BR.testrate.adv. 0116 bf.JPG

With clinic registration Natasha Anthony, right, takes information from people waiting outside for their turn as Ochsner starts community wide testing at Blackwater United Methodist Church located at 10000 Blackwater Road because Baker is considered a hot spot Thursday May 28, 2020, in Baker, La.

After conducting a study in New Orleans that found many people were unwittingly spreading the novel coronavirus even though they did not show symptoms, Ochsner Health is conducting a similar study in the Baton Rouge area.

The goal of this 'prevalence study' is to provide a clearer picture how far the virus has spread in Louisiana's two largest cities, which have seen the highest number of confirmed cases and deaths in the state.

Dr. Leonardo Seoane, Ochsner's chief academic officer and a senior vice president, said testing could start in late June or after the Fourth of July break. Ochsner is working with the Baton Rouge Area Foundation and tying down grant funding.

"We're working on it right now, the logistics, and I would say within the next month we will do a prevalence study in Baton Rouge and the greater Baton Rouge area, taking into account about 800,000 people," he said.

Seoane said the study will test 2,500 to 2,600 people in the region over a one-week period. That group will be based on a demographically and geographically representative sample so the results can be generalized to the broader population.

Ochsner's testing efforts in the Baton Rouge area informed creation of the sample for the prevalence study, Seoane said, revealing places with heavy viral spread or with limited testing.

Residents will be asked to take both an antibody test and a polymerase chain reaction, or PCR, diagnostic test, showing who has gotten over the virus and who has recently been infected. Pairing the two tests together is an important way of revealing how many asymptomatic spreaders of the virus are in the community.

Seoane said the antibody test is 99.6% accurate.

The Baton Rouge study will be similar to one Ochsner conducted in Orleans and Jefferson parishes with 2,500 participants. Though the study's methods and final results won't be disclosed until they are checked by outside scientists in a peer review, hospital officials say that study showed a "significant amount" of people unwittingly spread the virus.

Seaone said the findings of the studies could have important public health implications. They could, for example, reinforce the call for social distancing measures and protective face coverings — even for those who aren't sick — and they could give crucial information about where and how the virus is spreading.

Seoane said a recent study not conducted by Ochsner found that about 30% of people who tested positive for the virus got it from somebody who did not show symptoms. 

"It's one of the major ways the virus is spreading and understanding that in your community, I think, is key," he said.

The same premise is possible for the Baton Rouge area.

Many have noted that new positive cases likely greatly undercount the true spread of the virus because of the number of people who don't have symptoms and aren't tested. Based on data from March, the U.S. Centers for Disease Control and Prevention estimated that 35% of people who were infected were asymptomatic in the United States.

The planned prevalence study in the Baton Rouge area give a clearer picture of which specific areas the virus has hit hardest.

An earlier Ochsner analysis of 3,481 people who tested positive for coronavirus at its facilities lent support to what state and federal health officials have been saying: Coronavirus is having a disproportionate effect on the African American community.

That study found 70.4% of those who tested positive were black, though the hospital system's patient base is only 31% African American.

Among a number of health and demographic factors, including being African American, that study also found that living in a low-income area was associated with increased odds of being admitted into one of Ochsner's hospitals for the COVID-19 illness. 

Seoane said the prevalence study will be able to break down its information to the ZIP code level. The same  level of detail in the New Orleans-area prevalence study shows the virus is spreading out from epicenters.

"We absolutely proved, and I mean I think everybody intuitively knows too, it's not spreading just homogeneous throughout the entire state, much less the city of New Orleans, Jefferson Parish or Baton Rouge. There are certain neighborhoods where you see the spread and it'll kind of grow from those areas," Seoane said.

This kind of information can drive more focused testing and other efforts to help those hot spots, he said.

Meghan Parrish, spokeswoman for Baton Rouge General, said that hospital has talked about doing a prevalence study but hasn't yet. The hospital's antibody testing so far has shown that of about 2,000 people, 4.4% were positive, though that testing wasn't conducted randomly and can't be extrapolated to the entire community.

In a separate random sample constituting 5% of the hospital's employees — about 160 people — positive antibodies were found in 4.8% of those tested, she added.   

Other hospital systems in the Baton Rouge area are still in a conservation mode with their testing supplies, focusing on patients most in need. Daryl Cetnar, spokesman for Ochsner, said the hospital system has been able to leverage its purchasing power to source the necessary swabs and other supplies.

Mark Armstrong, city-parish spokesman, said the parish officials are excited about any new tool to fight the spread of the virus. 

Others, however, said they wanted to see more about how Ochsner was conducting its prevalence studies before they could speak to how representative the testing will be.

Susan Hassig, a Tulane University epidemiologist, said prevalence studies can be difficult and costly to construct to ensure a random and representative sample is collected.

She noted that some of the ways that Ochsner was soliciting participants for the New Orleans-area prevalence study was through television news stories and social media, both potential forms of self-selection that could undercut the randomness of any sample.

Random selection is a key factor in any study that seeks to generalize a smaller sample of people to the broader population. Public opinion polls work with the same statistical principles. 

"It really depends on how they actually implement it,” she said.

Seoane didn't provide all the details of Ochsner's methodology but did say that the hospital system plans to have multiple sites so they are close to residents and avoid bias in the sample. 


Email David J. Mitchell at dmitchell@theadvocate.com

Follow David J. Mitchell on Twitter, @NewsieDave.