Read our COVID-19 research and news.

Murky issue. EPA officials say they can’t give Congress data associated with studies of air pollution and health because it contains personal information, but a senior Republican disagrees.


In Battle Over EPA Subpoena, Privacy Remains Sticking Point

At an impasse over privacy issues surrounding decades-old health records, U.S. Representative Lamar Smith (R-TX) and the Environmental Protection Agency (EPA) are fighting the battle over data transparency one research paper at a time. A recent exchange of letters by the two sides is highlighting just how complex the issue of releasing government-funded health data can become.

Smith, the senior Republican on the House of Representatives science committee, issued a controversial subpoena to EPA on 1 August for all the raw data from a number of federally funded studies linking air pollution to disease. He says that congressional staff members and independent researchers should be able to review the data, which underpin a number of far-reaching federal regulations. Critics of the subpoena, including the panel’s Democrats, say the studies have already undergone rigorous scrutiny and that releasing the data would breach confidentiality pledges.

In its 19 August response to the subpoena, EPA submitted a letter and a collection of documents. These didn’t include raw data from either of the two major studies in question—one known as the Harvard Six Cities Study and the other by the American Cancer Society (ACS). Both were conducted in the 1990s. But the letter, obtained from the committee by ScienceInsider, describes EPA’s efforts to obtain and turn over data from several more recent follow-up studies, also requested in the subpoena. 

In a letter to Administrator Gina McCarthy on 3 September, Smith called EPA’s response “a combination of excuses, vague assurances, and already public information,” and gave the agency until 30 September to produce the data he asked for.

One problem, EPA Associate Administrator Laura Vaught wrote in the 19 August letter, is that much of this data belongs to the research institutions behind the two major studies—Harvard University and ACS, headquartered in Atlanta. EPA lacks the legal authority to force third parties to give up their data, she argued, and no subpoena can change that fact. While the agency will keep working with the institutions and the committee to satisfy the subpoena, “some data may remain outside of the EPA’s control and, therefore, outside of our ability to provide to the Committee,” Vaught wrote.

In contrast, Smith argues in his letter that EPA has “no legal excuse” not to obtain the data from Harvard, ACS, and other institutions involved in the research. He cited a circular from the White House Office of Management and Budget stating that the federal government can obtain and publish data from federally funded research and can also authorize others to obtain such data.

A Harvard communications officer tells ScienceInsider that, to date, the university has complied with every EPA request for data related to the Six Cities research. On 2 August, the university released data from two follow-up studies mentioned in the subpoena: a 2009 pollution study led by one of original Six Cities authors and an extended follow-up to Six Cities published in 2012. EPA gave the earlier study to the House, but says the second contains sensitive health data that could violate the confidentiality agreement made with participants under the federal Public Health Service Act. Because part of the data comes from the Centers for Disease Control and Prevention’s National Death Index, EPA turned to CDC for advice on what they can disclose and how to properly “de-identify” the data to remove any personal information. EPA says that it is still awaiting an answer.

Dismayed with the holdup, Smith again suggested the agency hand over the data as is and leave the de-identifying process to the committee—an option that House Democrats, including Representative Eddie Bernice Johnson, the panel’s senior Democrat, strongly oppose.  Smith has repeatedly promised not to make sensitive information public and said privacy concerns should not be cause for further delay. If EPA is truly struggling to make the data anonymous, “the Committee can recommend several experts who we are confident could de-identify the data in a matter of days,” Smith offered.