Skip to main content

Table 2 Goodness-of-fit and diagnostic performance of risk models for predicting importation of the Middle East respiratory syndrome (MERS)

From: Predicting the international spread of Middle East respiratory syndrome (MERS)

ID

Model

Number of parameters

AIC1

AUC2

Sensitivity (%)

Specificity (%)

1

Effective distance only

1

464.1

0.95 (0.54,1.00)

100.0 (88.3, 100.0)

79.6 (74.0, 85.2)

2

Effective distance + religion

2

461.2

0.87 (0.46, 1.00)

100.0 (88.3, 100.0)

69.2 (62.8, 75.5)

3

Effective distance + incidence

1

357.2

0.95 (0.54, 1.00)

100.0 (88.3, 100.0)

79.6 (74.0, 85.2)

4

All pieces of information

2

354.7

0.87 (0.46, 1.00)

100.0 (88.3, 100.0)

69.2 (62.8, 75.5)

  1. 95 % confidence intervals (CI) are given in parenthesis. 1. AIC, Akaike information criterion [30]. Note that the data used for parameterizing models 1 and 2 were different from those used for models 3 and 4, and thus, the comparison can be made only between models 1 and 2 and between models 3 and 4, respectively; 2. AUC, area under the curve, derived from the receiver operator characteristic (ROC) curve [31] to predict the risk of importing a MERS case