Update: After publication of this article, the Louisiana Department of Health removed the asthma-related emergency department data for 2010 through 2015 from its portal. LDH also amended the note on its website, referred to in the article, which previously stated the agency was, for some asthma indicators, displaying only “smoothed” data on its portal. The new note states the asthma-related data is temporarily unavailable.
The Louisiana Department of Health has publicly misrepresented the true rates of asthma-related emergency room visits in places like St. John the Baptist Parish, home to the controversial Denka manufacturing plant, according to Vickie Boothe, an environmental engineer and epidemiologist who spent a combined 33 years in the U.S. Environmental Protection Agency and the Centers for Disease Control and Prevention.
By presenting “smoothed data” without qualification, LDH essentially halved the actual number of asthma-related emergency department (ED) visits in any given year in its dataset, said Boothe, who’s now retired but works on a volunteer basis with the environmental groups 350 New Orleans and Climate Reality.
Smoothed data borrows from surrounding areas outside of a given geographical area, in this case, neighboring parishes. It can therefore be more workable in certain situations, Alyson Neel, communications director for LDH, told The Lens. The CDC smooths, based on its modeling, state-level data that health departments provide.
Since at least August 2021, LDH unintentionally displayed the wrong dataset regarding rates of asthma-related ED visits in the state, and is unable to revert to a different dataset due to a technical defect, Neel said Friday.
“Last year, the smoothed data was inadvertently uploaded,” Neel said. “Unfortunately, a technical issue is preventing any updates to the portal. We continue to work to resolve it.”
”While both the smoothed and un-smoothed data sets provided by CDC have value (e.g., the un-smoothed data set allows you to look at differences by parish, and the smoothed data set allows you to look at other types of breakdowns like differences in gender and race/ethnicity), we normally post the un-smoothed data set because we do think it can be more intuitive for some,” she wrote.
The agency’s website displays asthma-related ED visits for the years 2010 through 2015, when the agency was a grantee of the CDC’s National Environmental Public Health Tracking (EPHT) Network.
In response to questions submitted by The Lens, LDH updated its website on Friday to clarify that it is presenting smoothed data in certain circumstances. The agency also directs the public to the CDC’s website, which provides both smoothed and unsmoothed data.
Before August 2021, the Louisiana Health Data Portal showed – using the raw, or unsmoothed data – that the ED visits in St. John Parish from 2010 through 2014 were approximately twice the state’s average, clocking in at an average of nearly 98 annual visits per a population of 10,000, compared to the statewide average of nearly 57 visits, according to a report Boothe produced, which she shared with The Lens.
But with the smoothed dataset that LDH has displayed since last year, the average number of visits in those years dropped to 55. Boothe, who worked at the EPHT team that funded state health departments to submit asthma data when Louisiana was a full-fledged grantee, said the state agency’s current data presentation effectively misleads the public.
“We were very, very clear that you should never put [the smoothed] data in a chart, you should never put that data in a table,” she said. “And you should never suggest that that would be the rates for any particular county or in Louisiana’s case, parish, because in the less populated areas, you get a very wide statistical error.”
Had LDH displayed the smoothed data on its website during Boothe’s tenure at the CDC, she, along with her supervisor, “would have told them that, ‘You cannot do this – you cannot have the tables and the charts and suggest that those smooth rates represent the rate for any particular Louisiana parish,’” she said.
The CDC did not respond to a request for comment in time for this article’s publication.
But on a podcast produced by the CDC, Heather Strosnider, epidemiologist and the current section chief of the EPHT tracking section, said that “the smoothed view is an average over many counties. It shouldn’t be interpreted as the rate for one particular county,” (italics added in the CDC’s written transcript). The “smoothed view” is meant to bolster datasets from small geographic areas, like counties, she said.
It’s “like an average for a geographic area,” she said. “The results come from combining data from one county with data from neighboring counties; then the Network calculates an average to allow a pattern to emerge within the area.”
Boothe recalls participating in discussions during her time at the CDC, wherein she and her colleagues debated the merits of offering smoothed datasets at all to states, she told The Lens. They ultimately decided to do so for the benefit of rural counties in states like Maine and Wisconsin, she said.
But it appears that Louisiana is alone among other states with rural counties that have enjoyed EPHT-grantee status like New York, Maine and Wisconsin in displaying smoothed data at the county, or parish, level for asthma-related ED visits.
According to emails provided, in part, by LDH, when asked by Adrienne Katner, associate professor of environmental and occupational health sciences at LSU Health New Orleans School of Public Health, why there was a discrepancy on the website during August 2021, Kate Friedman, environmental health scientist, stated the agency neglected to mention in the “Info Tab/Metadata” that it was using smoothed data.
“We apologize for the oversight and are working to correct it now,” she said. “We are having some discussions here on whether to keep the smoothed data or revert to the original. I prefer the original data but due to COVID surge, we don’t have the resources in BHI to assist us to stratify the measures which are required by the grant,” she said, referring to the Bureau of Health Informatics, which serves as the “hub for many Office of Public Health-wide data and informatics efforts,” according to the bureau’s website.*
Friedman did not mention a technical issue that would have prevented the agency from updating its publicly facing data. She did provide an attachment that included the unsmoothed dataset.
Katner was not interviewed, consulted, nor the source of any information for this story.
Title VI Investigation, other research
Meanwhile, the EPA has found evidence that LDH and the Louisiana Department of Environmental Quality have left Black residents living near the Denka facility in St. John the Baptist Parish, along with those living throughout the industrial corridor, disproportionately vulnerable to harmful pollutants, according to a letter the federal agency sent last month.
Specifically, the EPA’s initial investigation shows that LDH may have failed to provide those residents with timely, critical information about the cancer risks associated with living near the Denka facility, given the exposure to chloroprene, Lilian Dorka, deputy assistant administrator for External Civil Rights at the EPA said in the letter addressed to LDEQ and LDH.
The EPA’s recommended maximum annual average air standard for chloroprene is set at 0.2 micrograms per cubic meter of air – which amounts to a 100-in-1-million cancer risk. But chloroprene levels, according to EPA data, remained as high as 23.677 micrograms per cubic meter of air in St. John the Baptist Parish as of September 2021, the legal nonprofit Earthjustice said in its complaint in January on behalf of the groups Concerned Citizens of St. John and Sierra Club.
Denka, for its part, previously pointed The Lens to the investments the company has made in reducing emissions, yet at the same time, said there’s no evidence its emissions are causing adverse health issues.
But a study conducted by the University Network for Human Rights found cancer rates to be significantly higher for the residents living near the Denka facility than what would be likely, after controlling for the factors of age, race and sex. The study also found that residents’ cancer rates were positively correlated with their proximity to the Denka facility.
Those findings may be relevant to concerns regarding the use of smoothed datasets in general, Kim Terrell, research scientist and director of community engagement at the Tulane Environmental Law Clinic (TELC), told The Lens. That’s because the hyper-local differences the University Network researchers identified demonstrate the hazards of averaging data from various geographical areas.
“We know from research that air quality varies across a pretty small scale,” she said. “And air quality is one of the main drivers of asthma, or maybe the main driver.” One study showed that air quality standards varied block-by-block in parts of California, for instance.
Terrell also helped produce a study at TELC, first reported by The Times-Picayune, that demonstrates Black residents in the state’s industrial corridor between Baton Rouge and New Orleans, dubbed “Cancer Alley” by some environmental groups and community members, are disproportionately exposed to harmful air pollutants, compared to their white counterparts. The study, which has been submitted for peer review, may help bolster the EPA’s Title VI investigations, Terrell said.
“For over 60 years, St. James Parish has concentrated heavy industry in the same two Black neighborhoods, one on each side of the Mississippi River,” Gail Leboeuf, St. James resident and co-founder of Inclusive Louisiana, said regarding the Tulane study.
While asthma is, of course, a different illness from others that might be top-of-mind for people concerned about the welfare of communities in the state’s industrial corridor, its effects shouldn’t be underestimated, Boothe told The Lens. The seriousness of the disease underscores the need to properly report its prevalence, she said.
“I know people don’t think of asthma as a deadly disease, but hospitalization is very serious,” she said, adding that children die from it in extreme situations.
Some language in this story was updated for clarity.
*Correction: A previous version of this story inaccurately referred to “BHI” as “behavioral health integration.” That error has been corrected.