The Environmental Protection Agency (EPA) plans to limit the scientific studies it uses to create regulations so that only reports with publicly available data are considered. The decision has drawn major ire over claims that opening research to the public risks breaching private health information, and that the policy could disqualify well-conducted studies from consideration if an institution refuses to make its data available.
But the EPA’s shift on transparency has the potential to be far more beneficial than critics make it out to be.
There’s an undisputed crisis of irreproducibility facing research today. Science, one of the world’s top academic journals, suggests, “more than half of researchers struggle to reproduce not only the results of their fellow scientists but their own experimental data.” The EPA policy amounts to giving independent researchers the opportunity to check one another’s work, and it’s a safeguard worth pursuing: If half of what we “know” about how chemicals and molecules operate in the body isn’t true, are current environmental or public health policies actually keeping us safe?
Consider the tennis match of nutrition science. One day, researchers announce that an evening glass of wine is healthy and increases longevity, while the next day others claim wine causes cancer. The flip-flops occur equally in other scientific fields, but the public generally loses interest once academics mention dose-dependent responses and molecular pathways. It’s uncertainty in these complex studies that cast doubt on the quality of research used not just by the EPA, but by state and federal agencies nationwide.
In 2013, the Obama administration tackled the issue by directing federal agencies to increase public access to data from federally funded studies.
In response, the National Institutes of Health — a federal agency that deals almost exclusively with sensitive personal health information — upheld public access to data as the gold standard for future research. NIH doesn’t disclose the identities of patients involved in HIV medication trials or cancer surveys, for example, but the agency has developed methods to anonymize data, strictly control access to potentially sensitive information, and require researchers to sign binding confidentiality agreements. The EPA’s latest move is part of its own ongoing attempt to fulfill President Obama’s transparency request.
Those who worry about the policy barring high-quality studies from consideration should consider this: If one or two studies form the basis of a major policy, perhaps it’s worth verifying that their conclusions are accurate.
When it comes to investigating potentially dangerous chemicals, the EPA has long relied on “linear models” that assume danger increases and decreases with the amount of a chemical present. Without independently running the numbers, scientists can’t know whether a j-shaped model (risk is high at low doses, but drops before rising again) or a threshold model (no harm is expected below a particular dose) would fit the data better. That’s one of the current arguments underpinning recent debate over the safety of BPA. Without verifying that the most accurate analyses were used to reach a conclusion, the EPA risks regulating on bad information.
And where a single study indicates that a chemical or pollutant is harming public health, proper science demands that the trend be investigated further to determine how and why it’s occurring, as well as whether the trend is actually occurring at all. Science is a replicative process, and that’s a very good thing.
The EPA policy may not be perfect, but it’s an aggressive step to address a problem plaguing the scientific credibility of every research institution in the world. For what it’s worth, data transparency is already becoming the norm. The federal government funds almost a quarter of all medical and health research in the U.S., and federal dollars mean public data. Even academic journal publishers like Elsevier, PLOS, and Springer Nature have longstanding policies requiring researchers to deposit pre-analyzed data into public data repositories.
It’s time we stand the old adage on its head. If it’s broken, someone’s got to fix it.
Dr. Joseph Perrone is the Chief Science Officer at the Center for Accountability in Science.
The views and opinions expressed in this commentary are those of the author and do not reflect the official position of The Daily Caller.