Enhanced detection of African swine fever virus in samples with low viral load using digital PCR technology

Detection of low viral load samples has long been a challenge for African swine fever (ASF) prevention and control. This study aimed to compare the detection efficacy of droplet digital PCR(ddPCR) and quantitative PCR(qPCR) for African swine fever virus (ASFV) at different viral loads, with a focus on assessing the accuracy of ddPCR in detecting low viral load samples. The results revealed that ddPCR had a detection limit of 1.97 (95% CI 1.48 - 4.12) copies/reaction and was 18.99 times more sensitive than qPCR (detection limit: 37.42, 95% CI 29.56 - 69.87 copies/reaction). In the quantification of high, medium, and low viral load samples, ddPCR showed superior stability with lower intra- (2.06% - 7.58%) and inter-assay (3.83% - 7.50%) coefficients of variation than those of qPCR (intra-assay: 8.08%-29.86%; inter-assay: 9.27%-34.58%). Bland-Altman analysis indicated acceptable consistency between ddPCR and qPCR for high and medium viral load samples; however, discrepancies were observed for low viral load samples, where two samples (2/24, 8.33%) exhibited deviations beyond the acceptable range (-46.18 copies/reaction). Moreover, ddPCR demonstrated better performance in detecting ASFV in clinical samples from asymptomatic pigs and environmental samples, with qPCR showing false negative rates of 7.69% (2/26) and 27.27% (12/44), respectively. McNemar analysis revealed significant differences between the two methods (