The U.S. Government Accountability Office (GAO) recommended the FBI improve its procedures after conducting a study on the bureau's statistical analyses and genetic test development process for samples of the anthrax-causing bacterium that was intentionally released in 2001.
GAO examined the genetic tests developed by four contractors to detect certain genetic mutations in samples of B. anthracis.
While the tests met the FBI's criteria and were scientifically verified and validated, GAO found a comprehensive framework lacking on the part of the FBI. Each contractor took a different approach to developing the tests, with one skipping the key step of verification testing. Standards of minimum performance for verification and validation would benefit future investigations, the GAO said.
The GAO outlined six elements of a statistical framework and measured the FBI's approach against them, finding room for improvement in three of the six areas.
First, the sampling procedures to gather the B. anthracis did not have rigorous controls. Second, the FBI didn't account for measures of uncertainty. Third, the FBI used genetic mutations to differentiate between the samples, but did not offer a comprehensive understanding of the conditions and methods that give rise to this distinction.
Additionally, the value of using genetic mutations to analyze samples is not clear. There is also a gap in the understanding of how bacteria mutate that could impact future testing.
Research into areas such as genetic mutation and genome sequencing methods funded by the Department of Homeland Security may help close these gaps, though the research results may not be available for several years.