The COVID-19 pandemic has increased the global demand for laboratory testing services including test reagents, sampling devices, laboratory instruments and personal protective equipment, at the same time the restriction on movement and travel have led to a massive disruption of supply chain around the world. As a result, there is scarcity of resources to meet the need of the hour, therefore we need newer assays with faster turnaround time and those can accommodate techniques like sample pooling to meet the diagnostic demands. Automation is the most efficient way of utilizing skilled manpower and is cost -effective. The Cobas 6800 has a uniform process which uses a universal sample preparation process to isolate, purify and extract the nucleic acids in the sample processing module rather than other platforms that use varied nucleic acid binding techniques optimized for preferentially isolation of nucleic acids. [3]
This study found that the sensitivity and specificity of the Cobas 6800 was 100% when compared to the standard assay. The performances of the Cobas 6800 when using pooled samples showed that specificity was 100% and the sensitivity varied from 95% in six pooled sample to 100% in both five and four pooled samples. These results agreed with the other validation study reports. [6–8]
Among the six-pool sample strategy the discordant results were identified among the low viral load samples with Ct values > 27. Among the low viral load samples that were detected in Labgun RT-PCR kit with Ct > 25, E target was the only gene that was detected with the Cobas 6800 assay. Similar observation was also reported in another evaluation of Cobas 6800, reiterating the manufacturer claim of higher sensitivity for E gene compared to the ORF1ab gene. [9]
This study evaluated the manufacturers claim of a lower limit of detection of 25 copies/ml for Target 1 and 32 copies/ml for Target 2. [4] Similar reports were found in this study where nine different dilutions ranging from 25 copies/ml to 78,125 copies/ml were tested and 25copies/ml was found to be the lowest detectable concentration for both the target genes. Other studies have also supported this claim that Cobas 6800 has a lower limit of detection for SARS-COV-2 detection and the Cobas E gene is more sensitive among the various other target genes in low viral load samples. [10,11 ]
The inter assay precision study done on five consecutive days using 6 pooled, 5 pooled and 4 pooled samples showed that the CV% varied between 1.5 to 4% which was within the acceptable range of 5% in reference to our lab standards. The study also observed a strong correlation between the Ct values of the target genes of both the assays, which was also supported by other studies that demonstrated a strong correlation between the SARS-CoV-2-specific targets obtained by cobas and the comparator assays.[7]
This study also demonstrated that there was no carry over between samples and the error limit was calculated as zero. False positive results may occur if carryover of samples is not adequately controlled during sample handling and processing. However, the Cobas 6800 has an inbuilt carry over contamination control system containing a uracil N-glycosylase which can enzymatically remove the PCR amplicon from a previous reaction without degrading naive DNA. [3]
The pooling technique run on Cobas 6800 was highly efficient when the positivity rate was low, during the months of August 2021 to December 2021, when the average positivity rate was around 0.3–0.1%. The pooling technique was proving to be cost effective and quicker with faster TAT when the positivity rate was low. However, after the emergence of the new omicron variant in the latter half of December 2021, the average daily cases reported in the UAE increased which thereby increased the slide positivity rates to up to 8% during the month of January 2022. When the slide positivity rate was high the pooling method led to an average positive repeat percentage of 27% therefore the pooling technique failed to be cost effective when the positivity rate increased. As supported by other studies which showed that pooling technique were effective only when the prevalence rate was low. [12–14]
When selecting a SARS-CoV-2 RNA assay the clinical performances of sensitivity and specificity alone is not sufficient there are other parameters that play a paramount role in selection specially when employed in large scale testing. Sample throughput, turnaround time (TAT), accommodation of techniques like sample pooling, hands on time, ease of method, availability of reagents and cost of testing are considered equally when it comes to large scale testing.
The Cobas 6800 assay uses primers designed to detect the ORF1 a/b genomic region, and studies have shown that ORF1ab, RdRp primers have shown to have better analytical performances in identifying the SARS-CoV2-RNA. [15] This facilitates Cobas 6800 to be more efficient in detecting the SARS-CoV-2 infections.
The Cobas 6800 is a fully automated platform with the turnaround time (TAT) of two hours, however the techniques of pooling require significant amount of time spend in the preparation of sample pools. The barcodes of samples pooled are scanned and captured in the system and after the pooling of samples into the secondary tubes, secondary barcodes are created and the system matches the secondary barcodes to the related primary sample barcode samples. for a single run of 5 pooled samples 480 samples need to be pooled and this preparation including sample barcoding and aliquoting alone required an average of two hours. Therefore, the total TAT to run 480 samples was about 5 hours. On the contrast Labgun Exofast RT-PCR kit has a TAT of 40 mins to run 96 samples, however the initial sample barcode scanning and aliquoting followed by nucleic acid extraction and purification with the MGI extraction kit requires an additional 90 minutes. This shows Cobas 6800 RT-PCR platform was still time efficient when compared to Labgun Exofast RT-PCR kit.
The Cobas 6800 is an expensive test compared to the Labgun Exofast RT-PCR kit as the price of the cartridge is higher compared to the comparator test, however the pooling techniques employed greatly reduces the per capita cost of the test. Further the automation process of the Cobas 6800 platform reduce the cost of skilled labor as it eliminates the need for skilled man power, whereas Labgun requires additional man power with specific skills who are required at different work stations such as nucleic acid isolation, extraction and to run the PCR assays. Hence when the sample load is high Cobas 6800 platform has proven to be cost effective than the comparator.
Our study has some limitations the study did not include the clinical data on the severity to correlate with the sensitivity which could have given further insight on the performance of the automated Cobas 6800 platform.
In conclusion the overall performance of the Cobas 6800 was superior than the standard assay for detection of SARS-COV-2 from an operational stand point. The automated system significantly improves work flow and processes large number of samples with shorter turnaround time. This study concludes that Cobas 6800 RT-PCR assay is a reliable platform for qualitative and rapid detection of SARS-CoV-2 and can be effectively utilized for pooling of samples with highly efficient performance when the disease prevalence is less.