Skip to main content
. 2023 Mar 24;94(3):1125–1135. doi: 10.1038/s41390-023-02558-6

Table 2.

Model characteristics, quality, and validation.

First author (year) Model characteristics Model performance Model estimation
Country Sex Study type Study setting Number of predictors (n) Sample size Method of discrimination assessed Method of calibration assessed Method of validation assessed
Kobayashi T (2006)21 Japan Both Cohort study 13 medical institutions 7

IVIG Resistance: 148

IVIG Responder: 528

Model AUC: 0.85 (95% CI: 0.81–0.88)

external validation AUC: 0.90 (95% CI: 0.85–0.96)

 HL and Calibration plot External validation
Egami K (2006)25 Japan Both Cohort study Database of medical institutions 5

IVIG Resistance: 41

IVIG Responder: 279

 model AUC: 0.79 (95% CI: 0.73–0.86) HL No
Yang S (2019)23 China Both Cohort study Multi-center/ hospital 5

IVIG Resistance: 22

IVIG Responder: 90

model AUC: 0.77 (95% CI: 0.72–0.83)

internal validation AUC: 0.77 (95% CI: 0.71–0.82)

external validation AUC: 0.69 (95% CI: 0.58–0.81)

external validation AUC: 0.63 (95% CI: 0.53–0.72)

 No

Internal validation

two external validation
Wu S (2020)28 China Both Cohort study Hospital 5

IVIG Resistance: 31

IVIG Responder: 246

 model AUC: 0.750 (95% CI: 0.666–0.834) No No
Piram M (2020)19 France Both

Predominantly prospective

cohort study

 National clinical and biological repository 4

IVIG Resistance: 92

IVIG Responder: 323

 Model AUC: 0.725 (sensitivity, 77%; specificity, 60%) No No
Wu S (2019)32 China Both cohort study Double centers 4

IVIG Resistance: 23

IVIG Responder: 259

Model AUC: 0.891 (95% CI: 0.837–0.945)

external validation (sensitivity, 70.0%; specificity, 75.1%)

 No External validation
Fu PP (2013)33 China Both cohort study Hospital 5

IVIG Resistance: 211

IVIG Responder: 966

 Model AUC: 0.672 (95% CI: 0.631–0.712) HL No
Gámez-González LB (2018)26 Japan Both cohort study Medical center 5

IVIG Resistance: 101

IVIG Responder: 318

 No AUC (sensitivity:76.2%; specificity:64.8%) No No
Tan XH (2019)18 China Both cohort study hospital 8

IVIG Resistance: 348

IVIG Responder: 4929

AUC: 0.74 (sensitivity:76%; specificity:59%)

internal validation AUC: 0.72 (range, 0.65–0.80)

 HL internal validation
Bar-Meir M (2018)20 Israel Both cohort study 9 medical centers 2

IVIG Resistance: 42

IVIG Responder: 270

AUC: 0.7 (95% CI: 0.6–0.8)

external validation AUC: 0.69 (95% CI: 0.59–0.8)

 HL external validation
Wang T (2020)24 China Both cohort study hospital 7

IVIG Resistance: 124

IVIG Responder: 520

model AUC: 0.7423 (accuracy: 0.8844; sensitivity: 0.3043; specificity: 0.9919)

external validation AUC: not reported

 No external validation
Tang Y (2016)27 China Both cohort study hospital 5

IVIG Resistance:46

IVIG Responder: 864

 model AUC: 0.77 (95% CI: 0.71–0.82) HL No
Hua W (2017)34 China Both cohort study hospital

IVIGRKD model:6;

IVIGRKD ≤ 6 months old model: 4

IVIG Resistance: 380

IVIG Responder: 1746

model AUC: 0.685 (95% CI: 0.652–0.717)

patients ≤ 6 months model AUC: 0.746 (95% CI: 0.665–0.827)

 HL No
Sano T (2016)35 Japan both retrospective cohort study seven institutions 3

IVIG Resistance: 22

IVIG Responder: 90

 No (sensitivity: 77%, specificity: 86%) No No
Tremoulet AH (2008)30 America both cohort study 2 clinical sites 4

IVIG Resistance: 60

IVIG Responder: 302

 No (sensitivity: 73.3%; specificity: 61.9%) No No
Lin. M. T (2016)22 Taiwan both cohort study hospital 3

IVIG Resistance: 22

IVIG Responder: 159

AUC: 0.86 (95% CI: 0.76–0.97)

external validation (sensitivity: 71.4% specificity: 81.0%)

 No external validation
Sato S (2013)17 Japan both cohort study hospital 3

IVIG Resistance: 21

IVIG Responder: 84

 No (sensitivity: 85.7%; specificity: 77.4%) No No

IVIG intravenous immunoglobulin.

H-L Hosmer–Lemeshow test.

AUC area under the curve.

95% CI 95% confidence interval.