Embedded Files
taken to address risks.28 Our prior work on safety reporting systems has noted that report formats should allow for a sufficient description of events that align with analysis decisions, including the ability to effectively perform root cause analysis on high-priority issues.29 In addition, a 2016 report we issued on biological safety in high-containment laboratories highlighted the importance of analyzing all incident reports to identify potential trends.30 27Centers for Disease Control and Prevention, 90 Day Internal Review of the Division of Select Agents and Toxins (Atlanta, Ga.: Oct. 22, 2015). 28GAO/AIMD-00-21.3.1. 29GAO, Biological Laboratories: Design and Implementation Considerations for Safety Reporting Systems, GAO-10-850 (Washington, D.C.: Sept. 10, 2010). 30GAO-16-305. Page 18 GAO-16-642 Inactivation of Pathogens Officials from the Select Agent Program stated that they plan to revise their reporting forms in response to the 2015 internal CDC review mentioned above and issue new guidance associated with the new forms, thus allowing officials to search the database for new and more consistent information. However, as of April 2016, Select Agent Program officials had not yet determined what additional information to include on the forms. Moreover, as of May 2016, NIH officials told us that they did not plan to update their reporting form. Second, the total number of incidents involving incomplete inactivation in high-containment laboratories that occurred from 2003 through 2015 is unknown because federal incident reporting, in general, is required only for (1) incidents that involve select agents, which are reportable to the Select Agent Program; and (2) incidents that involve recombinant pathogens, which are reportable to NIH. Thus, incidents involving incomplete inactivation of pathogens that are neither select agents nor recombinant pathogens, such as West Nile virus, are generally not required to be reported to any federal agency (see fig. 2).31 Figure 2: Federal Reporting Requirements for Incidents in High-Containment Laboratories 31Laboratory officials told us that in some situations, they may notify a local or state health organization if an incident with a nonselect agent raises significant public health or safety concerns. Absence of Reporting Requirements Page 19 GAO-16-642 Inactivation of Pathogens Reporting is required for select agents because the Select Agent Program has determined that they have the potential to “pose a severe threat to public and agricultural health and safety.” However, nonselect agents also have the potential to cause disease if not properly inactivated. For example, during the course of our review, we found an incident involving the incomplete inactivation of Mycobacterium tuberculosis, a nonselect agent that causes the disease tuberculosis, in one laboratory. Because there are no federal reporting requirements for incidents involving pathogens that are neither select agents nor recombinant pathogens, only the laboratory in which the inactivation procedure failed was aware of the incident. Some experts from our meeting noted that it would be beneficial if all pathogens in high-containment received the same level of scrutiny and had the same biological safety controls in place when leaving high containment, whether the sample is a select agent or not. However, one expert also noted that when the Select Agent Program was created, its focus was more on the biological security risks associated with pathogens rather than biological safety, which may contribute to what appears to be an artificial distinction between select and nonselect agents for incident reporting. As a way to address the issue of incident reporting in a broader scope, we recommended in 2016 that federal high-containment laboratories report all incidents, whether they involve select agents or not, to senior agency officials.32 Third, the total number of incidents involving incomplete inactivation that occurred from 2003 through 2015 is also unknown because there is currently no clear and consistent definition of inactivation in guidance or regulations issued by the Select Agent Program, NIH (for oversight of recombinant pathogens), or the BMBL manual.33 As a result, researchers may not consistently define inactivation, which potentially affects how and when they report incidents involving incomplete inactivation. Experts at our meeting stated there is a need for a clear, consistent definition of inactivation across key federal guidance documents, with some experts noting that the lack of a consistent definition can make it difficult to understand when an incident occurs. The BMBL manual also emphasizes the need for clear definitions to avoid misuse and confusion of key terms used in research, but does not itself define inactivation. In addition, our 32GAO-16-305. 33These guidance documents do include some information on inactivation, but do not include a definition. No Clear, Consistent Definition of Inactivation Page 20 GAO-16-642 Inactivation of Pathogens past work has shown that the use of standardized definitions is key to ensuring that information is reported consistently.34 Officials from the Select Agent Program, NIH, and CDC agreed that there is a need for a clear and consistent definition of inactivation across guidance documents. Specifically, the Select Agent Program is in the process of revising the select
Report abuse