COVID-19 has flooded the headlines with numbers.
There is so much data out there about the virus and things related to it, but not all of it’s helpful — and not all of it makes sense.
Frankly, much of the data veers into the domain of obscurity and confuses us — the regular folk, not scientists — more than helps us understand the pandemic.
Data is critical, though, because it helps people take calculated risks and make decisions, and it helps policymakers craft response plans to the pandemic.
That’s also true in Hawaii, of course. State Epidemiologist Dr. Sarah Park says a variety of data informs the governor and legislators in important decision-making discussions, including when to reopen the state to visitors without such a stringent quarantine.
“Good information is the bed rock of any kind of effective response,” said Dr. Cyrus Shahpar, director of Prevent Epidemics, a research team of the international public health organization Resolve to Save Lives. “There is not another option. The other option is to have a lot of people sick and dying.”
But there’s a glaring problem in all of this that hinders our ability as a country to respond to the pandemic effectively, according to a new report by Prevent Epidemics, which is led by the former director of the Centers for Disease Control and Prevention, Tom Frieden.
That problem is rampant inconsistency in data across the country.
Prevent Epidemics studied the public data dashboards of all 50 states, Washington, D.C. and Puerto Rico and found that every state is basically doing its own thing in terms of what data it publishes, how it is presented and what platform is used. That kind of variance leads to data that is “inconsistent, incomplete and inaccessible in most locations,” the report says.
The report did not think Hawaii’s data was impressive at all. The review found that Hawaii was only publishing about 13% of the information researchers thought the state should be publishing.
The Hawaii Data Collaborative, a nonprofit data research organization here, writes in a review of the report that it shows Hawaii is not reporting some critical data points, including COVID-19-like illness trends and average testing turnaround times, which “represents a serious gap in the information we as a community have available to us to understand.”
To be fair, the report found that most states were doing pretty poorly, because there are no national standards for how these things should be done. The highest scoring state, Minnesota, got only a 43% grade.
“There’s a lack of federal guidance on what information we should be looking at,” said Shahpar, who formerly worked as a team lead for the CDC’s global rapid response team.
But clear standards are critical for effective virus control.
Park of Hawaii said something similar about a lack of consensus about what specific set of indicators should be looked at across the board: “That is a national discussion right now among state epidemiologists and subject matter experts. I don’t think there’s any one agreement right now.”
The pandemic is still evolving, she said. So the discussion evolves, too.
The Prevent Epidemics team identified a set of 15 essential indicators, or data points, that it says states should be publicly reporting on their COVID-19 data dashboards to help inform their pandemic response. That data includes case and death counts, screening rates, daily hospitalization rates and the percentage of beds occupied by suspected and confirmed COVID-19 patients.
Prevent Epidemics also calls for more specific breakdowns of the data, including a seven-day moving average for some indicators and racial and age breakdowns for others.
Shahpar says those kinds of granular details are necessary because as the pandemic continues, states can no longer resort to “blanket measures” such as shutting the whole state down, and have to come up with more targeted and precise responses.
“To do that, we need data,” he said.
Hawaii does not provide many of those details on the dashboard, according to the review and Shahpar. For example, it doesn’t provide data for shared living facilities such as nursing homes.
That data is extremely important because such facilities are where the bulk of COVID-19-related deaths in the U.S. are coming from, he said.
There’s nothing about what experts call “syndromic surveillance,” Shahpar said. That’s when the government observes people with COVID-19-like or influenza-like illnesses. They can be recorded from places like emergency departments and can be a leading signal in detecting the virus, according to the report.
The state also does not provide contact tracing data beyond differentiating cases by “travel,” “community” or “travel-associated contact” on its COVID-19 dashboard. But in a state like Hawaii where there is a relatively lower number of cases, Shahpar said there should be way more information available about contact tracing. Fewer people are easier to track.
“It’s not that difficult to pull this information together,” he said. “I’m sure they have it.”
Hawaii’s COVID-19 dashboard also does not break down the data much beyond age groups. Race breakdowns are available for cases, but not for deaths or hospitalizations. The Prevent Epidemics report also points out that many states — including Hawaii — report testing figures, but do not specify what sort of testing: antigen, antibody or polymerase chain reaction.
The Civil Beat Virus Tracker pulls local data from the Hawaii Department of Health. We get a lot of questions from readers about why we don’t provide detailed information about cases. That’s because more detailed data is not available through the state.
The data that Hawaii collects are “pretty standard,” said Park, who heads the state’s disease outbreak control division. Hawaii won’t publish anything that would violate someone’s privacy or jeopardize its work, she added.
“We cannot compromise the cases that we’re investigating,” she added.
As far as some of the breakdowns — or “disaggregation” as some of the data people call it — go, Park said with Hawaii’s relatively low number of cases, breaking the figures down into subgroups may cause privacy concerns.
“When the numbers are really small, they can be potentially identifying. So we’re really careful,” she said.
There’s also the matter of what happens to the data once it’s out there, she said. People misinterpret the numbers and make their own conclusions, which can cause resources to be diverted to areas that may need them less, so Park says they try to be mindful of what gets put out there.
“I don’t care how transparent you think we need to be on that,” she said.
Shahpar of Prevent Epidemics said he can understand the privacy argument. But with more than 1,600 cases, Hawaii can do more disaggregation — or breakdowns of data — than it currently does. If data is being collected with public dollars, then it should be used to inform the public, he added.
Still, Hawaii is in a good situation compared with the rest of the country in the grand scheme of the pandemic, he said. It has a relatively low number of cases, low rates of positive tests and low hospitalization rates.
“Hawaii has some natural things that allow it to be more successful,” he said. It’s been geographically blessed with its climate and location.
That isn’t to say Hawaii shouldn’t have better data collection and reporting, just because it’s doing relatively better in terms of case counts and hospitalizations, he said.
“I think this is the time to really bulk up your data,” he said. “Be a leader, rather than saying, ‘We’re in a good place and we’re fine.’”
Read the full report here:
Civil Beat readership has more than doubled in the past nine months. That’s incredible growth for which we’re so grateful.
But for a small nonprofit newsroom that provides free content with no paywall, readership growth alone can’t sustain our journalism. The truth is that less than 1% of our monthly readers are financial supporters.
To remain a viable business model for local news, we need a higher percentage of readers-turned-donors.
Will you consider becoming a new donor today?