Application and Interpretation of Findings From PCR-Based Diagnostic Tests

Interpretation of test results should consider the validation data in relation to the PCR method used; that is, the analytical and diagnostic sen­sitivity and specificity for the method including both the DNA extraction and PCR detection steps.

Thus, individual diagnostic faecal PCR test results must be considered carefully in terms of how these relate to the disease status of the animal. As qPCR provides an indication of the amount of MAP present in the sample, such results can be stratified into ‘heavy’ or ‘high’ shedders, compared to moderate, low or non-shedding individuals. Interpretation of in­dividual animal results needs to consider the pathobiology of the disease and the fact that PCR can detect DNA from non-viable organ­isms. A negative individual faecal PCR test re­sult, as with individual faecal culture, does not necessarily mean that the animal is not infected as this can be dependent on the stage of disease and hence the shedding level and presence of MAP in the faeces. In addition, the age of the animal tested is important: young calves have been demonstrated to shed MAP (Wolf et al., 2015), however transient shedding following an exposure event may not necessarily be cor­related with infection status (Begg et al., 2018).

Due to the clinical course of the disease, surveillance is often focused at the herd or flock level. Herd-level testing strategies can involve collection of a mixed ‘slurry’ from the environ­ment, or the creation of pools from faecal sam­ples from individual animals for evaluation by PCR (Aly et al., 2009; Mita et al., 2016). In an evaluation of national surveillance methods for Irish dairy herds, pooled faecal testing (pool size of 5) was one of the methods with highest sen­sitivity, estimated to be 76% based on previous studies, both for the identification of infected herds and assurance testing for uninfected herds (Sergeant et al., 2019).

As a herd-level diagnostic sample, bulk tank milk (BTM) is cost effective and easy to obtain. PCR-based diagnostic tests have been applied to this sample type, as have MAP-specific milk an­tibody ELISAs, to examine cattle, goat and sheep dairy farms (Slana et al., 2008; Okura et al., 2012; Bauman et al., 2019). The issue with BTM testing using PCR-based diagnostics as a method of surveillance relates to the low overall diag­nostic sensitivity and specificity values in bulk milk samples in comparison with other pooled testing approaches. Bauman et al. (2019) evalu­ated a BTM PCR assay as a herd-level diagnostic surveillance test to identify paratuberculosis- affected goat and sheep dairy farms in Canada, for which the disease is a significant concern due to substantial economic and production-limiting impacts. The study compared diagnosis of flocks through testing of BTM using a PCR method, milk ELISA and faecal culture. The BTM PCR method had low diagnostic sensitivity (0-25%) when compared with individual faecal testing of 20 randomly selected >2-year-old animals by culture or PCR, with lower sensitivity than a milk ELISA on BTM (33.3-50%) based on the same comparison, but even these sensitivity es­timates may be over-inflated as the true disease status of the farms was not known.

Environmental sampling for herd-level surveillance is an option, but careful considera­tion needs to be given to how representative the sample type is for the purpose of the test and the production system (intensive vs extensive production, dairy vs beef). Species with pelleted faeces pose their own unique challenges and are more compatible with faecal pooling approach­es. The sensitivity of herd environmental testing protocols is reported to be high, with good agree­ment between culture and qPCR test approaches (Berghaus et al., 2006; Aly et al., 2010; Lavers et al., 2013).

19.9