Skip to main content

Table 3 Sensitivity and specificity of the univariate and multivariate classification, respectively

From: Evaluation of the biomarker candidate MFAP4 for non-invasive assessment of hepatic fibrosis in hepatitis C patients

  Multivariate modela Univariate model
  Unrestricted Lower limit of 80 % sensitivityb,c Unrestricted Lower limit of 80 % sensitivityb,c
Sensitivityc estimate [95 % CI] 88.8 % [82.2 %; 93.6 %] 88.8 % [82.2 %; 93.6 %] 73.1 % [64.8 %; 80.4 %] 86.6 % [79.6 %; 91.8 %]
Specificityd estimate [95 % CI] 63.2 % [58.4 %; 67.9 %] 63.2 % [58.4 %; 67.9 %] 75.0 % [70.5 %; 79.1 %] 54.9 % [49.9 %; 59.8 %]
Leave-one-out cross validation e
Sensitivityc estimate [95 % CI] 80.6 % [72.9 %; 86.9 %] 81.3 % [73.7 %; 87.5 %] 71.6 % [63.2 %; 79.1 %] 85.8 % [78.7 %; 91.2 %]
Specificityd estimate [95 % CI] 61.5 % [56.6 %; 66.3 %] 61.5 % [56.6 %; 66.3 %] 75.0 % [70.5 %; 79.1 %] 54.9 % [49.9 %; 59.8 %]
  1. aA multivariate logistic regression model considering age and gender besides MFAP4 serum levels was derived and the respective Youden optimal cutoffs were determined
  2. bCutoff optimization was restricted to a minimum sensitivity of 80 % accounting for the importance of the identification of high fibrosis stages
  3. cSensitivity was defined as probability of classifying stages F3 and F4 correctly
  4. dSpecificity was defined as probability of classifying stages F0 to F2 correctly
  5. eAs sensitivity and specificity values are prone to overoptimism in the analysis of the complete data set leave-one-out cross validation was performed to obtain unbiased estimates