Ep. 76: Strengthening Transparency or Silencing Science?

Word leaked a few weeks ago that the EPA is poised to finalize a rule to limit the types of scientific studies that can be used to create new regulations. The proposal -- named “Strengthening Transparency in Regulatory Science” -- would require public health researchers to release their raw scientific data in order for their work to be considered when the EPA sets regulations. These regulations dictate things like how much pollution companies can release into the air and water.

The rationale for the rule goes like this: these studies are “not transparent,” and that means researchers currently can’t release their raw data because it contains confidential information like personal health data, names and home addresses.

The proposed rule would only allow the use of studies that make all data publicly available for anyone to analyze. Former EPA Administrator Scott Pruitt first pitched the new rule back in 2018 as a way to make the agency's decision-making more, "transparent, objective and measurable.”

Dr. Mary Rice was one of five scientists to testify before the House Science Committee on the “Strengthening Transparency in Regulatory Science” proposed rule. Photo: C-SPAN.

When the rule was first proposed, the agency received nearly 600,000 comments, the vast majority of them in opposition. One of those came from Dr. Mary Rice. She’s a pulmonary and critical care physician at Beth Israel Deaconess Medical Center, where she treats patients with lung disease. She also studies the effects of air pollution on lung health and is an assistant professor of medicine at Harvard Medical School.

On November 14th, Dr. Rice was one of 5 scientists to testify before the House Science Committee on the proposed rule. She spoke on behalf of the American Thoracic Society (ATS), a 16,000-member medical professional organization of physicians, researchers, nurses, respiratory therapists and allied health professionals “dedicated to the prevention, detection, treatment, cure and research of respiratory disease and critical illness.”

Here is an excerpt of Julie Grant’s interview with Dr. Rice. You can listen to the full conversation on our podcast:

Julie Grant: Broadly speaking, if the rule that we're talking about went into effect, how would your work and research be affected?

Dr. Mary Rice: Well, this rule does not really interfere with how the research is conducted, what it interferes with is how the EPA uses published studies to make decisions about health standards after the research has been completed. So to give you an example of my own research, I have a study of patients with chronic obstructive pulmonary disease or COPD. And in that study, the study participants carry air quality monitors with G.P.S. trackers around with them every day to identify where they're located at any given time. And in order to add that piece to my study, I had to take extreme measures to make sure that nobody who ever got that code could figure out who the study participants were.

So to comply with this rule, I would have to make all the data of my study participants available in some kind of public database, and that would include their medical histories, the fact that they have COPD, where they're located on a given day from the G.P.S. data. And in the consent form that they signed, I guaranteed that I would protect their private data. So the end result would be that that particular study would be one that the EPA would not consider and that would really be the case for most research that involves real people, living in the real world, going about their daily lives.

Julie Grant: Why is it important that certain data be private, and how does it impact things if a study like that is not usable by EPA?

Dr. Mary Rice: Well, that's the part that I'm most concerned about. The research we're talking about is research that involves real people, real children and adults and how pollution affects their health. This rule is a process barrier that would block that kind of research from informing EPA policies. And these are the kinds of studies that have shown things like air pollution is a cause of premature death in older adults; that air pollution is linked to worse lung function in kids and slower lung function growth as they get older.

A well designed study is one that collects a lot of information about people that could confound the association between pollution and health -- things like habits, like whether or not you smoke, whether or not you drink a lot of alcohol. But that's sensitive information that no one participating in those studies would want to have shared. You wouldn't want your mental health history to be shared with the public or your smoking history, for example. And so we have very good processes in place to protect that kind of information. And this transparency rule, the way it is currently designed, prevents the EPA from using those very well-designed studies. And that's what worries me most.

Julie Grant: The EPA has said in response to criticism of the proposed rule, "Science transparency does not weaken science. Quite the contrary. By requiring transparency, scientists will be required to publish hypothesis and experimental data for other scientists to review and discuss, requiring the science to withstand skepticism and peer review.” So what's your response to that? The EPA is calling this an increase in transparency, but seems like you see it differently.

Dr. Mary Rice: Yes. The key issue is that this rule creates a barrier against the use of science that can't meet the EPA requirement. The American Thoracic Society, and the scientific community broadly, very much supports mechanisms that promote data sharing among scientists while protecting the privacy of study participants. You know, when doctors are taking care of patients, they are considering all different kinds of studies. Some are well-designed. Some are just case reports.

And doctors have to make medical decisions based on best available evidence that they have at hand. And over time, as more evidence emerges, medical practices change. When I take care of patients, I do my best to look at all the available evidence and to make my medical decisions based on the best available evidence that I have at hand to make decisions about what drug to give or how to manage a particular patient in the ICU.

And really the same applies to the EPA and the decisions that it makes about our public health. We would want our EPA to use all of the available evidence in making decisions about health standards and not creating a rule that limits the use of certain studies. And that's the piece that the scientific community is objecting to.

Julie Grant: You invoked the tobacco industry in your testimony. And you included in your testimony a letter from the tobacco industry from the 1990s. Could you explain the connection there and why you wanted to make that connection?

Dr. Mary Rice: Yes. In fact, what EPA is proposing comes straight out of the playbook of the tobacco industry and its attempts to discredit research findings that link environmental tobacco smoke to health problems. I shared an internal memo that was written in 1996 by tobacco lawyer Chris Horner. And in that memo, he laid out a detailed strategy for discrediting scientific findings. And he advised R.J. Reynolds to focus on, "process as opposed to scientific substance," because attacking the substance of the science (that secondhand tobacco smoke is bad for health) would be "a public relations nightmare."

If you review the memo, it's really uncanny — the same words are used: transparency, use of sound science, replication of science. I think it's pretty clear that this proposed rule is not about improving transparency, it's about creating a process barrier to discredit or ignore science about potentially harmful health effects of exposures that are found to be inconvenient.

Julie Grant: So do you know any scientists or epidemiologists who are in favor of this proposed rule?

Dr. Mary Rice: I do not. In fact, what I thought was interesting at the hearing was that none of the witnesses were in favor. They were all scientists. And the minority witness was Brian Nosek, the director of the Center for Open Science. He was also not in favor of the rule. All of the scientists on the panel were supportive of mechanisms to improve transparency.

Julie Grant: So who is in favor of this rule? You've you've talked about what's at stake, but who's to benefit here?

Dr. Mary Rice: Well, I think it's very clear that this proposal is not favored by scientists, but it is supported by certain industry groups. And some of them do pollute the environment and have a stake in the outcome of environmental regulations. And one concern that the American Thoracic Society has is that at its worst, this proposal could deliver sensitive health research data straight to the potentially misleading manipulations of special interest groups. And those special interest groups and industry groups would then be free to report their alternative findings without having to undergo peer review like the scientific community does when they publish your findings. And they would have access to these data without having to demonstrate that they have the skills and appropriate resources to analyze the data and interpret it in an unbiased manner. And all of these findings would be on the same playing field in terms of informing EPA policy. And that's a frightening thought.

Julie Grant: Now, when the rule was first proposed in 2018 by then Administrator Scott Pruitt, he talked about a replication crisis, referring to scientific studies that can't be reproduced. And then in September, the current head of EPA, Andrew Wheeler, said this to a congressional committee: "Good science is science that can be replicated and independently validated, science that holds up to scrutiny. That is why we're moving forward to ensure that the science supported agency decisions is transparent and available for evaluation by the public and stakeholders."What are your thoughts on this idea that good science requires the ability to reproduce studies?

Dr. Mary Rice: Well, sometimes studies are not replicated when they're repeated in a different setting by a different scientific team. In fact, that happens all the time in medical care. Sometimes our decisions are swayed very quickly by a single drug trial that has a very different outcome from what has been previously found. But in EPA policy, historically, decisions have not been made based on one or two studies. We're talking about, in many cases, hundreds of studies that inform EPA decision making. That is replication. So when you see the same findings over and over, by different scientific teams, looking in different populations and, in the vast majority of cases, the same conclusion is reached -- I would not call that a crisis. I would say that's consistent scientific evidence.

One example of where the scientific evidence has been particularly consistent is the link between particulate matter exposure and mortality. In this most recent policy assessment, for example, the EPA looked at the link between mortality and particulate matter exposure, and they relied on 21 U.S. and Canadian studies. So I think replication is very important. And historically, the EPA has made policies based on data that is very well replicated. When the same findings are seen over and over again in different studies, using different tools in different populations, that is a form of replication.

>>Listen to the entire episode below or wherever you get your podcasts.

Julie Grant: The New York Times reported that the new rule would require scientists to disclose all their raw data, including confidential medical records. But the EPA said in response that that's not accurate. The press release from the agency said the EPA "maintains protecting confidential personal information, just as other federal health agencies regularly do. The reporter clearly does not understand the terms in the context of science transparency." So can you help us figure out what's going on here?

Dr. Mary Rice: So part of this gets at what is really the meaning of "underlying data." In my field (air pollution research), the underlying data about environmental exposure, it all really depends on the location. Where does a person live or where does the person work or go to school? And that location is used to assign exposure to pollution. Dates are also really important. So, for example, in a study that's looking at mortality or death, we're looking at the exposure in that location on the dates leading up to that death when that person was at that location are important because that determines what the exposure was. So you need dates and you need location to do an environmental health study looking at outdoor pollution at least.

So to actually make the "underlying data" public in a way that could be verified and double checked, it would involve sharing information about dates and location. And that's protected information.

I want to make another point, which is that I believe that part of the EPA's argument there was that there are mechanisms to de-identify data. And the environmental health research communities are concerned also because, in our field, it may actually be especially easy to re-identify study participants because of this reliance on dates and location. Potentially, it would be awful for the future of the field because imagine how willing you might be to sign up for an environmental health study, or to enroll your child in an environmental health study, that's looking at pollutants in the air or the water if you knew that the EPA would take the data of that study and make it fully available to the public.

Julie Grant: We're trying to get a sense of how big of a deal this proposed rule is. On this podcast, every week we're talking about another regulation being rolled back or another rule being put forward or a new climate report with dire news. So does this rule stands out out as different from what's happened in other administrations? Is this something that you think is broader or bigger somehow?

Dr. Mary Rice: Yes, I think this rule is particularly scary because we're not talking about one pollutant or one policy or one health standard. This is a rule that would cut across the entire rule making process for all types of EPA policy. And I focused a lot in our discussion on respiratory health because I'm a pulmonologist. But there are so many health effects of pollution and we depend on the EPA to set standards that protect our health. Some of these health effects are things like childhood leukemias, adult cancers, effects on air pollution on the heart and the brain and endocrine system. And this is really far reaching.

And this proposed rule introduces a process barrier that would cut way back on the amount of research that the EPA would consider in its rule making. And once that bureaucratic machinery is put in place to block or obstruct the use of science in EPA policy, it might be really difficult to undo. I'm very concerned that the implications of such a change for our health and and the next generation could be extensive and even irreversible.