U.S. EPA has recently come under fire from its quasi-independent auditor, the Office of the Inspector General (OIG), which has issued multiple reports accusing EPA of not moving with sufficient speed or transparency to increase regulations of sources of EtO. In these reports, OIG pressed the issue by publishing EPA draft work files, over EPA’s objection, which identified many facilities potentially contributing to increased community cancer risks. But publishing the draft screening results for these facilities is likely to have unintended consequences including complicating information gathering efforts, and increasing private litigation.
OIG’s April 15, 2021 Report
On April 15, 2021, the OIG issued a report accusing the Trump administration of delaying communications about cancer risk to communities near industrial EtO emissions. The gist of OIG’s criticism was that EPA had conducted some short-term monitoring of EtO emissions at the Sterigenics facility in Willowbrook, Illinois, as part of the 2014 NATA, but when the monitoring suggested that the facility’s EtO emissions were contributing to EtO concentrations significantly above the IRIS EtO URE, EPA did not immediately publish the monitoring results and engage in community outreach, but instead waited until the release of the 2014 NATA many months later to publicly disclose the EPA modeling of the Sterigenics sterilizer.
EPA responded to this report by emphasizing that it had been hesitant to share draft findings publicly because “it is fundamental that we provide accurate information in communicating risks. As such, any preliminary air toxics risk information needs to be verified and quality assured prior to communicating with the public to avoid confusion and to build trust.” (See Appendix B of OIG’s April 15, 2021 report). EPA also noted that, EPA needed to “effectively understand the concerns of coregulators” when addressing air toxics issues.
EPA’s concerns about the effect of releasing the draft modeling results in 2018 were prescient. As soon as the modeling results for Sterigenics’ Willowbrook facility were released, local citizen groups started pressuring politicians, resulting in Illinois EPA actually ordering the facility to close on a temporary but indefinite basis, based solely on the IRIS EtO URE screening values, and without alleging any violation of the facility’s air permit. Plaintiffs’ attorneys quickly followed, with hundreds of lawsuits filed against the facility. In the face of this trifecta of community activists, regulators and plaintiff litigation, the Sterigenics facility soon closed on a permanent basis. The Illinois legislature then flirted with a bill that would have banned all medical sterilizers in the state, but settled for a law requiring some of the nation’s strictest controls in order to limit sterilizer emissions to levels near the IRIS EtO URE. The resulting state-level regulations have cost industrial EtO emitters that remained in Illinois approximately $10 million each in capital expenditures alone.
EPA’s concerns about the necessity of coordinating with coregulators proved well founded as well. In the aftermath of Illinois EPA’s closure of the Sterigenics Willowbrook facility, on Oct. 25, 2019, the FDA took the rare step of issuing a public statement warning that aggressive government actions that threatened further shutdowns of medical device sterilizers, even on a temporary basis, “could result in years of spot or nationwide shortages of critical medical devices, which could compromise patient care.” For instance, FDA noted that Illinois’ shutdown of Sterigenics’ Willowbrook facility had resulted in a temporary nationwide shortage of pediatric breathing tubes. FDA warned that future regulatory actions related to medical device sterilizers needed to holistically take into account not just the health risks of EtO, but also any public health risks associated with the closure of even a single sterilizer.
OIG’s May 6, 2021 Report
OIG was not persuaded by EPA’s concerns about avoiding potential public confusion or conflicts with coregulators. On May 6, 2021, OIG issued a report instructing EPA to immediately initiate new Residual Risk and Technology Reviews (“RTRs”) for all EtO-emitting-source categories and to tighten emission standards in light of the 2016 IRIS EtO URE. EPA responded by noting that it was already conducting RTRs and revising relevant NESHAPs in light of the IRIS EtO URE, but also noted that EPA has statutory discretion to propose revisions to NESHAPs separately from the periodic RTR process. EPA also noted some issues that might affect the timeline for issuing NESHAP revisions, including ongoing litigation EPA was engaged in with regard to some NESHAPs, and the information-gathering activities being pursued to allow EPA to best address the unique context of commercial sterilizers. But on the whole, EPA was in cautious agreement with OIG that relevant NESHAPs should be amended in the next couple of years, and that the proposed revisions would expressly be based on the IRIS EtO URE.
But OIG’s May 6 report went well beyond recommending that EPA revise emission standards where appropriate. Over EPA’s objection, OIG also published a list of 29 facilities that EPA had identified in internal analyses as potentially contributing to lifetime cancer risks greater than 100 for at least one census block, but not on a census tract level. This list was based on preliminary working files used by EPA to calculate census tract risks as part of the 2014 NATA, and was not previously published as part of the NATA. For context, census tracts are small subdivisions of a county that are designed by the U.S. Census Bureau to contain about 4,000 people and 1,600 housing units. A census block, on the other hand, is the smallest geographic unit for which the U.S. Census Bureau tabulates census data, and can be as small as a residential block.
OIG’s publication of this data built on a prior OIG report from March 31, 2020, which had published a list of 22 larger facilities that EPA’s internal analyses suggested were significant contributors to census tract-level cancer risks identified in the 2014 NATA. OIG’s March 2020 report had recommended that EPA engage residents near those facilities and hold public hearings to inform them of cancer risks. At the time, the Trump-era EPA had declined the invitation and instead proposed to conduct additional refined risk assessments for EtO and communities around EtO sources.
The May 6, 2021 report notes that EPA specifically requested OIG not publish the facility list. EPA explained to OIG that the list “likely reflects many false positives” because, among other reasons, (1) the list was generated from draft working files and was not part of the 2014 NATA itself; (2) the list was based on emission rates prior to 2014 and thus may be outdated; and (3) for those facilities that EPA’s model had only shown to potentially contribute to high (i.e., greater than 100 in 1 million) cancer risk in a block size area rather than a full census tract area, EPA had never checked whether the “high” concentrations were predicted to occur anywhere that people actually lived (which is important, because the IRIS EtO URE is cancer risk assuming continuous exposure for an entire 70 years at that rate of exposure, and thus if the high concentration is in a block or area where people do not actually reside, then no one is being exposed to the concentrations on a continuous lifetime basis). Given these and other data limitations, EPA told OIG, “Even at the census tract, the risk results can be uncertain and do not provide actionable risk information. The census block level risks used by the OIG in developing Appendix C are even less reliable.” Nonetheless, OIG published the list of facilities, responding to EPA’s hesitancy about the data accuracy and false positives by quoting from the 2014 NATA that “risk estimates are always uncertain to some degree,” and “uncertainty does not prevent EPA from making a statement of risk.”
Notably, EPA’s objections to the accuracy of the draft data published by OIG are not merely hypothetical. For example, following publication of the 2014 NATA in 2018, the Georgia Environmental Protection Division (EPD) undertook extensive statewide monitoring and modeling of EtO levels in the air both near all commercial EtO sources identified by EPD and in urban and rural sites away from industrial emission sources. GA EPD found that when more up-to-date information about industrial EtO sources was taken into account, EtO concentrations were not as high as suggested in EPA’s screening model. Whether this is due to decreases in emissions since 2014 or some other factor, it is in any case consistent with EPA’s comment that the data OIG is publishing is not actually certain enough to be used to draw concrete risk conclusions absent further study.
Consequences of Publishing Draft Risk Analyses
OIG’s laser focus on transparency also comes with a cost, especially when it disregards the risk of causing publicity for facilities that do not in fact impose significant cancer risks but are erroneously (or at least debatably) identified as doing so. For instance, any facility publicly identified as potentially causing increased cancer risk is at risk of private lawsuits, whether or not the draft reports are accurate. Accordingly, when OIG published its list of 22 major EtO emitters in May 2020, the results were predictable to those of us in the legal community: a spate of toxic tort lawsuits. Plaintiffs’ attorneys have no incentive to wait for further modeling refinements, and have in the past couple of years filed many lawsuits against EtO-emitting facilities armed with little else than reports from government agencies that purport to identify the facility as being associated with cancer risks based on the IRIS EtO URE. And because class actions and toxic tort lawsuits can be costly to defend and entail lots of negative publicity that can damage corporate goodwill, many companies may be tempted to settle whether or not they are actually imposing cancer risk.
This litigation risk caused by hastily publicized modeling data creates another unintended consequence: Companies are hesitant to voluntarily share their data and expertise with government agencies, out of the reasonable concern that the information may be published and then used as the basis for a toxic tort lawsuit. After all, recall that the IRIS EtO URE is orders of magnitude lower than the amount of EtO EPA has measured in background air away from industrial EtO sources, so it is not surprising that the area around an industrial source of EtO may also have concentrations of EtO above the IRIS EtO URE, no matter how efficient the facility’s processes and control devices are. These reasonable sensitivities to litigation risk from public disclosures in turn can complicate government agencies’ efforts to get the data they need from industry, by making industry reluctant to release it.
EPA has continued to push forward with information gathering activities to help it better assess EtO-related exposures and control options. For example, in addition to EPA’s recent information requests to commercial sterilizers, on April 29, 2021, EPA also announced a plan to expand Toxic Release Inventory (TRI) reporting to encompass more sources of EtO. The TRI program already requires various industries (including many medical sterilization facilities) to provide EPA with annual reports of their use and emission of certain quantities of certain pollutants (including EtO). The exact revisions to TRI requirements have not yet been proposed, but EPA’s press release indicates that the TRI program will be expanded to cover several smaller sources of EtO not previously subject to TRI reporting. The data generated from this more expansive TRI reporting will likely be used in future iterations of the NATA to generate screening level analyses of cancer risk based on more comprehensive and up-to-date emissions data. But even with more comprehensive and up-to-date data, the analyses are likely to continue to cause concern in local communities and regulated industries so long as the 2016 IRIS EtO URE continues to be used to calculate actual risk.
At the end of the day, because state and federal EtO cancer risk estimates vary by a factor of over a thousand between the IRIS EtO URE and the TCEQ’s 2020 study, just knowing the amount of EtO a facility may be exposing surrounding communities to is not particularly helpful in determining cancer risk and whether any additional regulations are necessary. Accordingly, until there is a shared understanding in the scientific and regulatory communities of the level of risk associated with EtO, and the level of exposure at which that risk becomes unacceptable, both EPA’s emphasis on facility monitoring and OIG’s emphasis on community publicity are likely to continue generating litigation risk and regulatory uncertainty.