Periappendiceal fat-stranding models for discriminating between complicated and uncomplicated acute appendicitis: a diagnostic and validation study

Table 4 Diagnostics for models used for discriminating complicated appendicitis

Models	Model odds ratio (per score increase)	P value	ROC c statistics	IDI (%)	P value	Hosmer–Lemeshow statistics (χ²)	P value
1. Bröker et al. [32]	4.18 (2.80–6.23)	< 0.0001*	0.778	Ref	Ref	0.467 (3 groups)	0.4942
2. Imaoka et al. [28]	3.65 (2.58–5.15)	< 0.0001*	0.800	0.919	0.7313	7.556 (3 groups)	0.0060*
3. Khan et al. [16]	1.92 (1.49–2.47)	< 0.0001*	0.694	− 9.450	< 0.0001*	3.200 (4 groups)	0.2019
4. TH Kim et al. [33]	3.61 (2.51–5.18)	< 0.0001*	0.777	1.811	0.5036	2.382 (4 groups)	0.3039
5. Kang et al. [34]	1.76 (1.51–2.09)	< 0.0001*	0.772	0.596	0.8310	2.767 (6 groups)	0.5975
6. Atema et al. [35]	1.44 (1.31–1.58)	< 0.0001*	0.826	5.916	0.0248*	7.289 (9 groups)	0.3995
7. Avanesov et al. [36]	2.29 (1.85–2.85)	< 0.0001*	0.806	3.511	0.2678	3.865 (5 groups)	0.2764
8. HY Kim et al. [37]	5.15 (3.44–7.71)	< 0.0001*	0.838	13.816	0.0006*	1.163 (4 groups)	0.5591
9. Lin (our model 1)	2.14 (1.79–2.56)	< 0.0001*	0.878	18.292	< 0.0001*	13.315 (10 groups)	0.1015
10. Lin (our model 2)	2.08 (1.75–2.46)	< 0.0001*	0.879	18.292	< 0.0001*	8.150 (8 groups)	0.2273

IDI integrated discriminatory improvement, ROC receiver of operating characteristics curve

ISSN: 1749-7922