Diagnostic performance of RT-PCR-based sample pooling strategy for the detection of SARS-CoV-2

Table 1 Evaluation of the strategy for pooling NS samples for analysis by RT-PCR

Analysis	Total pooling		Pooling Ct ≤ 25		Pooling Ct: 26–30		Pooling Ct: 31–35
Analysis	Value	95% CI	Value	95% CI	Value	95% CI	Value	95% CI
AUC-ROC	0.98	0.95–1.00	1.00	100–100	1.00	100–100	0.90	0.70–1.00
Sensitivity (%)	96.67	88.58–100	100	90–100	100	90–100	80	34.94–100
Specificity (%)	100	95.83–100	100	95.83–100	100	95.83–100	100	95.83–100
Predictive value + (%)	100	98.20–100	100	90–100	100	90–100	100	87.50–100
Predictive value – (%)	92.31	73.98–100	100	95.83–100	100	95.83–100	92.31	73.98–100
Likelihood ratio +	NE	NE	NE	NE	NE	NE	NE	NE
Likelihood ratio −	0.03	0.00–0.23	NE	NE	NE	NE	0.20	0.03–1.15
Youden index	0.97	0.90–1.03	1.00	100–100	1.00	100–100	0.80	0.45–1.15

AUC Area down the curve, ROC receptor operating characteristic, NE not estimable, NS nasopharyngeal swab, RT-PCR reverse transcriptase-polymerase chain reaction. The likelihood ratio could not be estimated in all groups and subgroups since the sensitivity and specificity had values of 100%

ISSN: 1476-0711