Background: Viral respiratory illnesses are common and treatment in most cases is largely supportive. Although extended testing, using PCR to identify specific viruses, is available, targeted treatments are limited to the influenza viruses. Hence, respiratory viral panel (RVP) test results may not affect clinical management for many patients. Our institutional infectious disease guidelines recommend reserving RVPs for patients with critical illness or immunocompromised state. Other options include a rapid influenza and respiratory syncytial virus PCR (Rapid), and point-of-care influenza (POC). The objective of this study was to describe the use of RVP testing at our institution, estimate its associated costs, and identify areas of potential overuse.
Methods: This study was based at an 800-bed academic teaching hospital in an urban setting and utilized data obtained from our Epic-based electronic health record (EHR). We identified all patients who had RVP, Rapid, and/or POC testing performed on the Emergency Department (ED) and Hospital Medicine services, from 11/2016 to 6/2017. We obtained data on encounter location, intensive care unit (ICU) status, length of stay, and diagnosis related group (DRG) illness severity, as well as direct cost and charge data from the hospital’s financial database. Linear regression was used to compare continuous variables. We then conducted chart review of a random sample of 30 non-ICU patients who underwent RVP testing to extrapolate the percentage of patients who were immunocompromised.
Results: We identified 1881 patients who received any respiratory viral testing, of which 991 underwent a RVP. The mean turnaround time for RVP results was 1.9 days and a sizable proportion of test results (26%, n=260) returned only after the patient was discharged from the ED or hospital. Chart review estimated that only 40% of non-ICU RVPs were performed in immunocompromised patients, leaving 52% (n=516) of the total RVPs as potentially unnecessary based on internal guidelines. The charge associated with each RVP test was $3450, leading to greater than $1.75 million in excess hospital charges over the 8-month time period. Compared to patients who only underwent influenza testing, the length of stay in patients in whom a RVP was ordered was 1.5 days longer, even after adjusting for DRG severity of illness in multivariable analysis.
Conclusions: We found that RVP testing was often obtained in patients for whom the results may not change clinical management. Due to the costs associated with excessive RVP testing, reducing unnecessary use could represent a significant area of value improvement.