Abstract
BACKGROUND: Rapid/point-of-care respiratory virus nucleic acid tests (NAT) may improve oseltamivir, antibiotic, diagnostic test, and hospital bed utilization. Previous randomized controlled trials (RCT) on this topic have not used standard procedures of an accredited healthcare and laboratory system. METHODS: We conducted a parallel RCT at two hospitals [paediatric = Alberta Children's Hospital (ACH); primarily adult = Peter Lougheed Centre (PLC)]. Patients with a respiratory viral testing order were randomized to testing at either a central accredited laboratory (standard arm) or with a rapid polymerase chain reaction test at an on-site accredited laboratory followed by standard testing [rapid on-site test (ROST) arm] based on day of specimen receipt at the laboratory. Patients and clinicians were blinded to assignment. The primary outcome for ACH was inpatient length of stay (LOS) and for PLC was the proportion of inpatients prescribed oseltamivir. RESULTS: 706 patient encounters were included at ACH; 322 assigned to ROST (181 inpatients) and 384 to the standard arm (194 inpatients). 422 patient encounters were included at PLC; 200 assigned to ROST (157 inpatients) and 222 to the standard arm (175 inpatients). The rate of oseltamivir prescription and number of doses given was reduced in PLC inpatients negative for influenza in the ROST arm compared to standard arm [mean 14.9% (95% CI 9.87-21.9) vs. 27.5% (21.0-35.2), p = 0.0135; mean 2.85 doses (SEM 2.39-3.32) vs. 4.17 doses (3.85-4.49) p = 0.022, respectively]. ROST also significantly reduced oseltamivir use at ACH, reduced chest radiographs (ACH), and laboratory test ordering (PLC), but not antibiotic prescriptions. ROST also reduced the median turnaround time by > 24 h (ACH and PLC). The LOS at ACH was not significantly different between the ROST and standard arms [median 4.05 days (SEM 1.79-18.2) vs 4.89 days (2.07-22.9), p = 0.062, respectively]. No adverse events were reported. CONCLUSIONS: In a RCT representing implementation of ROST in an accredited laboratory system, we found that a ROST improved oseltamivir utilization and is the first RCT to show reduced ancillary testing in both paediatric and adult populations. A larger study is required to assess reduction in paediatric LOS as ACH was underpowered. These findings help justify the implementation of rapid on-site respiratory virus testing for inpatients. Trial registration ISRCTN, number 10110119, Retrospectively Registered, 01/12/2021.