Abstract
BACKGROUND: Oral mucosal moisture (OMM) measured by using a portable device (Mucus(®), LiFE Co., Ltd., Koshigaya, Saitama, Japan; approval No. 22200BZX00640000) quantifies tongue moisture, yet its diagnostic value for intravascular volume depletion (hypovolemia) in critically ill patients remains unclear. OBJECTIVES: To evaluate the diagnostic accuracy of OMM for hypovolemia in postoperative cardiovascular surgical patients in the intensive care unit (ICU). METHODS: The present single-center, prospective diagnostic accuracy study (Japan, Nov 2024-Sep 2025) enrolled consecutive postoperative cardiovascular surgical adults (≥20 years). OMM (index test) was measured on postoperative day (POD) 2 or on POD 3, if the POD 2 data were unavailable. The reference standard was a composite requiring all of the following conditions: absence of venous congestion (venous excess ultrasound (VExUS) score = 0), at least one low-volume indicator (central venous pressure (CVP) <4 mmHg, small inferior vena cava (IVC) diameter, or positive collapsibility), and a positive left ventricular (LV) "kissing sign," acquired within a two-hour window. The predefined tongue dryness cutoff (<27.9%) and a data-driven Youden threshold were compared. We estimated the sensitivity, specificity, predictive values, likelihood ratios, and area under the receiver-operating characteristic curve with confidence intervals (CIs). Multivariable logistic regression was performed after adjusting for age, Acute Physiology and Chronic Health Evaluation II (APACHE II) score, and body mass index. Separately, the correlations between OMM and CVP, IVC diameter, VExUS score, and intraoperative water balance were examined. RESULTS: Among the 85 eligible patients, 57 were enrolled (age, 66.9 ± 10.7 years; 70.2% were male). The hypovolemia prevalence was 12.3%. At the cutoff Youden index (≤26.1), sensitivity was 0.714 (95% CI 0.290-0.963); specificity, 0.640 (0.492-0.771); positive predictive value (PPV), 0.217 (0.075-0.437); negative predictive value (NPV), 0.941 (0.803-0.993); and AUC, 0.640 (0.476-0.804). At the predefined cutoff (<27.9%), sensitivity was 0.857 (0.421-0.996); specificity, 0.460 (0.318-0.607); PPV, 0.182 (0.070-0.355); and NPV, 0.958 (0.789-0.999). The diagnostic accuracy did not differ between the two thresholds (McNemar p > 0.99). OMM was not independently associated with hypovolemia and showed no significant correlations with CVP, IVC diameter, VExUS score, or intraoperative balance. CONCLUSIONS: In this single-center postoperative cardiac ICU cohort study, OMM showed limited discrimination for hypovolemia and no clear association with intravascular volume. OMM is not supported as a stand-alone screening tool; any interpretation should be coupled with dynamic hemodynamic assessments.