Table 3 Performance and presentation of models.

From: Aspiration risk prediction models in patients with nasogastric enteral nutrition: a systematic review and meta-analysis

Author (year)

Model performance

Calibration method

Validation method

Model presentation

 

Internal validation

External validation

 

Jing13

C-index = 0.81[95%CI(0.71,0.89)]

Calibration curve

Bootstrap

–

Nomogram

Peng14

A: ACU = 0.955[95%CI(0.935,0.976)],

Sensitivity = 95.2%, Specificity = 84.2%

B: ACU = 0.891[95%CI(0.790,0.991)],

Sensitivity = 94.7%, Specificity = 85.3%

Calibration curve

Unclear

temporal validation

Nomogram

Guan15

C-index = 0.841[95%CI(0.316,1.366)]

Hosmer– Lemeshow test

Bootstrap

–

Nomogram

Guan16

ACU = 0.976

Decision curve

Bootstrap

–

β coefficient risk scoring formula

Zhang17

A:ACU = 0.895[95%CI(0.822,0.936)],

Sensitivity = 82.5%, Specificity = 73.6%

A:ACU = 0.902[95%CI(0.842,0.975)],

Sensitivity = 80.6%, Specificity = 75.8%

–

Unclear

–

Nomogram, CART

Hou18

A:ACU = 0.926[95%CI(0.889,0.968)],

Sensitivity = 88.4%, Specificity = 85.2%

B: ACU = 0.902[95%CI(0.828,0.986)],

Sensitivity = 91.9%, Specificity = 81.8%

Calibration curve

Unclear

temporal validation

unclear

Zhang19

B:ACU = 0.904, Sensitivity = 93.8%,

Specificity = 77.8%

ACU = 0.843, Sensitivity = 66.7%,

Specificity = 84.7%

ACU = 0.964, Sensitivity = 93.8%,

Specificity = 83.1%

ACU = 0.984, Sensitivity = 89.6%,

Specificity = 96.8%

ACU = 0.992, Sensitivity = 89.6%,

Specificity = 97.4%

ACU = 0.903, Sensitivity = 72.9%,

Specificity = 86.8%

Calibration curve

Bootstrap

temporal validation

Nomogram

Yu20

C-index = 0.857[95%CI(0.739,0.944)]

Hosmer–Lemeshow test

Bootstrap

–

Nomogram

Xie21

A:ACU = 0.812[95%CI(0.787,0.837)],

Sensitivity = 78.4%, Specificity = 81.3%

B: ACU = 0.885[95%CI(0.816,0.947)],

Sensitivity = 85.4%, Specificity = 88.2%

Hosmer–Lemeshow test

Unclear

temporal validation

mathematical formula

Shi22

ACU = 0.809[95%CI(0757,0.861)]

Hosmer–Lemeshow test

–

–

mathematical formula

Sun23

A:ACU = 0.93[95%CI(091,0.96)],

Sensitivity = 90.9%, Specificity = 88.6%

ACU = 0.96[95%CI(093,0.97)],

Sensitivity = 88.3%, Specificity = 96.2%

B: ACU = 0.91[95%CI(089,0.98)],

Sensitivity = 89.2%, Specificity = 89.7%

ACU = 0.920[95%CI(0.910,0.950)],

Sensitivity = 64.9%, Specificity = 96.6%

–

Bootstrap

temporal validation

Nomogra, CART

  1. A, development cohort; B, validation cohort.AUC, area under the curve. We considered AUC = 0.5–0.7 as poor discrimination, 0.7–0.8 as moderate discrimination, 0.8–0.9 as good discrimination, and 0.9–1.0 as excellent discrimination.
Back to article page