"It's February and this patient appears to have influenza. I need a result right now. I know there's a rapid test available."
Anyone involved with infectious disease laboratory testing has heard the above refrain from physicians in the emergency room, intensive care unit, or infectious disease service. Assay manufacturers are constantly developing rapid diagnostic tests for a large variety of infectious diseases. However, an assay's speed does not always correlate with an assay's performance. Magauran et al,1 in this issue of Infectious Diseases in Clinical Practice, report their experience in using rapid assays for detecting influenza and respiratory syncytial virus (RSV) in their hands during a 2-year time frame. Their laboratory, at the Cleveland Clinic Foundation in Cleveland, Ohio, USA, instituted rapid enzyme immunoassay (EIA) testing for influenza and RSV in the fall of 2004. Specimens that tested negative by the rapid EIA procedures were then subjected to standard direct immunofluorescence testing for RSV and viral culture for influenza. The rapid EIA tests failed to detect 83% of direct fluorescent antibody (DFA)-positive RSV specimens and 64% of the culture-positive influenza specimens. By the end of the 2004 to 2005 season, the clinicians had stopped requesting the rapid EIA tests. Overall, the rapid EIA assays had been performed on less than 11% of the specimens submitted for influenza or RSV analysis during the 2004 to 2005 winter season.
The following winter, Magauran et al1 instituted real-time reverse-transcriptase polymerase chain reaction assays (rtPCRs) for influenza A, B, and RSV as a replacement for the rapid EIA procedures. The reported increased sensitivities of these rtPCR assays over the EIA procedures meant that confirming DFA or culture procedures would not be needed on rtPCR-negative specimens. During the 2005 to 2006 winter, 33% of the specimens submitted for influenza or RSV were assayed by the rtPCR method. Although the rtPCR method is not as rapid as the EIA, it is much quicker than viral culture, requires less hands-on time, and is less subjective than DFA. During the respiratory virus season, the Cleveland Clinic performed this assay daily; thus, results were available for most specimens in less than 24 hours.
Molecular assays often offer improved sensitivity and specificity over enzyme immunoassay, immunofluorescent procedures, and, in some cases, culture. However, the price one often pays for this improvement is a more complex assay that requires specialized equipment, highly trained technologists, and a laboratory where workflow can be designed to minimize contamination. More recently, however, clinical molecular infectious disease testing has been automated and simplified so that the nucleic acid extraction, amplification, and detection steps are all performed in a closed container with little chance of contamination. At present, this capability is not present in all hospital laboratories; thus, it remains a specialized procedure. However, as more assays become available in this format, both clinical demand and assay performance will drive smaller institutions to invest in real-time PCR technology.
The lesson to be learned form this "Lessons Learned" paper is that a rapid result is not always the best result. Clinicians must understand that the availability of a rapid (or any other test) does not, by itself, mean that the results from that procedure are accurate. A laboratory may run the test flawlessly, with appropriate controls, but if the performance characteristics of the assay are poor, the results may lead the clinician to an incorrect conclusion. Open, consistent communication between laboratory directors and clinicians is essential for proper assay selection and interpretation. Laboratory directors must understand the needs of the clinicians and patients. Clinicians must interpret results in the context of the particular assay's sensitivity, specificity, and predictive value. Both clinicians and laboratory directors must understand that assay offerings at a particular institution are selected based on clinical needs and assay performance, as well as budgetary and personnel constraints imposed on the laboratory by the institution.
Finally, rapid turnaround time, although an admirable goal, by itself does not make for a useful clinical assay. Performance data are more important to proper interpretation. Magauran et al1 monitored their clinicians' ordering practices, evaluated their rapid EIA offerings, and decided that their procedures were inadequate for their clinicians' needs. They changed their procedure to an rtPCR assay with a slightly longer turnaround time. The data presented by Magauran et al1 indicate that their rtPCR method for influenza and RSV is a well-performing assay that provides a clinically useful turnaround time. Their paper is an excellent example of applying a lesson learned.
1. Magauran CE, Yen-Lieberman B, Dhar S, et al. Lessons learned: moving from a rapid immunoassay to a molecular platform for respiratory viral testing during influenza season. Infect Dis Clin Pract (Baltim, Md)