Pandemic meets ‘infodemic’: Is there too much data out there?
Covid info is everywhere, but it takes real skill to interpret it
An Iowa surgeon who does not, for the most part, treat Covid-19 patients, recently used his first Facebook post ever to say he leaves his lawn care to a landscaper and his car troubles to a mechanic. He’s taking the same approach with Covid-19 and asked that more of us do the same.
“Biostatistics is not easy. True research (not just looking through Google and social media) and reviewing data and studies and articles is not easy,” Dr. Jeffrey Dietzenbach wrote. The “expert” at your office, in your social media feed, and even at your dinner table, he said, probably isn’t one.
We’re living in an “infodemic,” and we’ve been told to use data to make important safety decisions, such as when to mandate masks at school or return to work, or gather in large groups. But most of us are navigating by guesswork.
News sites update case counts and death tallies and school outbreaks daily. State public health departments, including New Hampshire’s, have created dashboards with vaccination rates, hospital capacity, community transmission levels, equity gaps, and even the infection rate of health care workers. Lawmakers heard dueling testimony from medical providers this session, some reciting debunked claims that the vaccine kills, others citing evidence that it is the best protection against illness and death. Until recently, Gov. Chris Sununu, state epidemiologist Dr. Ben Chan, and the state’s other public health experts held weekly televised Covid-19 updates.
In short, there’s no shortage of easy-to-access data on the coronavirus or the state’s progress containing it. But is there too much?
“It’s not all bad. But a lot of it is bad,” said Dr. Eric Toner, senior scholar at Johns Hopkins Center for Health Security. “Social media, and the ability for information to be dispersed globally in seconds, can be a good thing for disseminating accurate, important information. But it also gives people who have either incomplete understanding, misunderstanding or intentionally false information the opportunity to spread that information equally fast.”
Other public health experts said the most useful information comes from hospitalization rates, hospital capacity, the number of new cases, and vaccination rates – if watched over time. Trends up or down rather than single-day snapshots are best.
Still, there’s “squishy” data that appears more straightforward than it is, Toner said. This includes even vaccination rates.
The state Department of Health and Human Services did not respond in time for this story.
Dr. Ali Mokdad, a professor at the Institute of Health Metrics and Evaluation, has calculated New Hampshire’s “immunity level” to forecast that we’ll see a surge in cases later this month.
To determine immunity, Mokdad added the state’s vaccination rate (54 percent are fully vaccinated) and its total cases (15 percent of the state has had a positive test), arguing both groups are largely immune to infection. That leaves about 30 percent of the population susceptible to Covid-19, he said.
Even fewer are susceptible under Mokdad’s additional belief that positive test results capture only 45 percent of the state’s Covid-19 cases because not everyone who has had the coronavirus has been tested.
Toner agrees that immunity extends to more than only the vaccinated and that positive tests are a fraction of Covid-19 infections. But he’s not as certain about the 45 percent calculation. “We’re fairly confident that there are many more cases than can have been confirmed, but we just can’t measure it precisely,” he said.
What we can infer with confidence, he said, is that immunity – and thereby success in containing the virus – is best achieved through vaccination.
“The large majority of the population has some level of immunities,” he said. “What we’re seeing right now is a combination of the people who are unvaccinated without previous infection who are now all getting very rapidly sick or infected, and a much smaller group of people who are having breakthrough infections.”
A number of news outlets and advocacy groups have promoted vaccination by drawing a connection between communities’ vaccination rates and their case counts. That sounds logical and, in some communities, it appears true. But it depends on how the rates are calculated.
According to the state’s dashboard, Deering has a low vaccination rate of 39.7 percent and it has had 69 cases, which means 3.5 percent of its population has tested positive. Exeter’s 79.9 percent vaccination rate is among the highest in the state, but its 1,010 cases means that 6.7 percent of its population has tested positive.
Adjust for population sizes and the results differ. Do you count all cases or only those in the last 14 days? All methods have been used to argue a correlation between vaccination rates and high case counts.
The problem, said Toner, is that while those comparisons work at the national and even county levels, the differences between local-level populations (size, age, and rural versus urban) are significant enough to make definitive conclusions unreliable. And, if you calculate the immunity rate, as Mokdad suggests, the results are different yet again.
The message holds – vaccination boosts protection against a surge – but the path there isn’t reliable enough to compare one community to another.
“I think (sharing of misinformation) is being done largely by people who are trying to be helpful,” Toner said. “They don’t really understand what they’re looking at, and they see something they think everybody else is missing and say, ‘I’m going to share this.’”
Experts, not ‘experts’
Toner and Mokdad said hospitalization rates are a good indicator of community transmission levels, but a lag in reporting (not uncommon, Toner said) can misrepresent the daily rate. And, hospitalizations are a late indication of spread because they are usually one to two weeks behind a surge.
New Hampshire, like most states, reports hospital capacity, a metric so important that Sununu, public health officials, and emergency responders traveled to Kentucky this week to see how that state is handling a surge in cases. It’s called in the National Guard and erected overflow tents.
Mokdad said he would not expect a surge like Kentucky’s because of New Hampshire’s immunity level and its hospital capacity. “Yes, there’ll be an increase in demand, simply because cases are going up,” he said. “But you are not one of the states we feel will have a major issue when it comes to hospital use.”
On Thursday, 18.7 percent of all hospital beds and 18.9 percent of ICU beds in New Hampshire were staffed and available, according to the dashboard. Eighty percent of the state’s ventilator supply was available. Toner said those numbers should not be cause for alarm.
Even that capacity data can be less clear than it appears. The bed capacity reflects a hospital’s licensed number of beds, Toner said, but not the additional beds it could accommodate if there was a surge. And beds alone are not enough. “We have talked about this in terms of space, staff, and stuff,” said Toner. A shortage of health care workers can impact the number of beds available today and a hospital’s ability to increase its capacity.
Additionally, those statewide percentages alone do not reveal that capacity is much higher in the northern part of the state (27 percent of ICU beds and 55 percent of all beds are staffed and available) and lower in hospitals along Interstate 93, from Nashua to the Lakes Region. Those regional numbers, available on the state dashboard by hovering over the map under the “hospital tab,” are the closest indicator of what hospital capacity is from community to community.
With so much data that can be interpreted in so many ways, what do experts recommend? Listen to the experts not the “experts.”
Get vaccinated and wear a mask, Toner said, adding that masks must be mandated because making them optional is ineffective. Mokdad put it this way: “We need to behave.”