Table 3.

Summary of the papers that examined cirrhosis/CLD patients against healthy cohorts. The arrows show the VOC abundance in the CLD group compared to the healthy group in the study design.

Author/Year	Study Design	Analytical Method	VOCs Identified as Significant	Discriminatory Performance
Friedman et al. 1994	24 cirrhotic CLD vs. 24 healthy	GC-MS	Hydrogen-sulphide ↑ Limonene ↑	Not reported
Van den Velde et al. 2008	52 cirrhotic CLD vs. 50 healthy	GC-MS	Acetone ↑ Dimethyl-sulphide ↑ 2-butanone ↑ 2-pentanone ↑ Indole ↓ Dimethyl-selenide ↓	100% sensitivity 70% specificity
Dadamio et al. 2012	35 cirrhotic CLD vs. 49 healthy	GC-MS	Dimethyl-sulphide ↑ Acetone ↑ 2-butanone ↑ 2-pentanone ↑ Indole ↓ Phenol ↓ Dimethyl-selenide ↓ Isoprene ↑ Ethane ↑ Pentane ↑	83% sensitivity 100% specificity
Morisco et al. 2013	12 cirrhotic CLD vs. 14 healthy	PTR-MS	Heptadienol ↑ Methanol ↑ 2-butanone ↑ 3-pentone ↑ 2-octanone ↑ 2-nonanone ↑ Monoterpene ↑ P-cymene ↑	83% sensitivity 86% specificity
Del Rio et al. 2015	31 cirrhotic CLD vs. 30 healthy	PTR-MS	Methanol ↑ 2-butanone ↑ Carbon-sulphide ↑ 2-pentanone ↑ Limonene ↑	97% sensitivity 70% specificity
Pijls et al. 2016	34 cirrhotic CLD vs. 87 non-cirrhotic CLD	GC-MS	Dimethyl-sulphide ↑ Terpene (limonene) ↑ 2-methyl-butanal ↓ Propanoic acid ↑ Octane ↑ Terpenoid ↑ 3-carene ↑ 1-hexadecanol ↓ C16H34 ↓	83% sensitivity 87% specificity
De Vincentis et al. 2016	65 cirrhotic CLD vs. 39 non-cirrhotic CLD	E-nose	Not available	86.2% sensitivity 98.2% specificity
Eng et al. 2015	49 cirrhotic CLD children vs. 55 healthy children	SIFT-MS	1-decene ↑ 1-heptene ↑ 1-octene ↑ 3-methyl-hexane ↑ 1-nonene ↓ (E)-2-nonene ↓ Dimethyl-sulphide ↓	0.97 AUC