Determining diagnostic sensitivity loss limits for sample pooling in duplex rtPCR surveillance testing: and .

To expand surveillance testing capacity through sample pooling, a thorough understanding is needed of how sample dilution through pooling affects the sensitivity of candidate assays. We validated a robust and representative framework for assessing the dilution effect of sample pooling using duplex rtPCR surveillance of and , both of which are causative agents of severe anemia in cattle and a serious threat to the cattle industry in Virginia and many other states. We used 200 known-positive samples with Ct values representative of typical surveillance results in a series of pools in which we re-tested each sample individually, followed by each sample diluted in equal volumes with negative samples to make pools of 2, 4, 6, 8, and 10 total samples. We compared the Ct values of the individual positives with the Ct values of each pool size to determine if Ct values increase past the limit of detection in the 45-cycle assay. We observed a maximum of 2% sensitivity loss (no more than 2 of 100 samples returned a false-negative result) for both and during the pooling series, with lower-than-expected average Ct increase and sensitivity loss. We conclude that pooling up to 10 samples would be acceptable for regional surveillance of and using our rtPCR assay. The described strategy is applicable to validate pooling for a wide range of single and duplex rtPCR assays, which could expand efficient disease surveillance.