Development of metabolic and inflammatory mediator biomarker phenotyping for early diagnosis and triage of pediatric sepsis

Table 2 Accuracy results of orthogonal partial least squares discriminant analysis models

Age group	Data	Outliers in PICU sepsis cohort, n (%)	R ²Y	Q ²	p Value	Sensitivity; specificity	PPV; NPV	Accuracy	AUROC (SD)
2–17 yr	Metabolites	6 (6 %)	0.69	0.60	3.9×10⁻³¹	0.86; 0.91	0.92; 0.86	0.89	0.96 (0.01)
	Mediators	13 (14 %)	0.58	0.42	4.9×10⁻¹⁷	0.79; 0.77	0.78; 0.78	0.78	0.88 (0.03)
	Combined	7 (7 %)	0.69	0.62	5.9×10⁻³³	0.90; 0.85	0.87; 0.88	0.87	0.95 (0.01)
2–5 yr	Metabolites	3 (8 %)	0.68	0.50	4.4×10⁻¹⁰	0.76; 0.95	0.93; 0.83	0.87	0.95 (0.03)
	Mediators	6 (17 %)	0.67	0.45	1.8×10⁻⁷	0.73; 0.83	0.76; 0.81	0.79	0.91 (0.04)
	Combined	5 (14 %)	0.78	0.65	8.0×10⁻¹⁵	0.87; 0.93	0.90; 0.91	0.90	0.98 (0.02)
6–17 yr	Metabolites	6 (10 %)	0.79	0.68	2.0×10⁻²⁰	0.94; 0.92	0.94; 0.92	0.93	0.97 (0.01)
	Mediators	9 (16 %)	0.67	0.45	4.5×10⁻¹⁰	0.86; 0.76	0.82; 0.80	0.81	0.91 (0.03)
	Combined	5 (9 %)	0.82	0.76	2.7×10⁻²⁵	0.94; 0.92	0.94; 0.92	0.93	0.99 (0.01)

Abbreviations: AUROC area under the receiver operating characteristic curve, NPV negative predictive value, PICU pediatric intensive care unit, PPV positive predictive value, SD standard deviation

ISSN: 1364-8535