The recommendations refer to patients with systemic hypoperfusion, independently of the underlying cause, and run parallel to the specific measures applicable to each disease condition (e.g., septic drainage, coronary vessel opening, and fibrinolysis in cases of pulmonary thromboembolism). The recommendations are addressed to all intensivists, residents in training and other professionals who care for critically ill patients in their daily practice.

We use mean arterial pressure (MAP) as an estimation of tissue perfusion pressure. Since from the physiological perspective the vascular system loses its autoregulating capacity at MAP values of less than 60–65mmHg, most studies have proposed a target MAP of 65mmHg. In septic patients, for example, a MAP cutoff value of 65mmHg during the first 48h of admission was found to be the value that best discriminated between survivors and non-survivors.9 Furthermore, the achievement of higher MAP values did not result in better outcomes. However, it should be mentioned that recent capillary videomicroscopic studies have shown that the microcirculation of certain patients may indeed benefit from higher MAP values,10 thus suggesting the possibility of individualizing MAP according to its effect upon the microcirculation. However, we lack prospective studies on the prognostic impact of the individualized optimization of MAP according to microcirculatory parameters.

Two special situations require mention in the management of MAP in acute critical illness: (a) uncontrollable bleeding in trauma patients and (b) patients with severe traumatic brain injury in the absence of systemic bleeding.11 In the first situation, it is advisable to maintain a MAP level of 40mmHg until surgical control of the bleeding is achieved. In situations of severe traumatic brain injury with neurological impairment but no evidence of systemic bleeding, since we do not know the brain perfusion pressure, it is advisable to maintain a MAP level of 90mmHg. Once intracranial pressure is monitored, we adjust the MAP level to ensure brain perfusion.12

In addition, to ensure correct tissue perfusion, we must adapt DO2 to the metabolic needs. Our aim in this regard is to increase DO2. Although hemodynamic resuscitation is fundamentally based on optimization of the flow variables (ensuring arterial O2 content), as a general rule we do not seek any predetermined values but rather the adaptation of these variables to secure normalization of the parameters that reflect the global oxygen supply-consumption balance. Accordingly, after discarding or correcting arterial hypoxemia, it does not seem logical to seek concrete values referred to DO2, cardiac output (CO) or hemoglobin, but to adjust them according to the global condition of tissue oxygenation.

Special mention should be made of the resuscitation of high surgical risk patients–a population in which there is enough evidence to recommend the use of certain levels of DO2 (>600mlO2/min/m2) as a target for hemodynamic optimization before, during and after surgery.13 However, it seems reasonable to assume that the use of variables which inform us about the balance between DO2 and VO2 offers advantages compared with the isolated use of DO2 variables, with a view to avoiding both over- and under-resuscitation (with the problems this entails).

As we have mentioned, beyond the achievement of predetermined DO2 or flow values, hemodynamic resuscitation seeks to normalize the global perfusion markers. In clinical practice we fundamentally have two extremely useful variables for this purpose: venous oxygen saturations and the lactate levels.

Although in recent years there has been growing evidence of the prognostic value of different microcirculatory parameters, no studies have evaluated their usefulness as targets or goals in the hemodynamic resuscitation process.8 Thus, although very promising, it is presently not possible to recommend the incorporation of these markers to clinical practice as guiding parameters in the reversion of shock.

Cardiac output (CO) is considered a marker or evaluator of global cardiac function. It moreover also offers data on the cause of shock and of organ failure. CO is therefore a fundamental parameter in the hemodynamic evaluation of the critical patient. Nevertheless, the value of CO must be integrated with other hemodynamic variables (measures of preload, contractility, postload), biological signs and tissue oxygenation parameters, in order to obtain full information with which to guide our treatment decisions.

There is little evidence in support of the systematic monitoring of CO in critical patients.1,2,11,21,22 In many cases the situation of hemodynamic instability can be resolved through more simple assessment and monitoring (physical examination, diuresis, blood pressure [BP], estimation of preload and volume response parameters, etc.), without having to resort to additional measures or procedures. However, some patients continue to show signs of hypoperfusion 3–6h after the start of treatment. In these cases it may be useful to use more exhaustive monitoring producing more detailed information on cardiovascular function, and which may allow us to understand the reason for initial management failure and thus more adequately guide the resuscitation measures. Such monitoring, which must include CO, should be introduced early once the patient is found to be refractory to initial treatment.2,23

In patients with severe initial hypoxemia and suspected heart failure, or in patients with complex cardiopulmonary problems, it seems reasonable to monitor CO from earlier stages, since the initial resuscitation measures (volume expansion, mechanical ventilation, etc.) can worsen cardiac and respiratory function. On the other hand, in cardiogenic shock, correct and early monitoring of CO is particularly important not only in terms of the diagnosis but also in guiding posterior treatment.1,2,21

Lastly, and as commented above, when dealing with high-risk surgical patients, the optimization of CO during the operation and in the hours immediately after surgery has a direct impact upon the patient prognosis.13

Many methods are currently available for monitoring CO, with important differences among them. Information can be found regarding the existing devices in the chapter dedicated to the estimation of CO of the “Update in hemodynamic monitoring” series.23 These monitoring systems can be classified according to their invasiveness. In this regard we have invasive devices (PAC), semi-invasive systems (transpulmonary thermodilution, lithium dilution, pulse wave analysis, esophageal Doppler ultrasound, etc.) and noninvasive techniques (ultrasound, bioreactance, Doppler technology, etc.).

Cardiac output obtained by thermodilution with the PAC has been regarded as the gold standard for the measurement of CO since its introduction in 1970.24,25 Most of the methods for estimating CO have been evaluated by comparison with the data afforded by thermodilution with the PAC, despite the fact that this technique has limitations and might not be the best comparator in this sense. The PAC moreover allows us to obtain relevant hemodynamic parameters such as pulmonary artery pressure (PAP), pulmonary artery occlusion (wedge) pressure (PAOP), and parameters referred to DO2 and VO2. However, the use of PAC has decreased because of its invasiveness and controversy regarding its possible complications and its indications.25,26

The semi-invasive methods for estimating CO in turn include transpulmonary thermodilution, lithium dilution and pulse wave analysis.

Transpulmonary thermodilution is a variant of the thermodilution method in which injection of the saline bolus is made through a central venous catheter, and the temperature change is detected by a sensor placed in an artery (femoral or axillary)–the CO value being calculated from a modification of the Stewart-Hamilton equation. Its use is subject to debate in situations characterized by important body temperature variations, the use of extracorporeal filtration systems, and intracardiac shunts. A number of studies have validated this technique in different critical patient populations.2,23,27,28

Lithium dilution is based on the use of lithium chloride as a tracer for calculating CO. Calibration is performed by injecting a lithium chloride bolus into a central or peripheral venous line, and an electrode placed in an artery detects the lithium concentration in arterial blood. Cardiac output is calculated using the area under the concentration–time curve (AUC). The technique is contraindicated in patients receiving treatment with lithium or non-depolarizing muscle relaxants, and in patients with intracardiac shunts. Lithium dilution has also been found to be useful in different studies conducted in Intensive Care Units (ICUs) and in the surgical setting.29

Many devices offer us the possibility of obtaining CO on a continuous basis using arterial pulse wave analysis. The systems currently found on the market are PiCCO® (Pulsion), PulseCO® (LiDCO), Modelflow® (TNO/BMI), Most Care® (Vygon) and FloTrac®/Vigileo (Edwards Lifesciences). They differ in the way of transforming the information provided by the blood pressure wave profile in beat-to-beat stroke volume (SV), in the algorithms used in each case, in the form of calibration (since some require manual calibration while others do not), the arterial catheterization site, the parameters analyzed, and the accuracy with which they are able to determine CO.30–33 Most of the techniques at the same time generate continuous information referred to preload, postload and contractility, and also allow calculation of percentage pulse pressure variation (PPV) or stroke volume variation (SVV), used to guide fluid therapy and analyze its effects. The main importance of these devices is that they allow us to calculate CO in a continuous and scantly invasive manner. On the other hand, there are resuscitation protocols based on hemodynamic parameters obtained with these systems that have been shown to reduce morbidity among postsurgical and critical patients.32,34 Their main disadvantage is that they lose precision and reliability in situations characterized by important changes in volemia and changes in vascular tone – these aspects being particularly important in the case of systems without external calibration.

Esophageal Doppler ultrasound estimates blood flow in the descending aorta and allows us to calculate SV. The limitations of this technique include operator dependency and the fact that the probe is poorly tolerated in non-ventilated patients.2,23,35 It should be noted that this system has been shown to be useful for fluid optimization in high-risk surgical patients, resulting in shorter hospital stay and fewer postoperative complications.

Among the noninvasive methods used to monitor CO, mention should be made of bioreactance, transthoracic Doppler ultrasound and echocardiography.

Bioreactance, used by the NICOM® system (Cheetah Medical), is based on analysis of the phase shift that takes place in the high-frequency electric impulses targeted to the thorax, produced by changes in blood volume. Promising results have been obtained in heart surgery patients, though there are still not enough studies on its usefulness and reliability in broader critical patient populations.2,23,36 Transthoracic Doppler ultrasound in turn consists of applying a blind Doppler probe over different thoracic areas to measure flow at different levels of the cardiovascular system. This technique has a fast learning curve, with no need for calibration, but is operator dependent. The most widely used system, and which is supported by the largest number of studies in the literature, is the USCOM® monitor (Pty Ltd.). Despite its purported advantages, however, the literature comparing its utilization with the PAC in the intensive care setting is scarce.2,23,37

Echocardiography allows us to determine CO in a noninvasive (transthoracic echocardiography, TTE) or minimally invasive manner (transesophageal echocardiography, TEE), and moreover affords abundant hemodynamic information. Despite its multiple applications and the rapid spread of its use in the ICU, the technique requires adequate training in order to guarantee the quality and reliability of the measurements.3,4

The choice of one technique or the other in estimating CO is influenced by a number of factors, some of which are related to the device itself (e.g., its advantages and limitations), while others may be of an institutional nature or related to the patient. On the other hand, it should be taken into account that the use of a less invasive system may be preferable if it is able to quickly and easily offer us information, even if such information is slightly less exact–particularly in situations in which rapid assessment of the patient condition is required. In turn, the comparability of the techniques for the monitoring of changes and trends in CO may be more relevant in clinical practice than the degree of concordance or agreement among the absolute values. In many cases, the choice of hemodynamic monitoring device depends not only on the technique for estimating CO but also on additional parameters provided by the system, the severity of the patient condition, and the etiology underlying shock.1,21,22 The indication and type of continuous monitoring will be addressed in another section of these recommendations.

Volume expansion constitutes the first line therapy in most situations of hemodynamic instability, despite the fact that only 50% of the patients admitted to the ICU respond to fluid administration by increasing SV and CO.38–40 Recent data suggest that resuscitation with early and “aggressive” volume administration can limit or revert tissue hypoxia, progression to organ failure, and improve the prognosis.15,41 On the other hand, a clear association has been found between the accumulated water balance and mortality among critical patients.42 These findings underscore how important it is for the hemodynamic parameters proposed in deciding volume administration to be able to identify those patients who will benefit from volume expansion by increasing SV (responders) and those who will not (non-responders)–with a view to avoiding useless and potentially harmful treatments.

The parameters traditionally used to decide fluid administration are the preload estimators known as static parameters: filling pressures (central venous pressure [CVP] and PAOP), and the volumes and areas (global end-diastolic volume, right ventricle end-diastolic volume and left ventricle end-diastolic area). Central venous pressure, in particular, and also PAOP are still used in daily practice as routine tools for deciding when to administer volume. In this respect, a recently published German registry has shown that over 92% of all intensivists and anesthetists use CVP for monitoring patient resuscitation following heart surgery.43 Likewise, the guidelines of the Surviving Sepsis Campaign44 recommend the use of CVP within an early resuscitation protocol, based on the results of the study of Rivers et al.15 However, failure of these parameters to predict volume response has been evidenced by many studies.40,45–47 No significant differences have been found in the baseline CVP or PAOP values between patients who respond and patients who do not respond to volume expansion. Other preload estimators such as the end-diastolic area obtained by echocardiography or the volumes obtained with PAC or PiCCO® likewise have not been identified as good predictors of cardiovascular response to volume administration. Nevertheless, it is accepted that very low CVP or PAOP values (<5mmHg) could be predictive of a favorable response.47 The reasons why the static parameters are not reliable estimators of volume response include technical measurement aspects and the possibility of inadequate estimation of transmural pressures secondary to the effect of positive end-expiratory pressure (PEEP), abdominal hypertension, or alterations of ventricular distensibility. Moreover, from the purely physiological point of view, the knowledge of a concrete preload value does not allow us to know the preload reserve of a ventricle, due on one hand to the fact that the Frank–Starling curve or ventricular function curve is curvilinear (ascending segment: dependence on preload; descending segment: independence of preload), and on the other hand to the fact that there is no single ventricular function curve but rather a whole family of curves that relate preload and SV according to the changes in postload or contractility. For this reason, and for a given preload value, the increase in SV will depend on the part of the Frank–Starling curve on which both ventricles are operating.

The use of dynamic parameters based on a “functional” hemodynamic evaluation has recently been proposed. This approach consists of introducing a reversible and transient change in cardiac preload, followed by observation of the effects upon SV or CO. The most widely studied dynamic parameters in patients subjected to mechanical ventilation are those obtained from the analysis of the changes in SV and blood pressure during a mechanical respiratory cycle, based on heart–lung interaction.38,39,47–51 The main hemodynamic effect of an increase in intrathoracic pressure is a decrease in venous return and in right ventricle ejection, which may give rise to a drop in preload and in SV of the left ventricle. The magnitude of the respiratory changes of SV depends on the position of both ventricles on the ventricular function curve. Thus, when both ventricles are operating on the ascending part of the curve (the zone of preload dependence), a change in preload induced by positive intrathoracic pressure gives rise to a significant change in SV and in pulse pressure (since the latter is directly proportional to SV). Many studies have found that PPV values ≥13% [(maximum pulse pressure−minimum pulse pressure)/(maximum pulse pressure+minimum pulse pressure)/2×100] and SVV≥10% [(SV maximum−SV minimum)/(SV maximum+SV minimum)/2×100] predict the response to volume expansion with high sensitivity and specificity in different critical patient populations. Furthermore, it repeatedly has been shown that the dynamic parameters are better predictors than the static parameters. In this sense, a recent systematic review of 29 studies involving 685 patients referred to the predictive capacity of dynamic parameters derived from the blood pressure curve showed the areas under the curve (ROC) for PPV and SVV to be 0.94 and 0.84, respectively, while in comparison the areas of the static parameters were 0.55 for CVP; 0.56 for the indexed global end-diastolic volume; and 0.64 for the left ventricle end-diastolic area.50 Other methods based on the same physiological concept, such as the variation in aortic flow velocity as determined by esophageal Doppler and the variation in peak velocity or the velocity–time integral of aortic flow determined by echocardiography, are also useful for predicting the response to volume expansion. Analysis of the amplitude of the plethysmographic signal in pulsoxymetry and the plethysmographic variability index has even been proposed.52–55 Likewise, other indices based on heart–lung interaction, such as variations of the superior and inferior vena cava, have been validated as reliable and precise predictors in patients subjected to mechanical ventilation.56,57

Despite the undeniable usefulness of these parameters in the evaluation of patient response to volume expansion, it is important to know their limitations. Firstly, the patient must be subjected to controlled mechanical ventilation, without spontaneous respiratory activity. On the other hand, these parameters have not been validated in the presence of cardiac arrhythmias, and their predictive value is lower in patients ventilated with tidal volumes below 8ml/kg.48,58

Raising the legs of the patient can also be useful in predicting the response to volume administration. Raising the legs to an angle of 45° with respect to the bed for at least 1min reversibly produces the cardiovascular effects of the administration of a volume of 300ml. The hemodynamic response to this maneuver–regarded as a test rather than a treatment–can be used in patients subjected to mechanical ventilation and with spontaneous breathing. Its predictive capacity has also been demonstrated in patients with arrhythmias. An increase in SV>10–12% estimated by echocardiography, PiCCO® system, esophageal Doppler or other devices during the maneuver is able to predict an increase in SV>15% following the administration of volume, with high sensitivity and specificity. A recent metaanalysis including the results of 8 studies confirms the excellent value of the leg-raising maneuver in predicting volume response in critical patients, with an area under the curve (ROC) of 0.95.59,60

In patients with spontaneous breathing, some studies have examined different maneuvers that could be useful in clinical practice, though confirmatory studies are still needed: the variation in arterial pulse pressure during the Valsalva maneuver, the variation in blood pressure with end-expiratory occlusion maneuvering, and the variation in right atrial pressure.40

Lastly, the evaluation of CO response to the administration of a certain volume of CO (fluid challenge) has been used for many years in clinical practice to assess the efficacy and safety of volume expansion therapy. This maneuver should be used in cases in which none of the above described parameters can be evaluated.61

Contractility can be defined as the capacity of the heart to generate external work independently of the loading conditions. Cardiac dysfunction is mainly caused by failure of the ventricular pump, which is unable to afford sufficient hydraulic energy to maintain effective circulation. Most of the indices used to evaluate contractility at experimental or clinical level are partially dependent upon preload or postload–a fact that may complicate the evaluations. The slope of the end-systolic ventricular pressure–volume ratio, known as end-systolic elastance (Ees), is regarded as the reference contractility index, due to its relative independence of the loading conditions and its sensitivity to changes in inotropism. The generation of complete systolic and diastolic data requires invasive left ventricle instrumentation for the simultaneous estimation of measures of pressure and volume, which complicates determination,62–65 though echocardiography could partially facilitate the recording of Ees at the patient bedside in a noninvasive manner. However, it is not possible to determine Ees in unstable patients through occlusion of the inferior vena cava and the administration of vasodilator or vasoactive drugs, since the pre- and postload alterations could trigger a life-threatening situation. A number of methods have been proposed for recording Ees through the estimation of a single heart beat, without the need to modify the loading conditions, though there is not enough clinical evidence in critical patients to apply such methods on a systematic basis.66

Traditionally, the parameter most widely used in the ICU for assessing ventricular function has been CO. However, it must be remembered that this hemodynamic parameter depends not only on contractility but also on preload and postload. The joint utilization of CO and filling pressures (CVP, PAOP) yields a series of hemodynamic patterns or references that can be very useful in clinical practice. In this sense, severe left heart failure is characterized by low CO and elevated PAOP. The most characteristic example of the usefulness of these patterns is the Forrester classification of acute myocardial infarction.67 However, the data obtained from the combination of CO and filling pressures does not afford sufficient information to know the mechanisms (alteration of ventricular distensibility, decreased contractility, increased postload, etc.) underlying the clinical and hemodynamic findings of the patient. Another parameter classically used in the ICU for the evaluation of contractility at the patient bedside is stroke work (SW), defined as the product of SV and the difference between MAP and PAOP. This index also depends on preload, but is independent of postload; as a result, the observation of low SW could serve to identify diminished contractile function in situations in which volume expansion has been adequate. Many studies in the fields of heart failure, ischemic heart disease and sepsis have used this parameter as an indicator of altered contractility, and it has even been related to patient prognosis.64,68

The heart may be regarded as a mechanical pump capable of generating hydraulic energy. Accordingly, the pumping capacity of the heart can be expressed as cardiac power (CP), defined as the product of the flow and pressure generated by the heart. Thus, CP is the product of CO and MAP determined simultaneously. Cardiac power is not significantly influenced by postload, though it can change with variations in preload. For this reason, we must make sure that the heart of the patient is not dependent upon preload before affirming that CP–and therefore heart pumping function–is diminished. In recent years, the maximum CP recorded after pharmacological stimulation or exercise has been used to evaluate patients with acute or chronic heart failure. In this regard, different studies have shown that the maximum CP reached is a more powerful predictor of the evolution of patients with chronic heart failure than VO2, CO or left ventricular ejection fraction (LVEF).69 These results have been confirmed in critical patients in situations of cardiogenic shock. Tan and Littler70 found the hemodynamic evaluation of cardiac reserve (difference between maximum CP maxima and basal CP) through dobutamine stimulation to clearly distinguish between survivors and non-survivors. Patients with maximum CP<1W (normal basal value for the healthy adult) or SW index <0.25j/m2 died, while all those with higher values survived for over one year. A posterior confirmation of the prognostic value of CP has recently been published in the same population of patients.71 For the time being, no data have been published referred to broader critical patient populations.

Echocardiography plays a key role in the evaluation of contractility, since it offers us a range of parameters that can be useful in the ICU for assessing contractility. The left ventricular ejection fraction is the parameter most widely used to evaluate contractility in patients with heart disease.3,72,73 In this regard, LVEF is considered fundamental in patients with heart failure not only because of its prognostic importance (the lower the LVEF, the poorer the patient prognosis) but also because most studies select patients based on their LVEF.72 At the time of its interpretation and evaluation in critical patients, we must take into account that LVEF is also dependent upon preload, heart rate and, fundamentally, postload. This dependence would be particularly significant in situations characterized by an important increase or decrease in postload, as occurs in sepsis. Furthermore, echocardiography offers us other parameters of known usefulness in the evaluation of contractility, such as the Tei index, dP/dt max, systolic tricuspid annular displacement, and the maximum velocity of the S-wave in mitral or tricuspid tissue Doppler imaging (TDI), which show greater or lesser sensitivity to changes in inotropism and loading conditions.74

At present, the new monitoring systems offer other contractility-estimating parameters such as the cardiac function index and global ejection fraction, obtained by transpulmonary thermodilution, though there are still not enough studies to allow us to recommend their use on a routine basis. Even the measurement of natriuretic peptides could play a role in the noninvasive assessment of ventricular dysfunction and in the evaluation of left ventricle filling pressures–though the existing data are not yet conclusive.64

Postload, defined as the “load” or opposition which the heart must overcome in order to expel its stroke volume, is an important determinant of CO for a given set of contractility and preload conditions. Although aortic pressure is one of the main components of ventricular postload, the parameter best able to estimate the latter remains the subject of debate. A form of assessment could be the estimation of systolic ventricular wall stress according to the equation of Laplace. Echocardiography can facilitate the determination of this parameter.

The estimation of postload in the ICU has been a traditional practice, despite its limitations, based on the determination of systemic and pulmonary vascular resistances using PAC. The current technological advances allow us to use other devices, including ultrasound, to obtain this same parameter in a less invasive manner.

It has been suggested that the ratio between PPV and SVV, called dynamic elastance by Monge Garcia et al.,75 could be of help in estimating vascular tone and, consequently, might be useful in hemodynamic resuscitation algorithms–though there is still not enough literature support in this respect.63,64,75,76

In conclusion, the new technological resources allow the combination of different hemodynamic monitoring techniques, affording information or a different approach to the same problems, with more precise evaluation of the contractility and postload alterations that may be found in critical patients.

Initial management of shock patients comprises adequate anamnesis and physical examination, together with electrocardiographic (heart rate and ECG) and pulsoxymetric monitoring (SpO2)–not only to assess possible cardiovascular failure, but also as necessary complements in diagnosing the clinical condition.

Initial basic hemodynamic monitoring of patients with potentially critical acute disease is determined by the definition of shock itself.8 Thus, the measurement of blood pressure and of at least one variable informing of global tissue oxygenation, such as serum lactate or SvcO2, is essential. The importance of monitoring MAP as an indicator of tissue perfusion pressure has been considered in detail in another section. Different studies have shown that the duration of hypotension in the first hours of shock exerts an accumulative effect upon the development of organ failure9–thus pointing to the need to establish frequent and precise blood pressure measurements. On the other hand, the monitoring of blood pressure using noninvasive systems loses precision in situations of shock. It therefore seems reasonable to indicate invasive and continuous blood pressure monitoring in the critical patient. In any case, in the early stages, the monitoring of blood pressure could be carried out on a noninvasive basis during admission to the hospital ward or in the emergency service.

The clinical detection of tissue hypoperfusion (with or without arterial hypotension) implies the need to adopt measures to improve DO2. In this dysoxia reversion or hemodynamic resuscitation process, the first step always includes the need to optimize volemia. Consequently, and in addition to the abovementioned goals of the resuscitation process, it is advisable from the start to monitor some parameter estimating preload dependency and which can be of help in making initial resuscitation decisions. However, the possibilities at the time of initial monitoring are limited, fundamentally when the patient has not been admitted to the ICU. In addition to an anamnesis and clinical exploration suggesting hypovolemia, the measurement of CVP by inserting a venous catheter with the distal tip positioned in the right atrium or superior vena cava will be the main parameter offering information on volemia, despite its limited capacity to predict the response to volume expansion. Rather than a specific parameter of volemia, or of preload dependency, CVP could be viewed as a marker of minimum circulating volemia. In this respect, extremely low CVP values are probably associated to insufficient volemia, though elevated CVP values are entirely nonspecific in predicting the response to volume expansion.39 Despite all its limitations, different multicenter studies have found initial volume expansion with CVP safety limits of 8mmHg (12mmHg in the case of patients subjected to mechanical ventilation), within the setting of a bundle of measures, to be associated with improved patient prognosis.15,44 Furthermore, if an invasive arterial catheter is available, we can use PPV as an estimator of the response to volume expansion in patients subjected to mechanical ventilation and with sinus rhythm.

Echocardiography currently plays a relevant role in the early diagnosis and management of patients with hemodynamic instability, as will be commented in the following section of these recommendations.3,77

In general, adequate initial management based on the information obtained from the clinical history, the physical examination and basic monitoring may suffice to secure a favorable outcome. However, in some patients, despite adequate initial resuscitation, persistent shock may be observed, or new complications related to the initial disease process or with the adopted treatments may develop. As has already been mentioned in the section referred to the estimation of CO, in patients with an insufficient response to the measures applied in the first 3–6h, or in patients with organ failure and/or comorbidities capable of causing interferences or of worsening during the resuscitation process, we should consider a greater degree of continuous hemodynamic monitoring to help optimize our interventions, quantify their effects and avoid complications resulting from the applied treatments.1,2,21–23

Continuous hemodynamic monitoring should provide information on CO and its determinants: preload/preload dependency, contractility and postload. Thus, in addition to monitoring of the variables used as goals or targets–whether MAP and venous O2 saturation and/or lactate clearance–an intensive resuscitation process will require the incorporation of technologies that allow the continuous evaluation of these parameters, with a view to correcting them. Monitoring of the mentioned target variables will remain essential, since they will mark the end of the resuscitation process. It will be necessary to measure these variables repeatedly after the treatment interventions to confirm that they remain normalized over time.

When choosing a monitoring system, and in addition to factors inherent to the patient, we also must take other elements into account, such as the technological resources available in the center, the experience of the team with each system, the place where monitoring and resuscitation will be carried out (emergency service, ICU, operating room, etc.), and cost-effectiveness criteria. The ideal hemodynamic monitoring system is simple, safe, relatively versatile, easy to use, operator-independent, cost-effective, reliable and precise, and should yield variables and information capable of guiding patient treatment.21,22 No system is presently able to satisfy all these conditions. Nevertheless, the following principles can help us choose a continuous hemodynamic monitoring system:

• -

  Monitoring has been shown to be particularly useful in the early phases of hemodynamic resuscitation, and proves less useful once organ failure has become established.

• -

  Although less invasive systems are preferable, they cannot always be used, since in complex situations we need full, precise and reliable information that is not always obtainable with less invasive techniques.

• -

  Noninvasive systems could be used in the hospital ward or emergency service to confirm a preliminary diagnosis, to assess the course in lesser risk patients, or for monitoring prior to admission to the ICU. For the time being, the use of such systems is not recommended in application to more critically ill patients. In general, the greater the severity and complexity of the disease condition, the greater the need for intensive treatments and increased precision in the measures obtained–and such conditions are currently associated with increased invasiveness of the systems used.21

• -

  Monitoring of hemodynamic changes in short periods of time is important (e.g., volume administration or the use of inotropic drugs). Moreover, continuous monitoring of the hemodynamic variables and the beat-to-beat estimation of CO can be particularly useful.1

• -

  The PAC can be useful for the management of complex circulatory conditions in which it is particularly important to know PAP, PAOP and tissue oxygenation parameters (e.g., acute left- and right-side failure, pulmonary hypertension, difficult weaning, patients subjected to heart surgery or pending heart transplantation).1,21,78

• -

  The transpulmonary dilution techniques that determine intrathoracic volumes and extravascular lung water can be regarded as the methods of choice in guiding hemodynamic management and the application of mechanical ventilation in patients with acute lung injury or acute respiratory distress syndrome. In these cases, PAC could also be used.2

• -

  The systems that determine CO through pulse wave analysis would be indicated in surgical areas, in situations of septic shock, or in other clinical scenarios without serious respiratory complications. Devices without external calibration lose reliability in the presence of important alterations in vascular tone.2,21