Shock states are defined by stereotypic changes in well-known physiologic parameters. While these well-known changes provide a convenient entry point into further evaluation of patients in shock or at risk for shock, use of such physiologic evaluation is not commonly seen in clinical medicine. A formal description of physiologic reasoning in the diagnosis of shock states is presented in this paper. Included with this conceptual framework is a discussion of key tests or findings that can be used to differentiate between possible diagnoses, and the pairing of treatment strategies to distinct classes of physiologic abnormalities. It is hoped that the methodology presented here will demonstrate the primacy of physiologic reasoning in the diagnosis and treatment of hemodynamic instability. Advantages of this method are speed and accuracy, efficient use of resources, and mitigation against sources of medical errors.
Despite what appears to be rather universal teaching of principles of cardiovascular physiology in the didactic phase of undergraduate medical education, the application of such knowledge in patient care is often sparse. A classic example is the diagnosis and management of hemodynamic instability. The diverse set of clinical conditions that can be termed “shock states” are characterized by stereotypic perturbations in well-known physiologic relationships while resuscitation principles are based upon understanding and correcting their underlying parameters. Despite the simplicity offered by consideration of the underlying physiology of shock, it is rare to see trainees frame the management of hemodynamic instability in these terms.
Often, patient assessment appears to be governed more by a hit and miss approach where clinical context and suspicions lead to a series of diagnostic studies aimed to either rule in or out a specific hypothesis, or to narrow the range of possibilities. As a practical issue, this approach often delays therapy due to the reluctance to commence treatment prior to arriving at a diagnosis. Further, the approach may unnecessarily expose patients contrast dyes, X-rays, and other invasive procedures. The stabilization of hemodynamic compromise cannot tolerate a long series of investigations prior to institution of therapy, and, in fact, poorer patient outcomes have been associated with this approach [
Below, a general formulation of shock-related organ dysfunction will be presented, followed by a discussion of diagnostic evaluation and therapeutic planning. With this discussion, the reader will be provided with a framework for understanding the etiology and therapy of critical illness that is suitable for everyday use on the wards and in the ICU.
For purposes of discussion, shock will be defined as a state of circulatory collapse of a magnitude that would lead to multiple organ dysfunctions if not rapidly corrected. While the basic classification of shock states into categories of cardiogenic, hypovolemic, and distributive shock is well known to all clinicians, these patterns are not always apparent to those caring for a patient. Rather, it is common to receive a call from a nurse or trainee reporting a patient with abnormal findings such as fever, tachycardia, or a drop in blood pressure. Starting from nonspecific findings such as these, a crucial part of patient evaluation includes determining whether the patient’s findings are consistent with one of the shock states, and to have a reliable approach to this question that is consistent with the known pathophysiology. Accordingly, the risk for organ dysfunction arising from hemodynamic compromise can be traced to departure from the normal ranges in either one or both of the following physiologic relationships. The autoregulatory curve describing the relationship between organ blood flow and mean arterial pressure [ The relationship between the rates of oxygen uptake (demand), and oxygen delivery (supply) [
The graphic representations of these functions are presented in Figure
Typical hemodynamic changes associated with three accepted categories of shock. Arrows show degree of change from baseline in blood pressure (BP), cardiac output (CO), systemic vascular resistance (SVR), and cardiac preload. Additionally, alterations in the relationship between oxygen delivery and demand (DO2/VO2) and mean arterial pressure and organ blood flow (MAP/OBF) are indicated. The asterisk (*) indicates the primary abnormality associated with each shock state.
Type of shock | BP | CO | SVR | Preload | DO2/VO2 | MAP/OBF |
---|---|---|---|---|---|---|
Cardiogenic | nl- | nl- | nl- | |||
Hypovolemic | nl- | nl, | ||||
Distributive | nl, | nl- |
Abbreviation: nl = normal range; arrows show increases or decreases.
The key determinants of organ perfusion are depicted. In (a), the relationship between oxygen consumption (VO2) and delivery (DO2) is indicated. Patients usually function on the rightward side of the curve where an excess of oxygen is supplied relative to demand. As delivery decreases relative to consumption, the patient moves left on the curve. A decrease in central venous oxygen saturation accompanies leftward movement on the curve. In severe cases where delivery is unable to meet metabolic demands, the patient slips beneath the critical oxygen delivery threshold. Organ dysfunction and lactic acidosis are regarded as evidence of pathologic oxygen supply [
The linkage between circulatory insufficiency and organ dysfunction involves the triggering of a state of malignant inflammation and microcirculatory dysfunction [
If the organ dysfunction resulting from most shock states can be traced to alterations in one or two key homeostatic relationships, then it seems reasonable that evaluation and diagnosis should involve a direct consideration of the same. Thus, for any patient exhibiting distress, change in solid organ function, or any suggestion of hemodynamic perturbation, the clinician bears the responsibility to consider the status of both the blood pressure/autoregulation and oxygen supply/demand relationships. Indeed, failure to evaluate these two relationships is a common source of misdiagnosis and poor patient outcomes in patients transferred from the ward to the intensive care unit [
In practical terms, evaluation of the relationships noted above focuses on blood pressure and oxygen delivery. If an abnormality in the either maintenance of blood pressure or oxygen delivery is present, then further analysis is guided by the derivation of these terms according to the diagram in Figure
Dendrogram describing the constituents of mean arterial pressure and oxygen delivery in the context of suspected decompensation. For each key abnormality, physiologic variables are indicated in
To provide an example, assume you are called to evaluate a patient with a new onset of agitated delirium a few days after an abdominal operation. The patient is picking at his drains and indwelling devices; the pulse rate of 120 beats per minute is up from previous readings in the 80 s. The vital signs are BP: 121/85, temp: 36.5, SaO2: 95% on room air, and a respiratory rate of 26 per minute. The temptation of many would be to medicate the agitation with an antipsychotic and perhaps a benzodiazepine if alcohol withdrawal is suspected. Instead, the approach advocated here is to consider the pulse of 120 and a failing organ (cognitive dysfunction) as potential early warning signs of cardiorespiratory collapse. The scheme advocated here is to consider the physiology of oxygen delivery and blood flow prior to jumping to diagnostic conclusions, or at least to do so in parallel as a means of confirmation.
Systolic blood pressure less than 90 mm Hg and mean pressures less than 60 mm are commonly used to indicate absolute hypotension. This is based less on rigorous testing and evidence than on tradition and classic large animal studies on autoregulation of blood flow. A baseline hypertensive patient would operate on a right-shifted antiregulatory curve, and may not have normal organ perfusion at mean pressures of less than 65–70 mm Hg. Retrospective analyses of trauma registries have supported the existence of age-related relative hypotension [
Continuing with the case above, if the patient does not seem to be way off his baseline blood pressures, does this mean the patient’s risk assessment is complete? No, an evaluation of oxygen supply demand status is necessary. Acute drops in arterial blood pH, changes in anion gap, and elevated lactate levels are associated with an oxygen debt or anaerobic metabolism, and can be rapidly assessed by a clinical laboratry or with portable point-or-care machines. Central venous blood with oxyhemoglobin saturation less than 70% (normal) indicates falling oxygen delivery relative to demand. Any centrally inserted lines including those inserted for dialysis or chronic care (PICC lines) can yield samples suitable for analysis. Acute elevations in the anion gap or lactate should be regarded as a sign of oxygen supply/demand imbalance until proven otherwise.
Disruption in the economy of oxygen supply/demand imbalance is due to either excessive consumption, inadequate delivery or a combination of both (Table
Disruptions in the normal economy of oxygen supply and demand can have profound impacts on solid organ function and long-term outcomes. Typically, changes in metabolic mode from aerobic to anaerobic as evidenced by high lactate or decreased central oxyhemoglobin saturation (sCVO2) can be explained by changes in either supply, demand, or both. Except in rare instances of hypermetabolism, most pathology in oxygen supply/demand can be traced to problems with delivery (DO2).
Increased demand (VO2) | Decreased supply, delivery (DO2) |
---|---|
Hemorrhage | |
Thyrotoxicosis | Decreased cardiac output |
Malignant hyperthermia | Decrease heart rate |
Decreased stroke volume syndromes | |
Fever | Low blood volume |
Systemic inflammation | Dehydration |
Shivering, thermogenesis | Hemorrhage |
Muscle contraction, fighting | Normal or elevated blood volume |
Heart failure, cardiogenic shock | |
Inadequate filling time, tachycardia | |
Valvular obstruction | |
Valvular insufficiency | |
Pulmonary hypertension | |
Obstructive shock | |
Pneumothorax | |
Pulmonary embolus | |
Cardiac tamponade |
Low stroke volume states can be divided into two broad categories: (1) a low intravascular volume leading to low stroke volume and (2) a normal or high intravascular volume that cannot be converted to a normal stroke volume due to obstruction, inefficient ventricular filling, pump dysfunction, intracardiac abnormality, or valvular incompetence. The “low circulating volume” class of low stroke volume shown in Figure
Significant blood loss as a cause of hypovolemia is diagnosed by sudden drops in hemoglobin, either alone or in the context of fluid administration. With instantaneous hemorrhage, isovolemic blood loss prior to fluid shifts would fail to reveal a depressed hematocrit, while still yielding the findings of increased low intravascular volume and increased vascular tone. Investigation would then lead to the etiologies of low cardiac output. With consideration and ruling out of a heart rate problem contributing to low CO, classes of stroke volume would then be considered. One may not yet arrive at the conclusion of hemorrhage, but would at least have the confidence that a low cardiac output due to a low preload is present, and begin infusion of fluids. A subsequent recheck of hematocrit during the fluid resuscitation would reveal a significant drop in this value, and promptly indicate hemorrhage. If blood loss was chronic or occurred more that 30 minutes prior to analysis, then interstitial to vascular fluid shifts would cause hemodilution and reveal a drop in hematocrit, and the cause of the elevated lactate and oxygen supply/demand imbalance would clearly be hemorrhage.
Cardiac output is a factor of both oxygen delivery and blood pressure; thus an abnormally low blood pressure, or a VO2/DO2 imbalance may arise from ineffective cardiac output. In reasoning through the roots of hypotension, the patient would have either a low CO or low SVR. Extremity warmth, brisk capillary refill, and hyperdynamic circulation are findings associated with a low SVR. Conversely, cold extremities, weak pulses with narrow pulse pressures, and delayed capillary refill would provide an indication that vascular tone is abnormally high, and most likely as an attempt to compensate for a low cardiac output. Low CO would again lead to consideration of heart rate and stroke volume, and causes of low stroke volume. Enlarged neck veins are characteristic of all low stroke volume syndromes except frank hypovolemia, which was previously discussed. Other classes of low stroke volume states include the following. Precardiac obstruction, as typified by tension pneumothorax and cardiac tamponade. Valvular obstruction, for example, mitral stenosis exacerbated by tachydysrhythmia. Acute heart failure as may occur with acute myocardial infarction, end-stage cardiomyopathy, acute volume overload, acute aortic insufficiency, and decompensation of pulmonary hypertension. Postcardiac obstruction such as pulmonary embolism or aortic valve stenosis compounded by heart failure. Mitral regurgitation.
Exam findings, monitor values and trends, and echocardiographic findings either alone or in combination would define which category of low SV is present. Findings from invasive monitors that differentiate the different syndromes are shown in red on Figure
In the case of low blood pressure with physical findings consistent with low vascular tone, a distributive shock picture can be assumed, and attention would then center upon further evaluation and empiric treatment of the conditions shown.
The utility of a physiologic evaluation as presented here is its ability to direct therapy appropriate to the class of abnormality once it is discovered. In distributive shock, for example, a patient with a low blood pressure clearly needs stabilization with a direct vasopressor. Because dilation of venous capacitance vessels accompanies profound arterial vasodilatation, fluids are also needed to restore preload. While the actual etiology of low vascular tone is not always apparent, the initial focus on the physiologic abnormality facilitates its rapid stabilization while diagnostic efforts continue.
Likewise, if the evaluation leads to a low cardiac output state with compensatory vasoconstriction, the physiologic analysis should attempt to differentiate whether the low CO was due to hypovolemia or one of the other classes of abnormalities discussed above. Evidence of a low oxygen delivery in the setting of volume overload indicates a need for inotropic support while the underlying condition is investigated further. For all of these cases, identifying that the patient is maximally vasoconstricted should lead away from using vasopressors, which would not create any meaningful improvement in the patient’s condition.
The tree branch diagram in Figure
Much controversy exists as to the most desirable endpoints for resuscitation from shock [
As resuscitation proceeds, it is important to continually reexamine VO2/DO2 and adequacy of blood pressure. Likewise, missed diagnoses (bleeding, myocardial infarction) and therapeutic mistakes (vasoconstrictors used instead of fluid) can be caught early in their evolution if these two physiologic relationships are frequently reevaluated. For example, a hypotensive patient with dehydration is given a vasopressor to elevate his blood pressure; a reevaluation of the serum studies showing a rising lactate and drop in pH may reveal that the initial perception of the problem was incorrect and that means of maintaining blood pressure with a greater contribution of CO over SVR should be pursued.
On critical care rounds one commonly hears statements such as: “If Mr. X becomes hypotensive tonight, we will give a vasoconstrictor.” Statements such as the above are a good entry point for a discussion of whether the low blood pressure comes from a low cardiac output or loss of vascular tone, and how one can differentiate these possibilities and their etiologies. A lecture on shock given during the ICU rotation uses the principles presented here as framework for understanding shock states and their treatment. Laminated cards containing the dendrogram in Figure
The methods used by different individuals to arrive at a diagnosis from a set of presenting conditions can be highly variable. Indeed, many texts have been written on diagnostic reasoning and the steps taken by master clinicians to solve challenging cases. Despite the breadth of work on the subject, there is no consensus on how clinical reasoning should proceed, or how students and young clinicians should be trained. There is a compelling body of evidence suggesting that in a growing number of conditions, delays and inefficiency in identifying critical problems and initiating treatment compromise survival [
This article presents an approach to the evaluation and initial treatment of critically ill patients that seems to satisfy a number of requirements posed by unstable patients at risk for organ failure. It is comprehensive in the sense that it is based on well-known physiologic alterations that underlie all known classes of shock and organ dysfunction. As such, it includes the evaluation of VO2/DO2, which is often ignored by clinicians, to the detriment of their patients. The method is efficient. Evaluation of vascular tone, blood pressure trends, hemoglobin content, acid-base status, and lactate production can be evaluated in a few minutes, while a patient’s interview and exam occurs in parallel. Diagnostic studies can be prioritized. If a stroke volume problem is suspected, one will know that obtaining an echocardiogram will be most useful in defining the etiology of the problem while physical exam can further narrow the set of possibilities. Understanding which studies contain the highest yield and building the use of diagnostic studies into a decision tree further improves efficiency and speed of care. It does not delay treatment. Gross abnormalities in vascular tone, intravascular volume can be identified in less than a minute and acted upon promptly in an effort to stabilize the patient. Additional diagnostic impressions can be gained from the response to such early treatments. For example, poor responses to vasopressor infusions may suggest concomitant adrenal insufficiency while a brisk increase in blood pressure to fluid infusion may confirm the suspicion of blood loss. It may improve diagnostic accuracy and efficiency. The key to defining deviations from normal physiologic values is that many diagnoses are coupled with specific patterns of abnormalities (see Figure Finally, physiology is a fairly universal language. The care of the critically ill often involves discussions with many different disciplines including surgeons, nephrologists, cardiologists, general internists. Fragmentation of care, however, has the other unfortunate consequence of compromising patient safety and survival [
The validity and efficacy of an educational strategy is difficult to directly evaluate especially with respect to changes in patient outcomes or other care process measure, and none has been attempted here. The tools and strategy presented have been formulated from careful consideration of the roots of critical illness, from instances where ICU admissions resulted from incomplete consideration of the relevant physiology, and from cases where slow diagnostic and therapeutic efforts have compromised patient survival [
Oxygen uptake
Oxygen delivery
Cardiac output
Mean arterial pressure
Systemic vascular resistance
Central venous pressure
Pulmonary artery occlusion (wedge) pressure
Millimeters of mercury pressure.