Effect of Respiratory Viral Panel Adoption on Antibiotic Use in Ventilated Patients

Abstract

Rationale: Rapid respiratory viral panel (RVP) testing has become widely used to aid in the diagnosis and treatment of acute respiratory failure. However, the impact of RVP on antibiotic stewardship in critically ill patients is unclear. Objectives: To assess if adoption of RVP testing at hospitals was associated with changes in antibiotic duration in intensive care unit patients receiving invasive mechanical ventilation. Methods: With data from the Premier Inc. database from 2016 to 2019, we used interrupted time series with multivariable hierarchical linear regression models to quantify trends in outcomes for 31,644 patients in the 12 months before RVP adoption, the level change in outcomes at the time of RVP adoption (estimand of interest), and changes in outcome trends in the 12 months after RVP adoption. Results: Hospital adoption of RVP testing (n = 62,603) was associated with a decrease in days of antibiotics by 0.5 days (95% confidence interval, -0.8, -0.1) in the first month after adoption. There was also a significant decrease in the risk of Clostridioides difficile infection by 0.9% (95% confidence interval, -1.6, -0.3). There were no significant changes in other outcomes, including hospitalization costs, hospital length of stay, or rates of ventilator-associated pneumonia. Conclusions: Hospital adoption of RVP testing was associated with modest reductions in both antibiotic duration and risk of C. difficile infection among intensive care unit patients with acute respiratory failure and suspected infection.

Keywords: DNA probes; point-of-care testing; polymerase chain reaction; respiratory tract infections.

Figure 1.

Respiratory viral panel use over time, according to hospital type. Adopter hospitals were defined as hospitals with no respiratory viral panel use (based on charge codes) in 2016, followed by first use in 2017 or 2018 and continued use into 2019; prevalent-use hospitals were defined as hospitals with respiratory viral panel use in each year of the study period; and intermittent-use hospitals were defined as hospitals with respiratory viral panel use in some years but not others and that did not meet criteria for alternative classification. Hospitals that never used respiratory viral panels are excluded from the figure.

Figure 2.

Change in respiratory viral panel use and antibiotic use before and after hospital adoption of viral respiratory panel testing. Trends in the rates of viral panel testing and antibiotic days before and after adoption of respiratory viral panel testing. Shown are (A) viral panel testing rates and (B) days of antibiotics binned by month (black points) and associated lines of best fit (blue lines). The thick black line denotes the month of hospital adoption of viral panel testing.