Objective To compare a point of care (POC) nucleic acid amplification based platform for rapid diagnosis of COVID-19 against the standard laboratory RT-PCR test and perform an implementation study.
Design: prospective clinical trial (COVIDx) and observational study
Setting: a large UK teaching hospital
Participants: patients presenting to hospital with possible COVID-19 disease and tested on a combined nasal/throat swab using the SAMBA II SARS-CoV-2 rapid POC test and in parallel a combined nasal/throat swab for standard lab RT-PCR testing. Implementation phase participants underwent SARS-CoV-2 POC testing for a range of indications over a ten day period pre and post SAMBA II platform implementation.
Main outcome measures: concordance and sensitivity and specificity of POC using the lab test as the reference standard, test turnaround time in trial and implementation periods; time to definitive patient triage from ED, time spent on COVID-19 holding wards, bay closures avoided, proportions of patients in isolation rooms following test, proportions of patients able to be moved to COVID negative areas following test.
Results
149 participants were included in the COVIDx trial. 32 (21.5%) tested positive and 117 (78.5%) tested negative by standard lab RT-PCR. Median age was 62.7 (IQR 37 to 79) years and 47% were male. Cohen's kappa correlation between the index and reference tests was 0.96, 95% CI (0.91, 1.00). Sensitivity and specificity of SAMBA against the RT-PCR lab test were 96.9% (95% CI 0.838-0.999) and 99.1% (0.953-0.999) respectively. Median time to result was 2.6 hours (IQR 2.3 to 4.8) for SAMBA II and 26.4 hours (IQR 21.4 to 31.4) for the standard lab RT-PCR test (p<0.001). In the first 10 days of the SAMBA II SARS-CoV-2 test implementation for all hospital COVID-19 testing, analysis of the first 992 tests showed 59.8% of tests were used for ED patients, and the remainder were done for pre-operative screening (11.3%), discharges to nursing homes (10%), in-hospital screening of new symptoms (9.7%), screening in asymptomatic patients requiring hospital admission screening (3.8%) and access to interventions such as dialysis and chemotherapy for high risk patients (1.2%). Use of single occupancy rooms amongst those tested fell from 30.8% before to 21.2% after testing (p=0.03). 11 bay closures were avoided by use of SAMBA over ten days. The post implementation group was then compared with 599 individuals who had a standard lab RT-PCR test in the 10 days prior to SAMBA introduction. Median time to result during implementation fell from 39.4 hours (IQR 24.7-51.3) to 3.6 hours (IQR 2.6-5.8), p<0.0001 and the median time to definitive ward move from ED was significantly reduced from 24.1 hours (9.2-48.6) to 18.5 hours (10.2-28.8), p=0.002. Mean length of stay on a COVID-19 holding ward decreased from 58.5 hours to 29.9 hours (p<0.001) compared to the 10 days prior to implementation.
Conclusions
SAMBA II SARS-CoV-2 rapid POC test performed as well as standard lab RT-PCR and demonstrated shorter time to result both in trial and real-world settings. It was also associated with faster time to triage from the ED, release of isolation rooms, avoidance of hospital bay closures and movement of patients to COVID negative open green category wards, allowed discharge to care homes and expediting access to hospital investigations and procedures. POC testing will be instrumental in mitigating the impact of COVID-19 on hospital systems by allowing rapid triage and patient movement to safe and appropriate isolation wards in the hospital. This is also likely to reduce delays in patients accessing appropriate investigation and treatment, thereby improving clinical outcomes.