Introduction

Hitherto, sepsis remains the main cause of mortality in intensive care units (ICUs) [1, 2], and shock accounts for one of the highest rates of mortality, approximately 50%, in many countries [3]. Therefore, early detection and management is of chief importance. The gastrointestinal tract has long been hypothesized to play an integral role in the occurrence and development of sepsis [4]. During septic shock, the body responds by redistributing blood to vital organs, such as the heart, brain, and kidneys. However, the gastrointestinal tract is the first organ to be affected by hypoperfusion, which is considered a common indicator of the development of multiple organ failure [5, 6].

Although vasoactive drugs such as epinephrine and norepinephrine increase the perfusion pressure as well as cardiac output, they are capable of diverting blood flow away from the mesenteric circulation [5, 7], leading to the destruction of the intestinal mucosal barrier, release of inflammatory factors, intestinal epithelial hyperpermeability, bacterial translocation, and multiple organ dysfunction via the circulation [8–10]. Although the underlying mechanism of total organ failure is not entirely understood, it can be avoided, since early diagnosis and management of sepsis have been associated with improved outcomes and reduced mortality [11].

The process of early detection of sepsis may be straightforward; despite this, the early stage of sepsis remains challenging for patients [11]. Routine assessments of splanchnic perfusion such as mucosal perfusion, mesenteric blood flow, mucosal oxygenation, intramucosal pH, gastric and arterial partial pressure of carbon dioxide, intestinal permeability, and lactate production can be used as preventative steps, these tools are not useful once sepsis has already developed [12]. Therefore, there is an urgent need for an accurate, reproducible, radiation-free, rapid, and non-invasive tool to provide information about intestinal perfusion without compromising the treatment of patients with sepsis [13–15].

The use of critical care ultrasound has increased in ICUs in recent years. It has been found to be useful in identifying valuable information to improve medical decision-making, especially in patients with septic shock [16]. Ultrasound can reliably assess arterial and venous blood flow [17, 18]. The SMA is the primary artery supplying the small intestine [19]. In the fasting state, approximately 25% of the cardiac output depends on the splanchnic vascular area. Postprandially, blood flow into SMA is doubled or even tripled [20]. Although effective therapy is still lacking, the therapeutic approach that is selected considerably influences mortality risk. Therefore, in the present study we hypothesized that SMA ultrasound is an effective technique to predict mortality in patients with sepsis.

Methods

Results

Among 72 patients with sepsis, there were 42 (58.3%) non-survivors and 30 (41.7%) survivors. The mean ages of the survivors and non-survivors were 71.07 ± 7.52 and 71.45 ± 6.89 years, respectively (P = 0.822). The mean APACHE Ⅱ and SOFA scores were higher in the non-survivors than in the survivors (APACHE II, 22.67 ± 4.28 vs. 17.43 ± 3.17 and SOFA score, 10.62 ± 1.67 vs. 7.83 ± 2.29, respectively; P < 0.001 for both). The lactate level showed significant difference between non-survivors and survivors (3.79 ± 1.05 mmol/L and 2.74 ± 0.92 mmol/L, respectively; P < 0.001; Table 1). The non-survivors showed lower PSV (78.92 ± 4.75 cm/s vs. 100.8 ± 7.10 cm/s), lower EDV (26.09 ± 1.51 cm/s vs. 29.55 ± 1.41 cm/s), and lower RI (0.67 ± 0.03 vs. 0.71 ± 0.02) than the survivors (P < 0.001 for all). However, no significant differences found in the gender, CCI score, original infection sites, occurrence of septic shock, platelet count, C-reactive protein, and albumin level between both the groups. Compared with survivors, non-survivors experienced higher dose of norepinephrine (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of septic patients (n = 72).

https://doi.org/10.1371/journal.pone.0273377.t001

The statistically significant variables included APACHE Ⅱ score, SOFA score, lactate, low PSV, EDV, and RI which evidenced to be the potential risk factors for mortality by the univariate analysis (P < 0.05) were used to establish a stepwise multivariate logistic regression model. However, only PSV was included in the model. Patients with higher PSV (100.8 cm/s) have a 70.5% lower chance of dying (OR = 0.295; 95% CI: 0.094–0.925) (survivors) than patients with lower PSV (78.92 cm/s) (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Univariate and multivariate logistic analysis.

https://doi.org/10.1371/journal.pone.0273377.t002

The ROC analysis showed that the AUC of PSV (0.99) was the largest among variables of the APACHE Ⅱ score, SOFA score, lactate level, and EDV. The sensitivity, and specificity of PSV were 0.99, and 0.96, respectively (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. The area under the receiver operating characteristic (ROC) curve (AUC).

(APACHE Ⅱ: Acute Physiology and Chronic Health Evaluation Ⅱ; SOFA: Sepsis-related Organ Failure Assessment; Lac: Lactate; PSV: Peak systolic velocity; EDV: End-diastolic velocity; RI: Resistive index).

https://doi.org/10.1371/journal.pone.0273377.g001

Discussion

The present results demonstrate that low PSV was an independent risk factor for mortality, and the results of the ROC curves analyses showed that the PSV value had a good diagnostic performance in predicting mortality in sepsis, which has substantial clinical value for detection in the early stage of sepsis.

In recent years, critical ultrasound has been widely used to evaluate and to aid in the treatment of critically ill patients. Previous studies measured the blood flow velocity in SMA in several pathologies, including patent ductus arteriosus, sepsis, and necrotizing enterocolitis [23]. However, most of these assessed patients with sepsis, allowing for the application of indirect techniques to estimate visceral blood flow, but splanchnic microcirculatory blood flow has rarely been measured [5]. We used SMA ultrasound for the evaluation of patients with sepsis and found that early monitoring of PSV and EDV is extremely valuable in predicting mortality. Although our finding can bridge the knowledge gap, it must be emphasized that our study aimed to conduct a qualitative assessment of the SMA flow, rather than an absolute quantitative comparison between the groups.

The gastrointestinal tract plays a crucial role in the pathogenesis of many diseases, and it is considered to be the motor of multiorgan failure [24]. Non-occlusive mesenteric ischemia is reported to often occur in critically ill patients or in those with significant comorbidities. It is caused by a circulatory failure as well as vasoconstriction in the mesenteric bed [19]. A previous study reported that as a result of non-occlusive mesenteric ischemia, blood flow in SMA reduced to a great extent compared to the cardiac output, and along with blood flow reduction, mesenteric vascular resistance remarkably increased [25]. Therefore, early monitoring of the mesenteric blood flow is of great importance in predicting and evaluating the outcome of severe diseases.

In this study, we observed low PSV, EDV, and RI in the non-survivor group. PSV and EDV alone showed high sensitivity and specificity in predicting the risk of mortality. Therefore, any therapeutic intervention that targets the splanchnic flow may be of great significance in critically ill patients [24]. Other approaches such as vascular catheterization, dilution techniques, and surface mucosal transducers may be more effective for monitoring blood flow; however, they are more expensive and invasive, and expose the patient to the risk of complications and radiation [14, 24]. The SMA ultrasound aids in the analysis of the superior and inferior mesenteric flow and provides several quantifiable parameters, such as pulsatility index, RI, systolic and diastolic velocities, and blood flow volume, to assess the signal from visceral vessels [26].

Platelets play an important role in the coordinated immune response to an infection; however, during a dysregulated host response such as sepsis, platelet activation can cause inflammatory dysfunction and contribute to organ damage [27]. A previous study reported that a sharp drop in the serum albumin level due to bacteremia seriously affected the prognosis [28]. However, no significant differences were observed between the survivors and non-survivors with regard to platelet count and albumin level in our study. Lactate is a reliable prognostic biomarker for poor prognosis in patients with sepsis [12, 28]. However, the mechanism of hyperlactemia is complicated, and several reasons can cause an increase in lactate levels, including tissue hypoperfusion, underlying diseases, drugs, and birth defects. Therefore, the diagnostic value of the lactate level alone may be insufficient [28].

Despite the promising results, our study has some limitations. As only 72 elderly patients were eligible for and analyzed in our study, the study may be prone to sampling bias. This sample size may not be an accurate representation of the target population, and caution must be exercised before extrapolating the results of our study to patients with sepsis of all ages. Despite this limitation, our results are internally valid.

SOFA scores are frequently used to assess organ dysfunction and are closely related to the risk of mortality, with low scores indicating better outcomes [28]. In our study, a strong negative correlation was observed between PSV and the SOFA score; we found that the PSV value was a better predictor of mortality than the SOFA score. Based on our results, SMA ultrasound should be used at an early stage after patients with sepsis are admitted to ICU, with or without septic shock.

Conclusion

In conclusion, low PSV, evaluated using the SMA ultrasound, as found to be an independent risk factor and a good predictor for mortality in patients with sepsis. This finding is important for performing serial measurements in future prospective SMA ultrasounds.

Supporting information