Embedded Files
structured to specifically indicate whether an incident involved incomplete inactivation, nor require laboratories to include “incomplete inactivation” in the description of the incident.25 As a result, officials cannot easily search their databases to identify such incidents once the information on the reporting forms is entered into the database. According to Select Agent Program officials, the current structure of the reporting forms could affect their understanding of the magnitude of the problem, and a more extensive analysis of the full report descriptions for all incidents in the database could identify additional incidents involving incomplete inactivation. Similar to the Select Agent Program, NIH has an incident reporting form and requires laboratories under its purview to report incidents involving recombinant pathogens. We identified four incidents involving incomplete inactivation of recombinant pathogens that occurred from 2003 through 2015, which were reported to NIH but were not provided to us, through a review of incident reporting forms (see table 4).26 NIH later confirmed that these four incidents did involve incomplete inactivation. For example, due to an equipment failure in 2014, researchers at an academic highcontainment laboratory inadvertently moved viable samples of a recombinant pathogen thought to be inactivated outside of containment. The university reported the incident to NIH, but NIH officials did not initially identify this incident as resulting from incomplete inactivation. As with the Select Agent Program’s form, NIH’s reporting form for incidents involving recombinant pathogens is not structured to specifically indicate whether an incident involved incomplete inactivation, nor requires laboratories to include “incomplete inactivation” in the description of the incident, leading to difficulty in identifying such reports in its database. 25The Select Agent Program uses a reporting document called a “Form 3” to document incidents of theft, loss, or release of select agents. 26These four incidents all involved pathogens that were also select agents—and thus also reported to the Select Agent Program and listed in tables 2 and 3 above—in addition to being recombinant pathogens. Equine Encephalitis Viruses Deceased Horse from a Venezuelan Equine Encephalitis Outbreak in Texas, 1976 According to the Centers for Disease Control and Prevention (CDC), the equine encephalitis viruses are mosquito-transmitted diseases that can cause severe inflammation of the brain (encephalitis) in horses and humans. Eastern equine encephalitis (EEE) most commonly occurs in the eastern United States and Canada while Venezuelan equine encephalitis (VEE) most commonly occurs in South and Central America but has spread to the United States. Although a rarely seen illness in humans, approximately 33 percent of individuals with EEE die from the disease and many survivors are left with significant brain damage; infected horses have variable death rates as high as 90 percent. VEE is the most infectious of the equine encephalitis viruses and the United States weaponized VEE as an offensive incapacitating agent before the termination of its biological weapons program. These viruses are highly infectious by the aerosol route and have caused more than 160 laboratory-acquired infections, according to the Biosafety in Microbiological and Biomedical Laboratories (BMBL) manual. Source: GAO analysis of scientific literature, agency documents, and the BMBL manual. CDC Public Health Image Library, James Stewart (photo) | GAO-16-642 Page 17 GAO-16-642 Inactivation of Pathogens Table 4: Four Incidents Involving Incomplete Inactivation of Recombinant or Synthetic Pathogens from 2003 through 2015 Identified by GAO and Confirmed by the National Institutes of Health (NIH) Laboratory type Pathogen Pathogen type Method of inactivation Year of incident Academic Burkholderia pseudomallei Bacteria Physical 2014 Academic Brucella abortus, Brucella melitensis & Brucella suis Bacteria Chemical 2012 Federal Francisella tularensis Bacteria Chemical 2011 Academic Francisella tularensis Bacteria Chemical 2007 Source: GAO analysis of information from NIH, 2016. | GAO-16-642 Note: These four incidents all involved pathogens that were also select agents—and thus also reported to the Federal Select Agent Program—in addition to being recombinant pathogens. These examples demonstrate challenges the Select Agent Program and NIH have with identifying which incidents involve incomplete inactivation for the pathogens they regulate. Because the Select Agent Program and NIH cannot easily identify which incidents involve incomplete inactivation on reporting forms and within incident databases, they do not know the frequency or reason these incidents occur, making it difficult to develop guidance to help prevent future incidents. A CDC internal review of the Select Agent Program issued in 2015 noted the need to include subcategories of “release,” “loss,” and other additional fields on reporting forms to more consistently identify and categorize incidents moving forward.27 In addition, Standards for Internal Control in the Federal Government states that agencies are to employ control activities, such as the accurate recording of events, to help ensure that management’s directives are carried out and actions are
Report abuse