Table 1.
Median responses for each set of samples with each of six antigens as % positive control for that antigen.
| Antigen or Antigen Combination Using R | MOD 171 | JRRR 121 | RT237 F2 | SMP 70 | ZAM 295 | ST 123 | ST123 + ZAM295 | ST123 + JRRR121 | ST123 + JRRR121 + ZAM295 |
JRRR121 + SMP70 + ZAM295 |
All Six Antigens | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| % Pooled Pos | Probability of positive (0 to 1) with antigen combinations |
||||||||||||
| 1 | PEI++ n = 16 median |
27.3 | 38.8 | 56.5 | 88.4 | 96.3 | 79.0 | 0.9 | 0.9 | 0.9 | 0.9 | 0.8 | |
| 2 | F5 plus F6 negative, n = 80 median |
24.7 | 17.4 | 48.8 | 54.2 | 50.7 | 23.7 | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | |
| 3 | p * (F5 plus F6 compared to PEI++) | ns * | <0.0001 | ns | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | |
| Sensitivity analysis for single antigens |
Sensitivity/specificity analysis for antigens
combined in R |
||||||||||||
| 4 | Optimal cut-off for >93% specificity | 57 | 40 | 76 | 93 | 91 | 68 | 0.8 | 0.63 | 0.73 | 0.63 | 0.78 | |
| 5 | Optimal sensitivity % | 25.0 | 50.0 | 6.3 | 37.5 | 75.0 | 62.5 | 56.3 | 75.0 | 68.8 | 68.8 | 68.8 | |
| 6 | Optimal specificity % |
95.0 | 96.3 | 93.8 | 95.0 | 93.8 | 93.8 | 93.8 | 97.5 | 97.5 | 97.5 | 100.0 | |
| 7 | Accuracy % (sens + spec)/2 |
60.0 | 73.1 | 50.0 | 66.3 | 84.4 | 78.1 | 75.0 | 86.3 | 83.1 | 83.1 | 84.4 | |
| Data for PEI++ relative to just F5 or F6 at cut-offs for above sensitivity | |||||||||||||
| 8 | F5 | Specificity % | 97.5 | 92.5 | 92.5 | 95 | 95 | 97.5 | 97.5 | 100 | 97.5 | 100 | 100 |
| 9 | p (F5 compared to PEI++) | <0.05 | <0.0001 | ns | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | |
| 10 | F6 | Specificity % | 92.5 | 100 | 95 | 95 | 92.5 | 90 | 90 | 95 | 97.5 | 95 | 100 |
| 11 | p (F6 compared to PEI++) | ns | <0.0001 | ns | <0.01 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | |
Median responses for each set of samples with each of six antigens as % positive control for that antigen. Optimal cut-offs are based on 16 PEI PCR- and IDEXX-positive samples (PEI++) compared to all 80 PEI F5 and F6 negatives (again as % positive control), and sensitivity and specificity of each group are based on those cut-offs. In some cases, such as for MOD171, the optimal cut-off by ROC analysis led to a very low specificity; cut-offs were then identified to give maximum accuracy (average of sensitivity and specificity) at >93% specificity. The data for F5 and F6 samples are also presented separately, using the same cut-offs. The final five columns show the median calculated negative or positive status (on 0 to 1 scale) of PEI++ compared to all 80 F5 and F6 negative samples, combining results with several specified antigens using R statistics; the calculated values for each serum sample have then been analysed by ROC and optimal cut-offs determined to provide the highest accuracy (average of sensitivity and specificity). * Significance values were calculated using the Tukey test; ns = not significant.