Microcirculatory alterations are frequently observed in critically ill and severe sepsis patients characterized by a decrease in capillary density and an increase in heterogeneity of perfusion. This derangements result in microcirculatory shunting and oxygen extraction deficit, and plays a major role in the pathophysiology of sepsis and multi-organ failure. Loss of hemodynamic coherence between the macro- and microcirculation results in persistent and incomplete reversal of microcirculatory and regional hypoxia that is the major factor contributing to the development of organ failure.

Current techniques permitting monitoring the microcirculation can be classified in two main groups: (1) Methods for evaluation of local tissue oxygenation as a surrogate of microcirculatory blood flow. (2)Methods allowing direct visualization of the microvascular network and microcirculatory blood flow.

Near-infrared spectroscopy (NIRS) is a non-invasive technique for evaluating tissue oxygenation in vessels with a diameter < 1 mm (arterioles, capillaries, and venules). Recent systemic review studies have showed that patients with severe sepsis or septic shock have lower levels of StO2, moreover, survivors present higher levels of StO2 compared with non-survivors. Reactive hyperemia during vascular occlusion test (VOT) can be considered an integral test of microcirculatory reactivity, evaluating the tissue's ability to adjust oxygen extraction capabilities to oxygen delivery after a hypoxic stimulus induced by a transient interruption in blood flow. Continuous StO2 measurement and VOT derived StO2 deoxygenation slope and StO2 recovery slope have been found to be predictors of mortality and organ dysfunction.

Sublingual microcirculation allows direct visualization of the sublingual microcirculation and for quantitative determination of variables of flow, capillary density, and flow distribution. Microvascular alterations including decreased functional capillary density, increased perfusion heterogeneity, and increased proportion of stopped and intermittently perfused capillaries, are more severe in non-survivors than in survivors. In addition, persistence of these alterations was strongly and independently correlated with multi-organ failure and mortality.

ARDS is the most severe form of acute lung injury in ICU with mortality about 45% without achievement in ten years. However, only few studies were focused on the microcirculatory alterations in ARDS patients.

Microcirculation consists of a branching network of small blood vessels (<100 μm diameter) that includes the arterioles, capillaries, and venules, and plays a pivotal role in the delivery of oxygen to tissue cells. Microcirculatory alterations are frequently observed in critically ill patients, and especially in patients with severe sepsis. These alterations are characterized by a decrease in capillary density that determines an increase in the diffusion distance of oxygen to tissues and an increase in heterogeneity of perfusion. This microcirculatory derangements involve the consequent presence of under or not perfused capillaries in close proximity to well perfused capillaries. Therefore, these functionally vulnerable microcirculatory areas may become hypoxic, resulting in an oxygen extraction deficit. This phenomenon has been termed "microcirculatory shunting" and plays a major role in the pathophysiology of sepsis and multi-organ failure.

Persistent and incomplete reversal of microcirculatory and regional hypoxia is considered to be a major factor contributing to the development and maintenance of multiple organ failure. Unfortunately, alterations in microvascular perfusion often persist after the correction of systemic hemodynamic abnormalities, and their severity is associated with a poor outcome. Such a loss of hemodynamic coherence between the macro- and microcirculation has been described in several clinical and experimental studies and has been found to be an independent predictor of adverse outcome and organ dysfunction. The microcirculation has been suggested to be the motor of sepsis.

Conceptually, current techniques that permit monitoring the microcirculation can be classified in two main groups: (1) Methods that allow evaluation of local tissue oxygenation as a surrogate of microcirculatory blood flow. (2) Methods that allow direct visualization of the microvascular network and microcirculatory blood flow.

Near-infrared spectroscopy (NIRS) has been used as a tool to monitor tissue oxygen saturation (StO2) in acutely ill patients and has been proposed as a tool to quantify microvascular dysfunction in patients with sepsis. According to Beer's law, the NIRS signal is limited to vessels that have a diameter less than 1 mm (arterioles, capillaries, and venules), but, as 75% of the blood in a skeletal muscle is venous, NIRS StO2 measurements mostly represent local venous hemoglobin O2 saturation. NIRS has been used in different clinical conditions such as severe trauma, hemorrhagic shock, septic shock, and cardiogenic shock or severe cardiac failure. The utility of and application of NIRS in intensive care mainly involved three main NIRS measurements: (1) continuous tissue oxygen saturation (StO2) measurement; 2) StO2 deoxygenation slope (DecStO2 slope) in response to vascular occlusion test (VOT test); and 3) StO2 recovery slope (RincStO2 slope) in response to VOT.

Tissue oxygen saturation (StO2) has been proposed as a marker of tissue perfusion depending on the device used and the location of the measurement. A systemic review and mea-analysis had showed that patients with severe sepsis or septic shock have lower levels of StO2, and RincStO2 slope. Moreover, survivors from severe sepsis or septic shock present higher levels of StO2 and RincStO2 compared with non-survivors. The VOT is a provocative test in which StO2 is measured on a distal site (such as the thenar eminence or forearm) whilst a transient rapid vascular occlusion is performed, using a sphygmomanometer, for either a defined time interval (for example, 3 minutes) or until the StO2 decreases to a defined minimal threshold. The ischemic tissue then induces vasodilation of surrounding arterioles, metarterioles and pre-capillary sphincters to decrease local vascular resistance and regain blood flow. Once this threshold is reached, the tourniquet is released and blood flow is restored. There is a reactive hyperemic response which represents the tissue's ability to auto-regulate blood flow and oxygenation. Several parameters arise from this technique, including: 1. the rate of deoxygenation (RdecStO2), thought to reflect the local metabolic rate, 2. the rate of reoxygenation (RincStO2), thought to reflects the time required to wash out stagnant blood and is thought to be determined by local cardiovascular reserve and microcirculatory flow, and the post-obstructive hyperemic response.

For the rate of StO2 increase (RincStO2), or termed StO2 recovery slope, it is assumed that when the StO2 recovery slope is reduced, the capacities of recruiting microvessels in response to a hypoxic stimulus are lower. Many papers have demonstrated that the RincStO2 is decreased in septic patients. The StO2 recovery slope had been shown to be lower in septic patients and that the presence of this alteration in the first 24 h of sepsis and its persistence were associated with a worse outcome. Among the septic patients, the RincStO2 was higher in survivors than in non-survivors and was most strongly associated with organ dysfunction and mortality.

However, StO2 is not a direct measurement of microvascular blood flow, but an index of tissue oxygenation which is dependent on the balance between O2 delivery (DO2) and oxygen consumption (VO2). Any change in StO2 can reflect a change in flow in the same direction and/or a change in metabolism in the opposite direction. More importantly, proportional changes in flow and metabolism may be associated with unchanged StO2. The vasoreactivity test evaluates microvascular reserve more than actual microvascular perfusion. Indeed, even though StO2 is slightly lower in septic patients compared to healthy volunteers, there is a huge overlap between the groups. No significant difference in StO2 between injured patients and healthy volunteers has been found among septic shock, post-surgery, and healthy subjects; between septic shock and normal volunteers; and between trauma patients and healthy subjects, or even higher tissue oxygen tension in patients with sepsis has been reported.

As a consequence, a vascular occlusion test (VOT) combined with StO2 measurement has been proposed to enhance the discriminatory power and to better evaluate the tissue micro-oxygenation in septic shock. The use of VOT has been shown to improve and expand the predictive ability of StO2 to scenarios such as trauma, severe sepsis and septic shock.

Micro-videoscopic techniques are assumed to be the gold standard for the study of the microcirculation as these allow the direct visualization of microvascular perfusion and characterization of its alterations. Recently, a third-generation lightweight handheld vital microscopes (CytoCam) was developed based on incident dark field imaging. Direct visualization of the sublingual microcirculation allows quantitative determination of capillary density, microvessel morphology, and dynamics of microcirculatory blood flow.

Variables of flow including Microvascular Flow Index (MFI) and proportion of perfused vessels (PPV), as well as capillary density including total vessel density (TVD) and perfused vessel density (PVD), and flow distribution Heterogeneity Index (HI) were calculated according to international criteria. The alterations seen in sepsis are often characterized by a highly heterogeneous perfusion, with stopped flow capillaries next to vessels with flowing cells.

Microvascular alterations, which are characterized by decreased functional capillary density, increased perfusion heterogeneity, and increased proportion of stopped and intermittently perfused capillaries, contribute to the defect in oxygen extraction observed in sepsis, and may be involved in the development of organ failure. Microcirculatory alterations are more severe in non-survivors than in survivors. The survival rate progressively decreased with quartiles of severity in alteration of the microcirculation. In addition, alterations in microvascular perfusion were one of the strongest predictors of outcome and remained independently associated with outcome in multivariate analysis. The time course of microvascular alterations also differs between survivors and non-survivors. Microvascular alterations improved over time in response to therapy in survivors but not in non-survivors. Conversely, the persistence of these microcirculatory alterations after the first 24 h was strongly and independently correlated with mortality secondary to circulatory failure in the early phase and to multi-organ failure in the late phase. For the flow parameter, in 2009, an international multicenter observational prevalence study in ICU patients involving 36 ICUs worldwide enrolling 501 patients identified microcirculatory alterations MFI < 2.6 in combination with tachycardia (heart rate > 90) as an independent risk factor for increased hospital mortality.

Among microcirculatory variables, there is no clear consensus on which microcirculatory perfusion parameter is most important. The decreased microcirculatory perfusion as measured by PPV, PVD, and MFI was associated with mortality. They reported that the PPV parameter was the strongest predictor of mortality and that this association was maintained in multiple logistic regression models for both early (< 24 h) and late (≥ 24 h) time points. In a similar study, impaired flow and increased heterogeneity of flow were significantly disturbed features of the microcirculation in non-survivors compared with survivors. Furthermore, in a study of 49 ICU patients in septic shock, there was no difference in microcirculatory perfusion parameters at the onset of shock, but survivors were able to restore their microcirculatory perfusion as indicated by significant differences in PPV. However, in the ProCESS trial, the measures of density, namely TVD, PVD, and De Backer score (an estimate of total density), to be associated with mortality at all time-points in a single model and at the 72-h time period. Therefore, which microcirculatory parameters having the most significant pathophysiologic impact still await for further studies to delineate.

Acute respiratory distress syndrome (ARDS) is the most severe form of acute lung injury in ICU. The hallmark of ARDS pathogenesis is marked increased permeability from endothelial and epithelial injury [58]. Persistent and marked tissue hypoxia and multiple organ failure are the main cause of death. The attributable mortality of ARDS remains high without achievement within the recent ten years. Recently, The Large Observational Study to Understand the Global Impact of Severe Acute Respiratory Failure (LUNG SAFE) study revealed that in the 29 144 patients admitted to participating ICUs, the hospital mortality was 34.9% for those with mild, 40.3% for moderate, and 46.1% for severe ARDS.

Therefore, from the aforementioned discussion for the two major assessment NIRS and sublingual microcirculation, the aims of our study are the followings:

1. To compare the predictive power of the NIRS and sublingual microcirculation for the prognosis of ARDS. Furthermore, whether the prognosis of ARDS can be predicted more precisely if investigators simultaneously assessing microcirculatory alteration with NIRS and sublingual microcirculation?
2. To delineate which parameters of microcirculatory alteration, variables of flow or density have the most significant pathophysiologic impact for the prognosis of ARDS?

Inclusion Criteria:

• ARDS patients conforming to the Berlin Definition of Acute Respiratory Distress Syndrome
• ventilatory support with mechanical ventilation

Exclusion Criteria:

• age < 18 years
• pregnancy
• established "Do Not Resuscitate" orders prior to enrollment
• acute cerebrovascular event (primary diagnosis)
• combined with acute coronary syndrome with myocardial infarction (primary diagnosis)
• acute and active gastrointestinal bleeding (primary diagnosis)
• requirement for immediate surgery
• inability to obtain written informed consent

admitted to medical ICU and the hemodynamics is stable (mean arterial pressure MAP > 60 mmHg with no change in vasopressor infusion rate for 2 hours) will be enrolled.