other underlying health conditions, their age and in some cases, their race.

Death certificates -- completed by doctors or nurse practitioners -- are generally filed in the town where a person died, though they are eventually gathered by the state Department of Public Health. But even under normal circumstances the department is months behind in cataloguing and even further behind in making them public.

“Death certificates are an important source of information when trying to assess the impact of an epidemic or pandemic, especially when this information is needed in near real time to guide the implementation of public health control measures,” said Cartter.

Cartter said it is likely that COVID-19 deaths have been underreported just as the case is with the regular flu every season. “I suspect that underreporting will also be true for COVID-19 for many of the same reasons as for influenza,” Cartter said.

COVID, or not?

The death certificates on file in East Hartford include those who died at the Riverside Rehabilitation and Health Center in East Hartford, one of the hardest hit nursing homes in the state, with 62 COVID-related deaths.

The deaths at Riverside overshadow the city’s numbers -- of 56 people who died in East Hartford in the month of April, 41 died at Riverside. Of the 41 deaths at Riverside, 29 were listed as COVID-related.

The Courant’s analysis found five deaths at Riverside that Gill said he later amended the death certificate to include COVID-19. Like Delphine Green, the cause of death initially listed for Loida Alemedina was dementia. And -- also like Green -- Gill did a port-mortem analysis and concluded she was found to be COVID-19 positive.

The death certificates of three others were amended: David Baker, who died on April 1, Rugelio Rivera, who died on April 12, and Evelyn Baker (no relation to David Baker), who died on April 22. Rivera and Evelyn Baker’s death certificates originally listed dementia as a cause of death, while David Baker’s listed cerebral palsy.

The accurate reporting of deaths during the pandemic has been complicated by a number of factors, including shifting guidance from the federal Centers for Disease Control. In early April, the CDC broadened the criteria for attributing a death to COVID-19. Many of those who were dying had never been tested for the virus.

Overall in the three cities analyzed by The Courant, Gill amended at least 10 death certificates to classify them as “suspected COVID-19.” In many cases, it was too late to do a COVID test so the diagnosis of suspected COVID was based on medical records.

In Milford, Gill’s office made a frantic attempt to find the body of Jean Auclair, who died at the Golden Hill Rehabilitation Center on March 30, even racing to the crematorium, but they were too late to test him for the virus.

Instead, Gill reviewed the medical records and changed the 92-year-old man’s death certificate to include COVID-19 as a likely cause. Golden Hill had one other case in early April in which the death certificate was amended.

Across the state, Gill has he has amended more than 175 death certificates, although he said recently, he has turned those cases over to DPH to review. The majority of the amended death certificates are either nursing home deaths or people who died at home.

Gill said he wonders about other nursing homes in the state and whether they have accurately reported all of their coronavirus deaths, particularly in the beginning of the pandemic, but he said that’s a more appropriate investigation for the state Department of Public Health to undertake.

Routine deaths in nursing homes are not typically reported to the medical examiner, but on March 30 the state Department of Public Health ordered the reporting of all confirmed and suspected COVID-19 deaths.

Responding to questions about the death certificates, a spokeswoman for National Health Care Associates, which owns Riverside, said “unfortunately, some of your information is incorrect, but due to HIPPA requirements and other health care privacy regulations we are prohibited from discussing specific patient circumstances.”

The company declined to elaborate.

“When faced with the unfortunate passing of one of its residents, Riverside followed directives from the state, including the requirement to notify the Office of the Chief Medical Examiner,” the spokeswoman said. “Further, only either a physician or a nurse practitioner may determine cause of death.”

East Hartford Mayor Marie LeClerc said nursing homes like Riverside were up against an array of factors as they dealt with the intensity of the virus.

“The impact has been significant and has been emotional for the staff, families and community,” she said. “I believe the size of the facility, the significant vulnerable population who have underlying and complicated health issues, along with the state coding all deaths as COVID-19 probable should be taken into consideration as you look at these numbers.”

Even more deaths possible

The Courant’s review of death records uncovered another category that could push the state’s final coronavirus death toll higher.

A number of deaths -- at least nine found by The Courant -- had certain earmarks in terms of cause or timing that Gill said he would like to have reviewed, but he was unable to because the body had either been buried or cremated. Under normal circumstances, deaths from issues such as respiratory failure, heart disease or other ailments are not routinely brought to the attention of the medical examiner.

Gill said, for example, his office wasn’t aware of Doris Jean Gunter’s death at Riverside until The Courant asked if he had reviewed her death certificate. Gunter died on April 12. Her cause of death was listed as dementia. Her family didn’t want to talk about her case.

The Wolcott Hall Nursing Center in Torrington had four cases that weren’t reported to the medical examiner that Gill said he otherwise might have reviewed. They included two diagnosed with dementia whose death certificates list “failure to thrive” as a cause of death, another who died of respiratory failure and a fourth who died of a heart attack.

“During this timeframe in March and April when testing was less available, Apple Rehab monitored all patients that were symptomatic and tested patients based on the recommendation of the medical practitioner,” said Karen Donorfio, a spoksewoman for Apple Rehab, which owns Wolcott Hall. “DPH guidance and strategy for all resident testing through a Point Prevalence Study was released on May 11.”

“If a resident were tested in our facility, then the result we receive is dependent on the lab and the type of test,” she added. “Apple uses the nasopharyngeal swab as directed by the DPH. It is possible that a different method is used post mortem. DPH is also aware that there have been false positives and false negatives. Apple Rehab reports based off the result provided by the lab.”

Donofrio also noted that the medical examiner “may find a different result than what is stated on the death certificate and we would be notified at that time. We were notified of a change for one patient.”

The Courant shared death certificate data with Gill to try and determine how people died in nursing homes and to see if there were possible deaths before the state announced its first COVID-19 death in mid-March -- something that Gill also has quietly been investigating. Cartter said epidemiologists also will study other non-COVID deaths as well as where people died to see if the virus had any impact those cases.

“I reviewed all deaths reported to us in February and March. I found one death in February of a man who died of pneumonia at a nursing home,” Gill said. “He did not have an autopsy. So there is nothing that can done at this point to further investigate the death.”

Another area that Gill has studied to get a better handle on the overall toll is at home deaths during the pandemic.

By Late April Gill had done 137 tests for COVID-19, of which 73 -- more than half -- came back positive. The majority of those tests were taken at funeral homes; some of the others were done during autopsies.

There were only four at home deaths in East Hartford in April -- a small number compared to Milford, which recorded 24 at home deaths in April. More people died at home in Milford in April than in any other location during that period. Only two of the 24 at home deaths in Milford were labeled COVID deaths.

But as the death toll piled up across the state in April, Gill was able to do fewer investigations of at home deaths because of time constraints.

One of those at home deaths was Nancy Griffeth, who died on April 24 in the East Hartford home in which she had raised her family. Her death certificate said she died of Alzheimer’s.

Her daughter Vickie Griffeth made a critical decision in late February when doctors said that her 68-year-old mother’s Alzheimer’s disease was getting progressively worse -- and perhaps she needed to go into a long-term care facility.

But as Griffeth saw the news that the coronavirus was arriving in Connecticut, she wondered if putting her mother Nancy in a nursing home would be a death sentence.

“I said, ‘I’m not going to risk her life. I know she’s dying, but I’m not going to let her go out that way. Not for my sanity,‘” Griffeth said.

Vickie Griffeth said in a recent interview she is secure in the knowledge that she did the right thing, considering the number of people who have died alone in nursing homes.

“I feel like I made the right decision,” she said. “I could hold her and tell her I love her.”

Heart attacks in Torrington

Outside of deaths caused by suspected exposure to the coronavirus, the data analyzed by The Courant raises a number of other questions that are hard to answer, such as: Could other deaths from cardiac arrest or strokes be attributed to the virus?

In Torrington, Fire Chief Peter Towey’s said he had a sense when going on emergency calls to people’s homes, they “definitely seemed sicker.”

The death data shows the city had 14 cardiac arrest deaths in April of people who were at home when they were initially stricken. Six died at home, while eight died at Charlotte Hungerford Hospital after being transported there by ambulance.

Were the heart attacks fatal because people feared going to the hospital? Is it possible they had COVID-19 but were never tested?

“I’m not sure we saw an increase in the number of calls, but there was an increase in the severity of the cases,” Towey said. “We talked about whether people were doing their best to stay at home and avoid emergency rooms or hospitals where they knew COVID was. That may be why people seemed sicker, but there’s no way to know that for sure.”

The Milford death records show four possible suicides in April and another still under review that occurred in March. Gill said he is still waiting for toxicology tests to determine the causes of death in those cases.

By contrast, Torrington had only one potential suicide and East Hartford none. Three of those Milford cases were men in their 30s. The Courant contacted two of the families, but both said the deaths were still too raw to talk about.

Torrington had the most overall deaths in April of the three cities studied, with 114 deaths, the majority at Charlotte Hungerford Hospital, where 63 died, which didn’t surprise Towey.

“Charlotte is the only hospital in the immediate area, so people from all over the region end up there,” Towey said.

The chief said there’s no doubt many of the hospital cases were COVID-19, and 51 of the hospital deaths were eventually classified as COVID related.

“I have an entirely new appreciation for hospital workers now.” Towey said. “It’s a totally different world now.”

Families frustrated, nervous

One of those deaths at Charlotte Hungerford Hospital was 46-year-old Eddie Miguez -- a resident of the Torrington Family Respite Center, a state-run group home. He was one of two people from the facility to die in late March and early April.

Family members hadn’t seen him for some time because the group home had barred visitors in early March as a precaution to keep COVID-19 out of the facility. So when the facility called at 5 one morning a few weeks into the lockdown to say that Eddie had fallen in the bathroom and was taken to the hospital, his brother John Miguez was shocked. The staff told John that Eddie must have had a seizure, which seemed odd to him.

“Eddie had Down syndrome and his vision was poor, but he had never had a seizure in his life,” John Miguez said.

Eddie went to the hospital but came back to the group home a day later seemingly unscathed.

“But not too soon after he got back to the group home, we get another call that Eddie had COVID,” John Miguez said. “I’m thinking he had it all along and that’s why he fell in the first place, especially since he died pretty quickly.”

The second day of Eddie’s hospitalization for COVID-19, the family got a call: Eddie wasn’t going to make it. He died the next day. His death certificate lists his cause of death as “acute hypoxic respiratory failure due to COVID-19 virus due to Downs Syndrome,” which irks his family.

“Eddie didn’t die because of Down syndrome. He died of the damn virus,” John Miguez said.

John Miguez said the holidays won’t be the same without his brother. The family always brought Eddie home for Christmas so he could be with his family.

“He loved Christmas. He’d be walking around hugging everybody,” John Miguez said. “His death was way too quick. I just want him to be alive and well.”

The information in this story was obtained with the help of genealogists Jenny Hawron and Richard Plant, who were able to access death records during the pandemic shutdown.

___

(c)2020 The Hartford Courant (Hartford, Conn.)

Visit The Hartford Courant (Hartford, Conn.) at www.courant.com

Distributed by Tribune Content Agency, LLC.