Why is Connecticut’s death toll so high?
State third per capita in COVID-19 deaths, behind only NY, NJ
With 2,436 coronavirus-linked fatalities so far, Connecticut has experienced one of the nation’s deadliest COVID-19 outbreaks, enduring more reported deaths than some entire countries.
Despite its small size, Connecticut has had more total fatalities than states such as California, Texas, Florida and Washington — as well as the nations of Mexico, Portugal, Turkey and Switzerland, among others. Per capita, Connecticut ranks third nationwide in deaths, behind only New York and New Jersey.
Fairfield, New Haven and Hartford counties all rank among the 25 U.S. counties with the most coronavirus-linked deaths, according to data from Johns Hopkins University.
“Our community obviously has been hit hard,” said Dr. Ajay Kumar, the chief clinical officer for Hartford HealthCare. “There are multiple factors. At some point when we are on the other side of the COVID-19 crisis, we’re going to dissect this a bit more and understand what we could have done differently.”
Beyond the numbers, the deaths have written a new and tragic chapter in the state’s history. Connecticut has seen the deaths of medical workers and nursing home residents, of long-married couples and vibrant young people. Through much of it, the state’s coronavirus victims have died alone and their loved ones have grieved alone.
There is no single answer to the question of why COVID-19 has been so deadly in Connecticut. But experts point to a confluence of factors, including the state’s proximity to New York, its density, the age of its population and its lack of significant social distancing measures until well into March.
Proximity to NYC
New York City has been the nation’s largest COVID-19 hot spot, accounting for more than 20% of all deaths nationally, and Connecticut has seemed to suffer from its proximity to the city.
Early on, at least, Fairfield County residents likely contracted the virus in New York, then brought it home.
“We have people still going into New York daily to commute to
work,” Norwalk Mayor Harry Rilling noted in late March, when his town already had more than 100 cases. “It’s hard to tell them to quarantine because they may be an essential employee in New York.”
The apparent path of the virus, from Fairfield County into New Haven and then Hartford, supports the idea that Connecticut’s outbreak largely began in New York, where an array of factors — from people riding packed subways to the city’s role as an international hub — have contributed to a devastating impact.
“What we saw was that the communities closest to New York City were hit first and were hit hardest, and then the virus spread across the state,” said Summer Johnson McGee, dean of the School of Health Sciences at the University of New Haven. “So certainly our proximity to New York contributed, because so many people work or live in New York who reside in southwest Connecticut.”
That idea is supported by a recent study from the Yale School of Public Health. Based on the genetic makeup of various COVID-19 strains, researchers found that Connecticut’s earliest coronavirus cases had roots in Washington state but that the mid-early surge appeared to have roots in New York state.
Population density
Connecticut is the nation’s fourth-most densely populated state, with a particular concentration of people along I-95 in the southern part of the state and I-91 through the central portion.
Experts and officials say the virus seems to spread more quickly in denser areas — both because density makes social distancing more difficult and because urban settings often have more shared spaces, such as apartment hallways and elevators, where the virus can linger on common surfaces.
These factors likely allowed the virus to spread silently and quickly in the state’s cities, before officials and the public were fully aware of its danger.
“In more rural states, we’re not seeing as rapid of a spread,” McGee said. “So I do think high population density, combined with — in those urban centers, where you really see lack of access to primary care and lack of chronic diseases, that also contributed as well.”
McGee noted that Connecticut’s black and Latino populations, largely concentrated in densely populated areas, have been disproportionately susceptible to COVID-19, likely due to underlying health disparities.
All seven Connecticut towns or cities with at least 80,000 residents have had at least 900 people test positive for the coronavirus. The six municipalities with the highest case rates — Stamford, Danbury, Norwalk, Bridgeport, New Haven and Waterbury — are all larger cities with relatively diverse populations.
Age of the population
Connecticut is the nation’s sixth-oldest state, according to 2017 estimates, and ranked tied for 14th in percentage of the population aged 65 or older, as of 2018.
That means more residents are particularly vulnerable to COVID-19 and live in nursing homes where the disease can spread quickly.
“What’s come out of this clearly is that older populations are at the highest risk,” said Dr. David Banach, epidemiologist at UConn Health. “If we have an older population coupled with a higher population density, that would be what we would associate with a higher mortality rate.”
An aging population could help explain why Connecticut has more deaths per confirmed cases than all but one other state (Michigan), though Banach warned that uneven testing makes that metric unreliable.
Prevalence of older patients appears to be a particularly strong factor in the Hartford area, where outbreaks in local nursing homes have driven up the death toll. Nursing homes residents accounted for 281 of Hartford County’s 442 deaths on April 22, or about 64%. No other Connecticut county had experienced more nursing home deaths.
As a result of the nursing home outbreaks, Hartford County has seen more total deaths so far during the outbreak than New Haven County, and nearly as many as Fairfield County — even with far fewer confirmed cases.
Overall, nearly half of all coronavirus-linked deaths in Connecticut have been nursing home residents, including 43 at Kimberly Hall North in Windsor as of April 24, making it among the nation’s hardest hit facilities.
“There’s truly a devastating number of deaths and cases,” said Dr. Richard Feifer, the senior vice president and medical officer for the company that operates Kimberly Hall North.
The state’s response to the potential for widespread infections in nursing homes has drawn criticism, with many familiar with a plan to segregate COVID-19 patients saying its launch was marred by delays, confusion and miscommunication.
Timing of social distancing plan
Gov. Ned Lamont has been praised for his early response to the state’s first coronavirus case, which included closing schools and public places relatively quickly — the governor declared a state of emergency just two days after the first
Connecticut resident tested positive for the virus.
But Kumar, of Hartford HealthCare, said the entire U.S. was slow to put public health measures in place. Even before the first cases surfaced publicly, the coronavirus seems to have been silently spreading, infecting patients even as public officials at both the state and federal level debated how far to go in shutting down the economy.
“The timing for how [social distancing] was rolled out, in a lot of ways there’s benefit to doing it earlier, but the actual impact that had based on when the decisions were made is a little tough to analyze in hindsight,” Banach said.
In Connecticut, although Lamont asked residents to work from home if possible on March 12, he didn’t order the closure of nonessential businesses until March 23. He also stopped short of the sort of shelter-in-place order other governors have implemented.
At a Tuesday press briefing, Lamont said that now, looking back, most leaders probably wish they had taken earlier action.
“In hindsight, I think everyone wishes they had done everything sooner,” Lamont said.
Lamont added that many of Connecticut’s public health measures were lockstep with those put in place in New Jersey and New York, even though the outbreak started earlier in those states.
New York Gov. Andrew Cuomo — who has also been praised for swift action and whom Lamont has worked with throughout the outbreak — told Axios that he now wishes he had acted as early as December or January.
The counting protocol
Connecticut’s high death rate can be partially explained by the protocol the state uses to count coronavirus-related deaths.
Connecticut, unlike some other states, counts deaths not only in confirmed COVID-19 cases but also in “probable” cases. This counting protocol, leading to a higher official death toll, aligns with the Centers for Disease Control and Prevention, which on April 14 began counting both confirmed and probable COVID-19 cases and deaths.
On April 20, when Connecticut conformed with this guidance and began counting probable cases, the state’s death toll jumped by 204.
Experts say places that do not count probable deaths are likely undercounting the actual number of COVID-19 fatalities. When New York City began counting probable deaths in mid-April, for instance, it added 3,700 victims to its count in one day, The New York Times reported. And in Massachusetts, the Boston Globe reviewed state death records and found the state is likely severely undercounting coronavirus fatalities.
Using data from 12 different countries, The New York Times found that tens of thousands more people have died during the pandemic than usually die during the same time frame — and that’s not counting deaths officially attributed to COVID-19.
In Connecticut, state epidemiologist Dr. Matt Cartter said that “probable” cases include people who were exposed to the virus and had symptoms of the disease but were not tested for the virus. The Office of the Chief Medical Examiner makes the determination of a “probable” case, Cartter said.
“Many of the states, including states around us, do not count probable deaths due to COVID-19 and are only reporting out … laboratory confirmed deaths,” he said.
Still, Cartter said, there’s no question Connecticut’s outbreak has been severe.
“I don’t want anyone to misunderstand,” Cartter said. “Connecticut, New York, the New York City metro area, parts of New Jersey, Boston, Providence — we’re all seeing levels of COVID-19 activity that many parts of the country are not experiencing.”
Widespread testing lacking
Kumar also said the U.S. was slow to implement COVID-19 testing in the first place — and that communities across the country are still paying for that lag.
Widespread testing would allow officials to identify COVID-19 cases quickly and instruct infected patients to self-quarantine to avoid spreading the virus. This is particularly important because some people infected with the coronavirus may have mild or no symptoms but still spread the virus to others.
Kumar said the slow identification of COVID-19 cases also delayed the contact tracing process, which can’t begin until a case of the disease has been identif i e d . T h i s may h ave kneecapped the state’s chance to contain the virus at an early stage.
In Connecticut, no more than 2.6% of the population had been tested for the coronavirus, according to state data provided Tuesday. All three of Connecticut’s neighbors — New York, Massachusetts and Rhode Island — have conducted significantly more tests per capita. Connecticut as of Saturday had identified 2 9, 2 8 7 COVID-19 cases, but experts and officials have repeatedly said that the actual number of cases is likely much higher.
McGee said all states, including Connecticut, would have benefited from more testing early on.
“There’s no doubt that, had we been able to do more testing sooner and more contact tracing, that we would have found more cases and prevented more community spread,” she said. “That’s something we have to get right as we think about opening.”
Courant staff writers Christopher Keating and Dave Altimari contributed to this report.