Earlier this month, one of the darkest images of the pandemic seemingly brightened overnight. On April 14, Massachusetts had one of the highest reported nursing home COVID-19 death rates in the country, with 9,018 dead.
The next day, it plummeted 39 percent to 5,502, according to the official state count.
Massachusetts officials had contacted the Globe in advance to warn that a drop in the death count was coming due to a new, stricter method of counting COVID-19 deaths in nursing homes. The new approach would follow national standards, they said, making Massachusetts more consistent with other states.
But the drop was more than twice as large as the state officials described in the briefing — and twice as large as the media initially reported. The bigger 39 percent drop instantly invalidated widely reported claims about the severity of the pandemic in nursing homes. Before the change, the math showed that roughly one of every four Massachusetts nursing home patients had died of COVID-19. Now, it is about one in seven.
Some critics and researchers are puzzling over the new approach, which they say may underestimate the true death rate.
“How will history count these deaths?” asked Representative Ruth Balser, a Newton Democrat who has been a leading advocate for greater public reporting of COVID cases and deaths. “We are left with questions.”
“It’s a mess,” added Barbara Anthony, former undersecretary of the state’s consumer affairs office and now a senior fellow in health care at the Pioneer Institute, who believes Massachusetts is now undercounting nursing home deaths.
The concern, to a large degree, revolves around how to ensure that each person listed as a nursing home death really lived in a nursing home and really died from COVID-19.
Until April 15, on its daily dashboard Massachusetts had used a sweeping definition for COVID deaths in long-term care. It was counting any resident or staff who contracted COVID at any time prior to their death in a nursing home or a hospital or rest home. Even those who tested positive but later died of an apparently unrelated cause were counted as a nursing home COVID death.
But the state was already reporting deaths in a more narrow way to the federal government, using the National Healthcare Safety Network tracking system, which the Centers for Disease Control and Prevention required states to start using last May. That system doesn’t include staff deaths. Nor does it count as COVID-19 deaths people who apparently recovered from the virus but later died of causes believed unrelated to COVID.
Over time, the gap between the number of nursing home deaths the state was reporting publicly and reporting to the federal government grew into the thousands. However, the state did not draw public attention to that, and reporters routinely relied on the dashboard for authoritative numbers. In late March, for instance, the Globe reported that “more than 8,800 long-term care residents have died of COVID-19 in Massachusetts in the last 12 months.”
Then, last week, state health secretary Marylou Sudders and other officials told the Globe that, starting April 15, the state would use only the more narrow National Healthcare Safety Network standard for counting nursing home deaths. That would reduce the death toll from 6,722 to 5,502, a drop of 18 percent, state officials said. The state purged references to the larger numbers of nursing home deaths from its dashboard.
“We see it as a maturation of our dashboard,” Sudders said. “It will allow us to be consistent with what other states are reporting on their dashboards. The more we learn about COVID-19 in this pandemic, the more it allows us to make our data more precise.”
But that wasn’t the whole story. In fact, applying the new federal standard for COVID nursing home deaths had cut the number from 9,018 on April 14 to 6,722. Then, the state took an extra step: It asked all licensed nursing homes and rest homes to review deaths among their residents with the more narrow standard in mind. The state relied on nursing home administrators, who had more firsthand knowledge of each case, to winnow the list down to 5,502 deaths.
Balser said she was puzzled by the change, recalling the frantic families last spring who were receiving only murky information about the hundreds of frail seniors dying in nursing homes. Public outrage prompted Balser to sponsor legislation that now requires more detailed reporting on the state’s website about such COVID-related deaths.
But the new system, Balser said, is confusing — especially after the state health department for so many months had highlighted the much larger nursing home death numbers on its daily dashboard.
“We owe it to the people who died and to their families [to get answers], and that includes accurate numbers and clear definitions,” Balser said.
Other states in recent weeks, including Ohio, have updated their data-collection and reporting systems, creating confusion about their nursing home deaths, too. The changes have sparked questions about how to measure the effectiveness of specific policies during the pandemic and for future use, such as locking out visitors to nursing homes for months to protect vulnerable seniors against infections, but exposing them to greater isolation.
In an updated briefing with the Globe this week, state health officials said the new system brings more clarity about COVID’s deadly reach into nursing homes.
But Anthony said the state’s approach is confusing because the state has not changed the way it counts overall COVID deaths. It still uses the more expansive method to calculate total deaths, which, as of Thursday, stood at 17,168. That, she said, means the public can’t get an accurate picture of how the number of long-term-care deaths compares to overall COVID deaths.
Before the change, it was widely reported that nursing home COVID deaths accounted for roughly half of all COVID deaths in Massachusetts. But a similar comparison can’t be made anymore because the state now uses two different methods to count total deaths versus the ones in nursing homes.
Officials at Kaiser Family Foundation, which runs a popular website that compares death rates by state, say that their efforts have been hampered by widespread inconsistencies in how states collect and report data.
“We don’t have a clear idea of what [state] policies worked and what policies didn’t, because the data has been such a mess,” said Priya Chidambaram, a senior policy analyst at the foundation.
David Grabowski, a health policy researcher at Harvard Medical School, said that when the numbers are eventually sorted out, Massachusetts will still rank near the top when measuring the rate of nursing home deaths because “things were so out of control here early on.”
More broadly, he noted that the newer CDC reporting system has its own weaknesses since it relies on nursing homes to accurately report deaths, amid continuing ambiguity about COVID’s lasting effects.
“How do we know if a resident has a stroke, was that stroke related to COVID?” Grabowski said. “It’s not clear to me at all that facilities are following it through.”
Anthony, with the Pioneer Institute, said getting accurate and consistent numbers nationwide is important, but the record also needs to reflect the human tragedy behind the numbers.
Massachusetts was “late to the game in terms of securing the safety of nursing home residents, and it’s still a large percentage of nursing home residents who did die — even using the new definition,” Anthony said. “I don’t want to ever lose sight of that.”
Rob Weisman of the Globe staff contributed to this report.
Kay Lazar can be reached at firstname.lastname@example.org Follow her on Twitter @GlobeKayLazar.