What’s slowing Miami’s COVID spread? Partial ‘herd immunity’ may play a part

As a deadly summer wave of virus continues to recede, Miami-Dade County officials and scientists are trying to figure out what combination of factors may have contributed to slowing a surge of COVID cases that at one point threatened to topple South Florida’s healthcare infrastructure.

Social distancing measures, face mask orders and curfews certainly helped, public health experts say, but so did other factors that they’re still working to understand — specifically, the seasonality of the virus and so-called herd immunity, which occurs when enough people in an area are infected with a virus to nearly eliminate transmission.

In Miami-Dade, Florida’s hardest hit county, there have been more than 140,000 confirmed cases — more than twice the number of cases in the next highest county, Broward — and a number that is certainly an undercount, according to blood surveys by county and federal officials that estimate the true infection rate is at least five times higher.

If those estimates are correct, that would mean Miami-Dade’s total infection rate could range from 10% to as high as 30% of the county’s population. That’s still far from reaching the threshold where herd immunity could reduce transmission to zero. The threshold for the novel coronavirus is still being debated, but is generally thought to require at least a 60% infection rate.

Opponents of public health restrictions have argued that herd immunity alone can stop transmission in hard-hit places such as Miami and, earlier in the course of the pandemic, New York City.

But if the coronavirus were allowed to spread and reach full herd immunity in the United States, the “death toll would be enormous,” White House coronavirus advisor Dr. Anthony Fauci said this week.

Scientists emphasize that partial herd immunity only contributes to a decline and is not enough on its own to end the pandemic.

“What’s important to understand is that all of these things add up,” said Marc Lipsitch, a professor of epidemiology at Harvard’s T. H. Chan School of Public Health. “A modest amount of help from herd immunity combined with partially effective control measures, combined with the fact that it’s summer.”

But he also cautioned that the slow, downward slope of new cases in South Florida isn’t a reason to abandon any safeguards at all. The safeguards are one of the main reasons for the declines.

“Each of those is making a contribution to the trajectory in Miami and elsewhere. ... Herd immunity alone would not be enough to get us out of this pandemic at the current levels,” he said.

At Jackson Health System, Miami-Dade’s public hospital network, CEO Carlos Migoya told the Miami Herald this week that the downward trend in new cases and hospitalizations could be reflective of partial herd immunity but like Lipsitch, he emphasized that personal protective measures are a major piece of the equation.

“Maybe we have enough people in this community who have been there,” he said, “plus we’ve continued masking and distancing.”

What does a high infection rate mean?

In an effort to shed more light on what’s slowing transmission, some scientists have created simulation models to calculate the number of infections while others have conducted serology surveys using blood tests to help them figure out what percentage of people have antibodies to the virus, indicating they’ve been infected in the past.

Youyang Gu, an independent data scientist, uses artificial intelligence and infectious disease modeling to create projections that are cited by many public health experts. Gu’s model is cited by the Centers for Disease Control and Prevention in the agency’s forecasts for COVID deaths.

Gu’s simulation model estimates that 20% of Florida residents have already been infected. And in Miami-Dade, he said the model indicates that 30%, or about 829,000 people, have been infected — an estimate so high that it spurred a debate among influential epidemiologists on social media earlier this month.

But that doesn’t mean they’re discounting it outright. Bill Hanage, an epidemiologist at Harvard’s public health school, said while the estimate does look high given the total number of deaths in Miami-Dade — 1,999 as of Friday — many of those currently infected may have a better chance of survival than those initially infected.

“I suspect the model is also reflecting accurately a large number of infections in younger age groups who are least likely to suffer the most severe outcomes,” Hanage said. “I hope this continues to be the case and Florida is able to protect its retirement communities.”

Gu himself has acknowledged that his model’s estimate for Miami-Dade’s infection rate is alarming.

“It is pretty high. It’s one of the highest I’ve seen, but that is what the data is saying,” he said. “I can’t adjust it afterwards and say, ‘Yeah, it’s too high.’ ”

Under Gu’s projection as of Friday, Miami-Dade would suffer almost 1,000 more deaths by Nov. 1, for a total of 3,362. That’s more people than were killed in all the attacks in the U.S. during 9/11, when 2,977 people died.

Trevor Bedford, a scientist at the Fred Hutchinson Cancer Research center in Washington state, also cited Gu’s estimate for Florida — that about 20% of the state’s population has already been infected — when theorizing on social media that some level of herd immunity was slowing spread in the Sunshine State.

“I believe the substantial epidemics in Arizona, Florida and Texas will leave enough immunity to assist in keeping COVID-19 controlled,” Bedford wrote on Twitter. “However, this level of immunity is not compatible with a full return to societal behavior as existed before the pandemic.”

Natalie Dean, a University of Florida biostatistician, said that the public should view all modeling results with caution. The data underlying the models, Dean said, are inconsistently reported and can lead to confusion.

“In practice, we like to compare several models,” Dean said.

What scientists really need, she said, is widespread blood testing to look for evidence of virus antibodies — proof that a person was infected — and “on-the-ground data.”

What our blood taught us

Miami-Dade was one of the first places in the country where there was a coordinated effort to randomly sample people’s blood for evidence of past infections.

The study was a joint effort between county officials and the University of Miami, but testing halted in May after regulators issued new guidelines for antibody testing and researchers decided to stop using rapid finger-prick tests.

On Thursday, a county official and a lead researcher of the study both said the tests would not be resuming, saying it would be hard to determine who had been infected if their antibodies were months old and waning. UM added that the state is also now doing antibody tests at some state testing sites, in Miami-Dade and other counties.

When Miami-Dade and UM halted their work at the end of April — barely a month into the pandemic — they estimated that about 6.5% of the population had been infected, dwarfing the number of reported cases at that time.

Though scientists using different methods may disagree on the actual rate of infection in a community, most agree that population-based immunity plays some role in slowing transmission of the disease — when combined with other measures.

Harvard’s Lipsitch emphasized that only one factor is in our control: personal behavior. He said that when New York became the epicenter of COVID-19 in the United States, residents of neighboring states, including New Jersey and Connecticut, responded by staying home, social distancing and other behaviors that worked to reduce transmission of the disease.

Government interventions are not always needed, he said, because people will often act out of a sense of self-preservation.

“When a community is really hard hit as Miami and others in Florida have been,” Lipsitch said, “I do think it focuses the mind and leads to people taking the control measures more seriously even if the law doesn’t change.”