Background: Rapid influenza detection tests (RIDT) have poor sensitivity compared with polymerase chain reaction (PCR) method and low positive predictive value even during time of high influenza prevalence. National Guidelines suggest considering reflex PCR testing for negative RIDT results.
Methods: Intermountain Healthcare (IH) operates a nonprofit system of 22 hospitals, 41 urgent care centers, and more than 185 physician clinics serving the Intermountain West. We identified RIDT (BD Veritor) ordered during high/moderate influenza activity (Dec 11-Mar 11) for the 2016-17 season at any one of the IH locations and calculated the proportion of RIDT reflexed within 48 hrs. We compared the outcomes for RIDT tests order in the ED alone vs reflex testing. Outcomes included admission after ED visit, oseltamivir prescription, 7-day readmission to ED or hospital for patients discharged home after initial ED visit with RIDT, and antibiotic prescriptions within 14 days of RIDT.
Results: 9,623 patients were evaluated during the study period with overall reflex to PCR within 48 hrs of 9.6%. The frequency of PCR reflex by patient setting is provided in Figure 1. The impact of influenza testing patterns ordered in the ED is shown in Figure 2. Among patients with a negative RIDT, hospital admissions of patients with a positive or negative reflex PCR were comparable (57% vs. 64%; p=NS), but was significantly higher than those when no reflex PCR was performed (57% vs. 12%; p<0.001) and (64% vs. 12%; p<0.001). The use of oseltamivir was low and variable among those with a positive test, but significantly higher compared to patients with a negative test (10% vs. 1%; p<0.001).
Conclusion: Despite the known problems with the negative predictive value of RIDTs during periods of high influenza activity, only 21% had a more sensitive PCR performed. PCR positive results appear to lead to increased oseltamivir use. Patients with RIDT negative and PCR negative results are admitted more frequently than patients with RIDT negative with no PCR results.
E. K. Korgenski,
B. K. Lopansri, None
C. L. Byington, BioFire: Collaborator and Grant Investigator , Licensing agreement or royalty and Research grant
J. Wilkes, None
P. Gesteland, None
A. Pavia, None