Skip to main content

Table 5 Sensitivity and specificity (95% credible interval) estimated from each model for each diagnostic test

From: Estimating sensitivity and specificity of diagnostic tests using latent class models that account for conditional dependence between tests: a simulation study

Model Name | Assumed dependence structure

Model 0

CIndep

Model 1

CDP among two tests

Model 2

CDP among two tests

Model 3

CDP

Model 4

CDN

Model 5

CDPN

Effect Type Used

NA

Fixed

Fixed

Random

Random

Random

Test

Measure

      

Culture

Sea

61(53–69)

62(54–69)

62(54–69)

60(52–69)

74(59–97)

67(57–79)

Spb

100c

100c

100c

100c

100c

100c

Serology: IHA

Sea

73(66–80)

73(66–79)

73(67–78)

70(63–76)

72(65–79)

69(62–75)

Spb

87(79–93)

86(79–93)

86(79–92)

84(75–92)

75(61–88)

76(66–84)

Serology: Igm ICP

Sea

81(75–86)

80(74–85)

80(74–86)

77(71–83)

80(72–86)

76(69–82)

Spb

65(56–74)

64(55–74)

65(56–73)

62(53–72)

56(45–67)

55(46–64)

Serology: IgG ICT

Sea

91(86–95)

91(86–94)

90(86–94)

88(82–92)

89(84–94)

87(81–92)

Spb

76(67–85)

75(66–84)

75(66–84)

74(64–85)

62(48–77)

65(54–74)

Serology: ELISA

Sea

77(70–84)

78(70–84)

78(71–84)

75(68–78)

82(74–88)

80(72–86)

Spb

97(93–100)

98(94–100)

97(93–100)

97(92–100)

88(72–99)

95(82–100)

  1. Values shown are mean estimates with 95% credible intervals
  2. aSe = sensitivity, bSp = specificity, cSpecificity assumed perfect. Models 0–4 were considered in the original work of Limmathurotsakul et al. while Model 5 is the additional analysis considered in this paper