Defining a Disease-Specific Antral Area Threshold for Confirming Empty Stomach in Chronic Hepatitis B Patients

1. Introduction

Identification of perioperative risk is a cornerstone of safe peri-operation anesthesia management regarding preoperative fasting and therapeutic drug aspiration [1]. Aspiration is a serious complication which can lead to severe morbidity and mortality during perioperative anesthesia practice [2,3]. At present clinical work, anesthesiologists only assess the risk of aspiration based on the patients’ medical history and compliance with fasting guideline during the period of anesthesia. But in fact, fasting duration is not the gold standard to diagnose the risk of aspiration and recommended fasting intervals are only reliable in healthy patients [5]. Hence, the quantification of the gastric status may stratify individual aspiration risk, which is of a major importance for anesthesiologists.

In recent years, point-of-care ultrasound (POCUS) is widely applied in clinical practice due to the valuable advantages it offers, such as consistency, efficiency, reliability and repeatability. Point-of-care gastric ultrasonography is a noninvasive tool that provide both qualitative (empty, solid food, clear fluids) and quantitative information (gastric volume estimation) about gastric contents [6,7,8,9,10].

In the non-pregnant adult, there is a significant positive linear correlation between cross-sectional area (CSA) and gastric fluid volume with the Pearson correlation coefficients ranging 0.6-0.91. In addition, the coefficient is larger in the right lateral decubitus (RLD) vs other positions [11]. The CSA 340 mm² cut-off value in the supine position for detecting a 'risk stomach', defined by a liquid volume > 0.8mL/kg, has a high consistency [12]. Hence, alternatively, a patient at low risk of aspiration could be defined by a stomach with an antral cross-sectional area ≤ 340 mm² and with no visual gastric contents.

However, applying this uniform cut-off value to patients with chronic liver diseases, particularly chronic hepatitis B (CHB), presents a significant clinical dilemma. CHB is a systemic viral infection that triggers not only hepatocyte damage but also profound extrahepatic manifestations, heavily impacting gastrointestinal motility [13]. The deterioration of liver function in CHB patients is frequently accompanied by portal hypertensive gastropathy (PHG), altered gut-liver axis signaling, and autonomic dysfunction. These pathophysiological cascade effects lead to persistent mucosal injury, reduced pyloric sphincter tension, and substantially delayed gastric emptying [14].

Because of these intrinsic pathological alterations, the baseline fasting CSA in CHB patients is anatomically and physiologically larger than that of healthy individuals. Therefore, strictly enforcing the traditional 340 mm² threshold for CHB patients inevitably leads to an unacceptably high rate of false-positive "full stomach" diagnoses. This overestimation of aspiration risk forces anesthesiologists into unwarranted defensive interventions, such as unnecessary cancellations or postponements of elective surgeries, or the mandatory use of rapid sequence induction (RSI), thereby imposing significant psychological stress on patients and tremendous financial burdens on healthcare systems. To date, no literature has systematically investigated or established a disease-specific sonographic threshold for this vulnerable population.

To address this critical knowledge gap, we conducted this prospective study on patients with chronic hepatitis B to calculate the cut-off value of CSA measured in the RLD for the diagnosis of an 'empty' stomach (gastric fluid volume ≤0.8mL/kg) in patients with chronic hepatitis B.

2. Materials and Methods

Study Design

The study was conducted in accordance with the Declaration of Helsinki. Ethical approval for this study (Ethical Committee [2016] 11) was provided by Tianjin Second People’s Hospital Medical Ethics Committee, Tianjin, China (Chairperson Guo Yang) on 1^st January 2016. This prospective observational cohort study was registered in the Chinese Clinical Trial Registry (ChiCTR1800016698) on June 18, 2018 and conducted between January to December 2018 in Tianjin Second People’s Hospital. Initially, an unregistered exploratory pilot phase was conducted between January and early June 2018 to evaluate the general application of point-of-care gastric ultrasound. Upon serendipitously observing an enlarged baseline antral area specifically in CHB patients during this phase, we formalized the study protocol and officially registered the trial on June 18, 2018. To maximize statistical power, data from both the initial exploratory phase and the subsequent prospective phase (June to December 2018) were included in the final analysis. The methodology followed the recommendations of the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) statement [15]. To ensure measurement consistency and eliminate inter-observer variability, all point-of-care gastric ultrasound (POCUS) assessments were strictly performed by a single, dedicated physician, with experience of at least 100 gastric ultrasound examinations in chronic hepatitis B patients. Patient recruitment was sequentially aligned with the clinical schedule of this designated sonographer to maintain methodological uniformity. Written informed consent was obtained from all participating patients prior to enrollment and any study-related procedures.

Study Population and Eligibility Criteria

Chronic hepatitis B patients were recruited to participate in the study if they met the following clinical inclusion criteria: aged between 18-75 years old, American Society of Anesthesiologists physical status (ASA) from I to Ⅲ, body mass index (BMI) < 35 kg/m² (with an absolute weight of 45–100 kg and height >150 cm), fasting duration time more than 8 hours and stop gastrointestinal motility drug for three days at least before examination. Exclusion criteria were refusal to take part in the study: a history of upper gastrointestinal surgery, pregnancy, gastrointestinal hemorrhage within one month, gastroesophageal reflux disease and severe mental illness.

Sample Size Determination

The sample size was prospectively calculated based on the primary objective of evaluating the diagnostic performance of the antral CSA for detecting an empty stomach using receiver operating characteristic (ROC) curve analysis. Assuming an expected area under the ROC curve (AUC) of 0.80, a minimum total sample size of 200 patients was required to construct the ROC curve with a statistical power of 0.85 (1-β) at a two-sided significance level (α) of 0.05. Anticipating potential protocol violations, dropouts, and suboptimal ultrasound imaging quality (e.g., obscuration by gastrointestinal gas), the target enrollment was conservatively increased by approximately 25%, resulting in a final planned recruitment of 223 patients.

Ultrasound Protocol

A bedside ultrasound device (ACUSON P300, SIEMENS, German) fitted with an abdominal probe (3 to 5 MHz) was used for the ultrasound examinations. Ultrasound scans were performed for chronic hepatitis B patients arriving in the gastroscopy room. The patients lay in the RLD. Following established standardized protocols [10], the transducer was placed in the sagittal or parasagittal plane in the epigastric region. The gastric antrum was distinctly identified using the left lobe of the liver and the abdominal aorta (or superior mesenteric artery) as constant internal anatomical landmarks (Figure 1).

An ultrasound assessment of the antrum was performed by the same operator. The maximal anteroposterior diameter (D1) and longitudinal diameter (D2) of a single section of the gastric antrum were determined and the antral cross-sectional area was calculated using the formula, CSA=(π×D1×D2)/4. After normal gastroscopy examination, the fluid in the stomach was collected by the gastroscopy and the volume was recorded. Any failure for the qualitative assessment of CSA was noted and the patients were excluded from the analysis.

Endoscopic Aspiration (The Reference Standard)

Immediately following the sonographic assessment, patients underwent their scheduled esophagogastroduodenoscopy (EGD). Under direct endoscopic visualization, all residual gastric fluid was completely aspirated, and its total volume was precisely measured and recorded. This endoscopically aspirated fluid volume served as the objective reference standard ("gold standard") to definitively categorize patients as having either an "empty stomach" (defined as gastric fluid volume < 0.8mL/kg or a "risk stomach" (volume > 0.8 mL/kg).

Clinical and Biochemical Data Collection

Comprehensive baseline demographic and clinical data were collected for all enrolled patients, including age, sex, BMI, precise fasting duration, and the nature of the last ingested meal (e.g., clear fluids, light meal, or solid food). Detailed medical comorbidities, specifically diabetes mellitus, hypertension, and coronary heart disease, were also documented. Furthermore, concurrent preoperative laboratory parameters, predominantly liver function indices and coagulation profiles, were extracted from the electronic medical records. These included alanine aminotransferase (ALT), aspartate aminotransferase (AST), CHE, total protein (TP), ALB, total bilirubin (TBIL), direct bilirubin (DBIL), indirect bilirubin (IBIL), prothrombin time (PT), and blood ammonia levels.

Statistical Analysis

All statistical analyses were performed using Stata version 18.0 (StataCorp LLC, College Station, TX, USA). The normality of continuous variables was evaluated using the Shapiro-Wilk test. Normally distributed continuous variables were expressed as mean ± standard deviation (SD) and compared using the independent Student’s t-test or one-way analysis of variance (ANOVA) followed by the Bonferroni post hoc test, as appropriate. Non-normally distributed data were expressed as median (interquartile range, IQR) and analyzed using the Mann-Whitney U test. Categorical variables were presented as frequencies (percentages) and compared using the Pearson Chi-square test or Fisher’s exact test. The diagnostic performance of the antral CSA in the RLD position for detecting a safe, empty stomach (defined as gastric fluid volume < 0.8mL/kg) was comprehensively evaluated using ROC curve analysis. The overall predictive accuracy was quantified by the AUC with its 95% confidence interval (CI). The optimal cut-off value for the antral CSA was strictly determined by maximizing the Youden's index (Sensitivity + Specificity - 1), while ensuring a clinically acceptable high sensitivity threshold to prevent the misclassification of high-risk patients.

To explore independent pathophysiological factors affecting the baseline CSA, a multiple linear regression model was constructed. Variables demonstrating a potential association in the univariate correlation analysis (Pearson or Spearman, depending on the data distribution) were subsequently included in the multivariate model. Multicollinearity among independent variables was rigorously assessed using the Variance Inflation Factor (VIF), with a VIF > 5 indicating significant collinearity. The overall fit and explanatory power of the regression model were reported using the adjusted coefficient of determination (Adjusted R²). All statistical tests were two-sided, and a P-value < 0.05 was considered statistically significant.

3. Results

3.1. Patient Enrollment and Baseline Characteristics

223 patients met the inclusion criteria. Figure 2 (Flow Chart) shows the recruitment and loss of patients in the trial. The baseline demographic and clinical characteristics of the study population are summarized in Table 1. No chronic hepatitis B patient had hemodynamic issues while lying in the right lateral decubitus position. Nineteen patients had diabetes mellitus. 216 patients had solid food for the last time, nature of last food was fluid for 7 patients. Fasting duration time more than 8 hours.

9 patients were excluded prior to analysis: 7 patients (2.7%) were excluded because the antrum was not clearly identified during the ultrasound assessment, 1 patient (0.4%) was excluded because of overweight and 1 patient (0.4%) was excluded because solid food was in the stomach. Consequently, 214 patients were included in the final statistical analysis.

3.2. Diagnostic Performance of Antral CSA

The frequencies of the antral cross-sectional area measured in the right lateral decubitus position in the whole population studied is set out in Figure 3. It is worth noting that the cross-sectional area of the gastric antrum in the right lateral position of patients with hepatitis B is much larger than the ultrasound-measured cross-sectional area of the gastric antrum in normal individuals, which is 340mm² (Figure 3).

Based on the endoscopic reference standard, the diagnostic performance of the antral CSA in the RLD position for predicting a safe, empty stomach (gastric fluid volume < 0.8 mL/kg) was evaluated using ROC curve analysis. The overall AUC was 0.763 (95% CI: 0.685-0.842), P < 0.001 (Figure 4).

Guided by the maximization of the Youden's index (Max Youden's index = 0.471) while prioritizing a conservative threshold to avoid false-negative aspiration risks, the optimal cut-off value of the antral CSA was determined to be 835 mm². At this cut-off value, the diagnostic measurement yielded a positive predictive value (PPV) of 94.4%, and a negative predictive value (NPV) of 36.3% (Table 2). The overall diagnostic coincidence rate with the endoscopic gold standard was 66.1%.

3.3. Univariate Analysis of Factors Affecting Baseline CSA

The relationships between baseline clinical characteristics, medical history, and the antral CSA in the RLD position are summarized in Table 3 and Table 4. Univariate analysis revealed that CSA in the RLD was significantly higher in males than in females, the difference was statistically significant (median, 8.59 (7.00, 10.77) vs. 7.56 (6.19, 9.56)cm²; Z = 2.227, P = 0.026) (Table 3), and CSA in different type of hepatitis, diabetes mellitus, coronary heart disease had no statistical significance (P > 0.05). However, the antral CSA was significantly larger in the hypertensive group compared to the non-hypertensive group (median, 9.49 (7.42, 10.98) vs. 8.03 (6.47, 10.27) cm²; Z = -2.239, P = 0.025).

3.4. Multivariate Linear Regression Analysis

To identify independent predictors of an enlarged baseline antral CSA, a multivariate linear regression model was constructed (Table 5). A multivariate linear regression model was established by taking the CSA in RLD as the dependent variable and the statistically significant index (included sex, CHE, ALB, TBIL, DBIL, Blood ammonia) in univariate analysis as the independent variable. The results showed that CHE and ALB had statistical significance in the model. For each additional unit of CHE, CSA in RLD will decrease by 0.001 units. For each additional unit of ALB, CSA area in the RLD will increase by 0.039 units.

4. Discussion

In this prospective observational study, we established a disease-specific sonographic threshold for the RLD antral CSA to confirm an empty stomach (gastric fluid volume < 0.8 mL/kg) in patients with CHB. Our principal finding demonstrated that an antral CSA cut-off of 835 mm² yielded a high PPV of 94.4% for predicting a safe gastric state, offering a reliable point-of-care tool for anesthesiologists to stratify perioperative aspiration risk in this specific population. Identification of aspiration risk is a footstone of safe anesthesia management regarding preoperative fasting and therapeutic drug aspiration [4]. Inaccurate risk assessment is often a leading cause of aspiration events [16]. Although the aspiration risk is highest in emergency situations, it occasionally occurs in patients who have followed fasting guidelines and are considered at low risk. This baseline risk is approximately 1:4,000 [17]. Perlas [18] found that CSA can accurately reflect the gastric contents. The measurement of CSA was used to determine the patient's gastric content state by point-of care ultrasound, which is a clinical diagnostic method with high accuracy and easy to perform [12,20]. A cut-off value of 340 mm² in the supine position was reported for the diagnosis of risk stomach (no solid content, gastric fluid volume < 0.8mL/kg, or both), with high sensitivity and high negative predictive value [19]. Some literatures reported that CSA measured in the RLD correlated most strongly with gastric volume in non-pregnant adult patients [12,18,21,22,23,24,25]. In the early stage of our study, there was a statistically significant difference in CSA between healthy volunteers and chronic hepatitis B patients in the supine position. This study found that if the CSA of 340 mm² is still used as a cut-off value to evaluate the aspiration risk in chronic hepatitis B patients, it may lead clinicians to over-assess the aspiration risk. Such overestimations force anesthesiologists into unwarranted defensive interventions, such as the unnecessary use of rapid sequence induction (RSI) or unwarranted surgical postponements, thereby increasing the psychological and financial burden on patients. Therefore, this study focused on the relationship between the cut-off value of CSA in the RLD and risk of aspiration about chronic hepatitis B patients. We believed that it was relevant to assess whether ultrasonographic measurement of CSA is able to detect the risk stomach.

In the present study, the overall success rate for ultrasound examination of the antrum was 96.87%. Nevertheless, among the 9 excluded ultrasound examinations, only 8 patients were not possible to identify the antrum in the right lateral decubitus position. This result corroborates those of preliminary studies performed on volunteers, reporting that a complete cross-sectional view of the antrum was obtained in 100% of 18 volunteers in the study by Perlas [17] et al and 98.5% of 22 volunteers in the study by Bouvet et al [18]. The results of the current study performed on a large population of patients allowed the conclusion that the measurement of antral CSA during the preoperative period could be used in clinical practice. In our study, the cut-off value of CSA 835 mm² in RLD for the diagnosis of risk stomach is associated with a sensitivity of 86.1%, a specificity of 61.1%, a positive predictive value of 94.4%, and a coincidence rate of 66.1%; therefore, it could be proposed for clinical practice. This cut-off value is more suitable for the diagnosis of risk stomach in chronic hepatitis B patients.

Hepatitis B virus (HBV) is not only hepatotropic, but also pantropic virus, which can be found in many tissues, responsible for extrahepatic clinical manifestations [26,27,28,29], and gastric mucosal injury is one of these manifestations. HBV localizes within the gastric mucosa, and the resulting antigen-antibody complexes precipitate microvascular endothelial inflammation, profoundly impairing the gastric mucosal barrier [13]. Concurrently, the deterioration of hepatic function exacerbates autonomic neuropathy and gastrointestinal dysmotility, ultimately leading to reduced pyloric sphincter tension and delayed gastric emptying [14]. This unique disease-induced anatomical distension explains why the traditional diagnostic cut-offs for healthy individuals are clinically obsolete for CHB patients. With liver function deterioration, the expression of HBV antigen in gastric mucosa and the degree of gastric mucosal lesion also increased, and the gastrointestinal symptoms were more prominent, as well as clinical manifestations of acid regurgitation, abdominal distension and so on [30]. Relevant studies have shown that HBV can lead to gastric tissue lesion and is also an important factor leading to gastric disease in infected patients [31,32,33,34]: Once a cell is infected with HBV, it may be directly damaged by the cell-killing effect [35]; and high positive rate of hepatitis B virus antigen results in the formation of high titers of antigen-antibody complexes. The deposition of the antigen-antibody complexes in microvascular endothelial cells can cause inflammatory reaction in patients’ gastric tissue, the inflammatory damage can weaken and destroy the gastric mucosal barrier [13]. As a consequence, long-term gastrointestinal dysfunction and damage to the gastric mucosa will have an effect on the patients' pyloric sphincter tension and increase CSA of the empty gastric antrum in patients with chronic hepatitis B [36].

At the same time, we analyzed the related factors that may affect CSA in the RLD. In our study, the CSA in the right lateral decubitus position was significantly higher in males than in females. This result was different from a previous clinical study [37] that showed no significant relationship between sex and CSA. Meanwhile, the present study found that CHE and ALB were independently associated with CSA. The reason could be that these liver function indexes can independently reflect the degree of liver damage in chronic hepatitis B patients, indirectly indicating the gastrointestinal dysfunction in these patients. Diabetes can cause delayed gastric emptying and reduce gastrointestinal function [38]. But in our study, there was no significant relationship between CSA and diabetes mellitus. Investigating its cause, the sample size of diabetes mellitus patients was too small to be accurately evaluated. Interestingly, our univariate analysis revealed that concurrent hypertension is associated with a significantly larger baseline antral CSA. Pathophysiologically, this may be attributed to the chronic systemic autonomic neuropathy and elevated sympathetic tone frequently observed in hypertensive patients, which can reflexively inhibit gastrointestinal motility and subsequently delay gastric emptying [39]. We speculate that the impact of hypertension on the antral CSA might be confounded by other predominant factors, such as age or the severity of the underlying hepatic dysfunction.

Our study also has several limitations. All the antrum ultrasound examinations were completed by only one physician. To minimize confounding factors, the ultrasound examinations should be performed by two operators double-blinded to each other’s results. It would be of interest in further studies to assess the part of interobserver variability in the estimation of the reliability and reproducibility of this method. Furthermore, many participants presented with longer fasting intervals than the minimum required by current guideline, and the findings may not apply to patients who adhere more closely to minimum guidelines before gastroscopy examination. At the same time, we didn’t stipulate the calories and species of food taken on the previous day. Although this may limit the generalizability of our results, it strengthens our suggestion on CSA cut-off value for fasted chronic hepatitis B patients.

5. Conclusions

For most patients, a preoperative ultrasonographic measurement of antral CSA was feasible. A customized cut-off value of 835 mm² provides excellent diagnostic confidence for confirming an empty stomach, thus preventing unwarranted defensive anesthetic interventions. We advocate for the routine incorporation of disease-specific POCUS thresholds into the perioperative airway management of patients with chronic liver diseases. Further studies are required in chronic hepatitis B patients to confirm our results and to define more precisely the usefulness of ultrasonographic measurement of CSA in preventive strategies of pulmonary aspiration of gastric contents.