Skip to main content

Table 3 Parameters estimates (posterior means) with their corresponding 95% credibility intervals and the model goodness-of-fit to the data of the best model for four tests with and without covariates

From: Latent class models for Echinococcus multilocularis diagnosis in foxes in Switzerland in the absence of a gold standard

 

Model

Model with “age”

Model with “cestodes”

Model with “sex”

SCT

 Se1

0.885 (0.827–0.934)

0.879 (0.816 to 0.931)

0.876 (0.811 to 0.930)

0.878 (0.814 to 0.930)

 Sp1

1a

1a

1a

1a

Egg-PCR

 Se2

0.548 (0.485 to 0.610)

0.544 (0.482 to 0.608)

0.544 (0.482 to 0.606)

0.546 (0.483 to 60.8)

 Sp2

0.934 (0.873 to 0.991)

0.936 (0.872 to 0.990)

0.940 (0.874 to 0.992)

0.940 (0.874 to 0.993)

pAb-ELISA

 Se3

0.560 (0.480 to 0.639)

0.558 (0.477 to 0.637)

0.551 (0.471 to 0.631)

0.557 (0.476 to 0.638)

 Sp3

0.659 (0.558 to 0.756)

0.659 (0.555 to 0.758)

0.648 (0.540 to 0.749)

0.659 (0.552 to 0.759)

mAb-ELISA

 Se4

0.632 (0.553 to 0.708)

0.629 (0.550 to 0.706)

0.623 (0.544 to 0.701)

0.629 (0.549 to 0.707)

 Sp4

0.700 (0.601 to 0.794)

0.701 (0.600 to 0.797)

0.693 (0.590 to 0.791)

0.701 (0.598 to 0.799)

 Prevalence

0.595 (0.431 to 0.664)

na

na

na

 Cov = 1b

na

0.697 (0.594 to 0.794)

0.500 (0.398 to 0.606)

0.596 (0.594 to 0.794)

 Cov = 0b

na

0.558 (0.312 to 0.784)

0.692 (0.464 to 0.857)

0.631 (0.312 to 0.784)

 Intercept

na

0.83 (0.38 to 1.34)

0.00 (−0.04 to 0.43)

0.39 (−0.01 to 0.83)

 Slope (OR)c

na

0.55 (0.31 to 0.94)

2.24 (1.31 to 3.90)

1.16 (0.96 to 1.96)

 DIC

1507.0

1501.9

1497.2

1506.2

  1. aSpecificity of necropsy fixed to 1
  2. bPrevalence for respective covariate = 1 (adult, with other cestodes and male) and covariate = 0 (young, without other cestodes and female)
  3. cOdds ratio
  4. Abbreviations: Se sensitivity, Sp specificity, Egg-PCR polymerase chain reaction, pAb-ELISA polyclonal enzyme-linked immunosorbent assay, mAb-ELISA monoclonal enzyme-linked immunosorbent assay (cut-off for both ELISAs determined by considering necropsy and SCT as the gold-standard test), DIC deviance information criterion, na not applicable