Viability of Archived Respiratory and Influenza Specimens Used for Laboratory Preparedness Exercise

BACKGROUND:  The Department of Defense Global Laboratory-Based Influenza Surveillance Program, based at the U.S. Air Force School of Aerospace Medicine, utilizes a sentinel-based system to conduct respiratory surveillance across the Department of Defense.  Preparations for the influenza season include training laboratory staff and ensuring laboratory supplies are sufficient for an unpredictable workload.  The School of Aerospace Medicine recently relocated to Dayton, OH, from Brooks City-Base, San Antonio, TX. In an average respiratory season, approximately 8,000 specimens are processed, but this number increased to 20,000 specimens during the 2009 H1N1 pandemic.  As the 2012-2013 season approached, an exercise to simulate a sustained high volume of specimens was designed to test laboratory processes and procedures, utilizing the laboratory’s specimen archive. Goals of the exercise included streamlining work-flow, optimizing scheduling, improving communication, and testing the viability of archived specimens. Due to specimen age and the relocation, which potentially added temperature stress to the specimens, this study focuses on specimen viability.

METHODS:  Original nasal wash and nasopharyngeal respiratory specimens were selected from the 2008/2009 archive. During the week of 10-14 September 2012, 236 specimens were stripped of all personal identifiers and tested using the lab’s standard viral culture and molecular detection techniques. The exercise results were then referenced to the archive results using agreement, sensitivity, specificity, and Cronbach’s alpha.

RESULTS:  There was high agreement between exercise molecular and virology results and archive results from prior testing. Combined molecular and virology results had an agreement of 97.8% (95% confidence interval (CI): 95.6%-99.9%).  Molecular results agreed with the archive result data 97.7% (95% CI: 94.8%-99.0%) and virology results agreed with the archive results 98.1% (95% CI: 96.3%-99.9%). With the archive result being the gold standard, the sensitivity for molecular testing was 99.3% (95% CI: 96.0%-99.9%) and specificity was 95.2% (95% CI: 88.3%-98.1%). Virology testing had a sensitivity of 99.4% (95 % CI: 98.2%-100.4%) and a specificity of 94.6% (95% CI: 91.6%-97.6%). Cronbach’s alpha, which measures the internal consistency between variables, was determined to be 0.9 (out of 1.00). When results were removed for nonconsistency, there was an excellent alpha score of 0.98.

CONCLUSIONS:  The exercise demonstrated both success in handling a workload 3-5 times heavier than the past year and an overall high agreement, sensitivity, and specificity between the exercise molecular and virology results and the archive results. There was a strong correlation despite the archived specimens being 3 to 4 years old and being shipped from San Antonio, TX.