Table 2.

ROC analysis of the diagnostic performance of clinical variables in predicting testicular torsion

Variable	Cut-off	AUC	p value	95% CI	Sensitivity	Specificity
Combined Model	-	0.839	< 0.001	0.784–0.893	75.6%	81.5%
GPR	21.8	0.738	< 0.001	0.679–0.796	66.7%	65.0%
WBC	8.6	0.695	< 0.001	0.631–0.760	63.8%	63.0%
SII	596.1	0.649	< 0.001	0.572–0.725	63.0%	62.7%
SIRI	1.13	0.616	0.003	0.538–0.695	58.3%	55.1%
AISI	320.3	0.666	< 0.001	0.576–0.724	63.0%	62.0%
NLR	2.5	0.625	0.001	0.547–0.704	58.3%	62.6%
PLR	114.1	0.606	0.011	0.519–0.692	53.4%	53.6%
MER	3.4	0.633	0.001	0.555–0.711	61.6%	59.7%
NER	31.3	0.638	< 0.001	0.599–0.717	61.1%	60.3%

The data were analysed using a receiver operating characteristic (ROC) curve. The optimal cut-off values were determined based on the highest Youden index. The Combined Model was created using binary logistic regression probabilities, including Age, GPR, and WBC variables. p-values below 0.05 were considered statistically significant

Abbreviations: AUC Area under the curve, CI Confidence interval, GPR Glucose/potassium ratio, WBC White blood cell, SII Systemic inflammation index, SIRI Systemic inflammatory response index, AISI Aggregated index of systemic inflammation, NLR Neutrophil-to-lymphocyte ratio, PLR Platelet-to- lymphocyte ratio, MER Monocyte-to-eosinophil ratio, NER Neutrophil-to-eosinophil ratio