Welcome to the session on shock. My name is John Cveransky, and I will be providing you with an overview of the diagnosis and management of patients with shock syndromes. The purpose of this talk is to review a common cause of admission to most types of intensive care units, namely the diagnosis and management of patients who meet shock criteria. In this talk, we will define shock both broadly and in relation to specific forms of shock. We will provide a brief overview of the individual types of shock and review general principles of managing shock, including treatments to be considered for all forms of shock, as well as specific therapy for individual forms of shock. I should note that in the multiprofessional critical care review course, there will be several discussions in different talks that include the treatment of patients with sepsis, the treatment of patients with acute myocardic shock, and the treatment of patients with trauma and traumatic shock. Let's start out with a basic question. What is shock? Well, as most of you know, it is inadequate oxygen delivery to the tissues. Shock is, in general, the physiologic state characterized by reduction in systemic tissue perfusion that results in decreased tissue oxygen delivery. It is perhaps more precise to state that shock is an imbalance between oxygen delivery and oxygen consumption, which places the patient at risk for cell death, organ damage, and if left unchecked, multisystem organ failure and death. This imbalance between oxygen delivery and oxygen consumption previously led to the idea of goal-directed therapy and supernormal oxygen delivery, although clinical trials did not bear out the benefit of doing those type treatments. It's also important to note that there is a small set of people who have impaired cellular respiration as a cause of their shock that can be seen in poisoning, but potentially also in patients with sepsis. For all forms of shock, it is helpful to consider the basics of therapy. These include ensuring the adequacy of the patient's airway in conjunction with the decision of whether or not to control the patient's work of breathing with sedation and potentially neuromuscular blockade. You want to optimize circulation by considering the individual form of shock and figuring out how to best treat this. You want to ensure that patient has adequate end-organ oxygen delivery, and this can be achieved by supplementing some of the components of oxygen delivery, including supplemental oxygen. And ideally, you want to reach the end points of adequate resuscitation, however you are managing the resuscitation, which potentially could include following the mental status of the patient, measuring capillary refill, or by using a biomarker such as lactate or SCV02. Together, these fit the ABCDE rubric for airway, control work of breathing, optimize circulation, ensuring oxygen delivery, and achieving the end points of resuscitation. Perhaps the area that is most dependent upon a clinician's skills, with perhaps the least amount of data informing this choice, is the timing of intubation. The advantage of early intubation includes minimizing the work of breathing, which can use up to 50% of a patient's oxygen consumption, and also minimize their endogenous catecholamines, at the risk of using sedatives that potentially can lower a patient's blood pressure, decrease preload, and decrease endogenous catecholamine production. Most of us were taught that volume resuscitation prior to intubation may avoid hemodynamic collapse. The PREPARE study that was published in Lancet Respiratory Medicine in 2019 did not, however, show a benefit to a 500cc crystalloid bolus prior to intubation on preventing cardiovascular collapse. The respiratory workload can be quite high in shock states, and the diaphragm is among the muscles that can have inadequate tissue perfusion in shock. Respiratory distress can also prompt agitation, which can increase oxygen demand and consumption. In the event of high respiratory muscle demand, this can contribute to the production of lactic acidosis. It is in this setting that intubation and the delivery of invasive positive pressure ventilation can decrease the work of breathing. For most forms of shock, with the possible exceptions of some forms of cardiogenic shock and decompensated right heart failure, volume resuscitation is a mainstay of optimizing circulation. There are many ways to assess the adequacy of circulation, using capillary refill, dynamic measures of volume resuscitation, such as passive leg raise. Following somebody clinically, including following heart rate and mental status, makes sense and is reasonable. There are people who follow static filling pressures, such as CVP, but many have moved on from this to more dynamic measures of volume resuscitation. There are a number of ways to maintain oxygen delivery and to measure the effectiveness of your interventions. Supplemental oxygen is perhaps the most basic method. It is important that you deliver enough but not excessive oxygen, and many clinicians will aim for a saturation somewhere between 92% to 94% and to ensure that a hemoglobin to which oxygen is bound and delivered to the tissues is adequate and greater than 7%. There are many ways to check the adequacy of oxygen delivery, including capillary refill, which the adromenous shock trial showed was non-inferior to following lactates. Following lactates is obviously a second method, and following central venous or mixed venous oxygen saturation are additional reasonable surrogates. The ultimate goal of resuscitation is to minimize the patient's morbidity and improve clinical outcomes. For short-term resuscitation, it is reasonable to follow capillary refill, oxygen, or central venous saturation. We have our second question, what are the most common forms of shock? The answer is three, which reviews the Shubin-Weil classification of shock, including cardiogenic, obstructive, hypovolemic, and distributive shock. In more details, here are the Shubin-Weil classifications. Things that can cause cardiogenic shock include acute MIs, both left ventricular and right ventricular, decompensated heart failure, dysrhythmias, and toxic ingestions. Obstructive shock includes pulmonary embolism, pericardial tamponade, tension pneumothorax, and hypovolemic shock can be broadly classified as hemorrhagic or non-hemorrhagic volume loss. Finally, distributive shock is the most common of all of these etiologies, which include septic, anaphylactic, and neurogenic shock. Classically, many of the measurements made to develop the Shubin-Weil classification were made using a pulmonary artery catheter. A pulmonary artery catheter was frequently used in the past because clinicians as a whole did a poor job of predicting cardiac output and filling pressures across all levels of training and experience. There is a substantial literature and experience in using pulmonary artery catheters, and they are still used frequently in cardiovascular intensive care units and in cardiac intensive care units. Outside of specific cardiac patients, there is little literature to support their routine use and substantial literature that shows no proven value in high-risk surgical patients, patients with shock, and patients with lung injury. In addition, to correctly interpret the data that comes from pulmonary artery catheters, there are a number of assumptions that need to be made in order to properly interpret the values, including placement in the proper west lung zone, absence of valvular issues, and normal heart compliance, or just several of them. Here are some of the classic findings derived from pulmonary artery catheters in differentiating forms of shock. Distributive shock with an elevated cardiac output and a decreased systemic vascular resistance, hypovolemic shock with a decreased cardiac output and an increased systemic vascular resistance, cardiogenic shock with a decreased cardiac output and an increased systemic vascular resistance, and obstructive shock with a decreased cardiac output and an increased systemic vascular resistance. More recently, there have been much interest in using non-invasive measures to help differentiate patients with shock. Here you can see four echo recordings showing a patient with a small LV and RV in hypovolemic shock, a large RV in a thrombus in transit for a massive PE, a circumferential pericardial effusion with movement of the ventricle in the setting of pericardial tamponade, and a poorly contracting left ventricle in the setting of cardiogenic shock. Now that we have differentiated the types of shock, let's move on to the diagnosis and management of patients with shock. Let's start out with a case, 59-year-old man who presents to the emergency room with abdominal pain, nausea, and vomiting. He's eight days status post-exploratory laparotomy in which he had a bowel resection and primary anastomosis at an outside hospital. On arrival, he's tachycardic, hypotensive, tachypneic, requiring oxygen, and febrile to 39.5. He has a high white count, hemoglobin of 9.5, and a creatinine of 2.1, and abdominal CT shows an anastomotic leak. As you have surmised, the patient, given his clinical findings, history, lab findings, this patient meets criteria for sepsis from an anastomotic leak. This patient has distributive shock from sepsis, the most common form of shock, with sepsis being the most common form of distributive shock. Anaphylaxis and neurogenic forms are other forms of distributive shock. In general, cardiac output is increased and systemic vascular resistance is decreased in patients with distributive shock. In septic shock, once the diagnosis has been made, which is sometimes the most complicated part of treating patients with shock, is making the diagnosis of the form of shock. For patients with septic shock, treatment includes volume resuscitation, early empiric antibiotics after cultures, and source control. For anaphylactic shock, general treatment includes stopping the causative agent, ensuring an airway in adequate circulation, epinephrine to treat the shock in the airways, and most importantly, antihistamines, H1 and H2 blockers, either separately or combined, which serve as a mainstay of therapy. This will give corticosteroids, and bronchodilators are useful for the treatment of bronchospasm. For neurogenic shock, treatment is primarily supportive, including managing an airway, volume resuscitation, aiming for an adequate mean arterial pressure, using vasopressors or inotropes as needed, atropine can be considered for severe bradycardia, but the role of steroids is still unclear. As previously mentioned, septic shock is the most common form of shock, and is defined by sepsis 3 as sepsis with hypotension requiring vasopressors and a lactate greater than 2, or the upper limit of normal. Patients with septic shock have an increased risk of death compared with those who don't have septic shock, for example, patients with sepsis who require vasopressors but don't have an elevated lactate. It is important to note that the sepsis 3 criteria just mentioned are associated with increased risk of mortality, but were not validated when developed to be useful in diagnosing patients with sepsis. And again, you'll hear more about this in the lecture on sepsis. While there are theoretical reasons why colloid would be a superior choice for volume resuscitation over crystalloids, there are, unfortunately, no proven benefit for the use of colloids. For some forms of colloids, such as hetastarch, there is an increased risk of renal failure and perhaps mortality in a dose-dependent fashion. Most support crystalloids as first-line therapy, including the surviving sepsis campaign. It is beyond the scope of this talk to discuss whether chloride-rich versus balanced salt solutions are the best crystalloid. I will briefly mention that the BASICS randomized trial that was published in August of 2021 did not show a benefit of balanced salt solutions over normal saline. But again, that will be discussed in other talks in this session. Anthony Rivers revolutionized the way we treated patients with sepsis in suggesting that we treat patients with sepsis the same way that we treat patients with trauma, acute MI, and CVA, namely, identify them early and treat them early. His single-center trial showed that initial early therapy with fluids and packed cells vasopressors, dobutamine, and mechanical ventilation prevented later need for delivering the same. In fact, early aggressive treatment prevented need for later aggressive treatment. After the first six hours, patients required less vasopressors, less mechanical ventilation, and less use of invasive monitoring devices. Additionally, those who received early goal-directed therapy had improved survival compared with usual care patients. What was not clear from the trial, which again was a single-center trial, whether it was the early part that was important, namely identifying and treating patients early, or the goal-directed therapy that was the important part. And towards that end, there were three large multi-center trials on three separate continents, the PROMIS trial, the PROCESS trial, and the ARISE trial that were all designed to answer this particular question. Perhaps the most effective way to look at these three trials is to examine the PRISM investigation, which was a pre-planned individual patient data meta-analysis, which again was planned by the three groups together to combine their results to be able to look at the results of all of the goal-directed therapies on different continents in one individual patient data meta-analysis. In total, there were over 3,700 patients at 138 hospitals in seven countries. You can see here the survival curve is superimposed, suggesting that there was no benefit to early goal-directed therapy when you combine the results from PROCESS, PROMIS, and ARISE. And this, in fact, parallels the lack of benefit in the individual trials. Importantly, in subgroup analyses, there were no subgroups that benefited from the early goal-directed therapy, which includes patients with worse shock, such as having a higher lactate, hypotension with a higher lactate, or a higher predicted risk of death. Perhaps the most important take-home from these trials and the meta-analysis is what happened to the usual care group. In the usual care group, patients got antibiotics roughly within an hour after emergency department presentation. Usual care included two liters of fluids prior to randomization, which happened roughly two hours after presentation to the emergency room. And the two liters of fluids that they got is just under 30 cc per kilogram. So usual care groups got actually pretty quick care. One can argue that Rivers' concept of identifying patients early led to improvements in usual care. And perhaps the best way to look at this is to compare what happened to the control groups in 2001 and 2017. In Rivers' study, those who were treated in the control group had a central venous oxygen saturation of, on average, under 50%, 49.2%. Whereas in the ARISE trial, the median SeVO2 was 70, meaning that half were lower and half were higher. In Rivers' study in 2001, the mean lactate was 7. In the ARISE trial, the median lactate was 4.3. Thus, again, one can argue that we've gotten better at identifying and treating patients in the intervening 16 to 20 years. There's a subset of patients that will remain hypotensive despite adequate volume resuscitation. And in those patients, clinicians need to add vasopressors to maintain adequate mean arterial pressure to ensure cerebral and renal blood flow, the two organs that autoregulate blood flow across a wide variety of pressures. The largest trial to test which vasopressor is best is the SOAP2 pragmatic trial, which compared norepinephrine versus dopamine in patients with all forms of shock, although the majority of these patients had septic shock. Mortality was numerically but not statistically lower in patients treated with norepinephrine. There were additionally fewer dysrhythmias in the norepinephrine-treated patients. Since the SOAP2 trial, there have also been several meta-analyses that have shown decreased mortality to the use of norepinephrine compared with dopamine as first-line therapy. In addition, the Surviving Sepsis Campaign, based on these meta-analyses and the SOAP2 trial, also recommend the use of norepinephrine as a first-line vasopressor. How about vasopressin? Well, the VAST trial compared giving higher doses of norepinephrine to patients already on norepinephrine versus adding vasopressin. There was no difference in mortality between giving the higher doses of norepinephrine versus vasopressin in conjunction with norepinephrine. In a prospectively defined stratum of less severe septic shock, the mortality was lower in the vasopressin plus norepinephrine compared with the norepinephrine alone group. Having said that, most clinicians will add vasopressin to norepinephrine when the doses of norepinephrine are higher, and there is a weak recommendation to do this in the Surviving Sepsis Campaign. How about giving vasopressin as monotherapy? The VANISH trial, which was a 2 by 2 factorial design comparing vasopressin versus norepinephrine and steroids versus no steroids to patients with shock. In this 2 by 2 factorial design, there was no difference in outcomes, including mortality and kidney injury in those randomized to vasopressin versus norepinephrine. Those patients who were randomized to vasopressin had a lower need for renal replacement therapy, but no difference in renal failure-free days. Of note, the dose of vasopressin used in this study was higher, 0.06 units per minute, than that used in the VAST trial, which was 0.03 units per minute. I should note that in this talk I am not going to discuss angiotensin II, other than to say that it was licensed primarily on its ability to raise blood pressure, but there is little data available on patient-centric outcomes with the use of this agent. Steroids have a long history of being used in septic shock. As an immune-modulating therapy, there is evidence that the use of high doses of steroids given to patients with all forms of sepsis do not work from some older studies in the 1980s. There is some data suggesting that patients with septic shock have impaired responsiveness to endogenous steroids. This is highlighted by the study by Annan and colleagues here, which grouped patients into three subgroups based on baseline cortisol and response to short cosentropin tests. This initial trial led to a study of patients with refractory shock, patients with hypotension despite vasopressors, suggesting a benefit in patients who were randomized to hydrocortisone plus fludrocortisone. This benefit included being able to come off vasopressors and a reduced 28-day mortality, and this was primarily found in those patients who did not respond to the short cosentropin test. This led to the CORTICUS trial, which had 500 patients with septic shock who were randomized to receive hydrocortisone or placebo. It's important to note that the entry criteria were a little bit different than the previous study in that patients either needed hypotension or the need for vasopressors, not both hypotension while on vasopressors. In this group of patients, hydrocortisone therapy did not improve survival, and there was no prediction of survival by the short cosentropin test. Given the differing outcomes between these two trials, two larger multicenter trials were done in the subsequent years. The largest of these trials was the ADRENAL trial, which enrolled 3,800 patients with septic shock. They enrolled patients who were on vasopressors and the ventilator for more than four hours. The intervention here was seven days of continuous infusion of hydrocortisone. There was no difference in mortality between the two groups. However, patients treated with vasopressors came off both the vasopressors and the ventilator faster. They did have a slightly higher incidence of adverse events, including hyperglycemia and hypernatremia. The APPROACH trial is a little bit more complicated. This was a two-by-two factorial design trial to assess whether hydrocortisone and fludrocortisone, given either with or without activated protein C, improved mortality at 90 days. The trial was stopped twice, once after activated protein C was removed from the market in 2011, and then the trial was terminated at 1,294 patients when the expiration date of the trial drugs was reached. The patients who were enrolled into this trial were sick, quite ill, on vasopressors greater than 0.25 for over six hours with at least two organ failures. Patients randomized to the hydrocortisone plus fludrocortisone group had a lower mortality of 43%, compared with 49.1% in the placebo-treated group at 90 days. They also came off vasopressors faster, but not the ventilator faster, compared with the placebo-treated group. Steroid-treated patients had more hyperglycemia and also numerically, but not statistically, more weakness. When you have differing results from large randomized controlled trials, it is sometimes helpful to do a systematic review and meta-analysis of the data. Here is one of the ones done after the adrenal and the approach trial that included more than 9,000 patients, 42 trials over 31 years. There was a numeric but not statistical improvement in mortality at 28 days. There was also a statistical improvement in long-term mortality. Patients treated with corticosteroids came off the ventilator and off vasopressors faster at the cost of having more weakness, more hyperglycemia, and more hypernatremia. Let's move on to the second case, 61-year-old woman presenting to the emergency department with fatigue and dyspnea. She's diabetic, obese, and has hypertension, and her husband notes that she's confused. On arrival, she's bradycardic, hypotensive, tachypneic, and her lab findings are notable for a normal white count, a normal creatinine, and an elevated troponin. Her ECG shows ST elevation in the inferior leads. She meets criteria for cardiogenic shock. Cardiogenic shock is simply defined as cardiac pump failure that leads to a decrease in cardiac output. Most common causes are myocardial infarctions, both right ventricular and left ventricular, decompensated heart failure, dysrhythmias, valvular defects, and on occasion, medications. In cardiogenic shock with ischemia, you get loss of left ventricular or right ventricular function. If 40% is lost, shock will occur. When you get decreased cardiac output, this will lead to tissue hypoxemia, lactic acidosis, after compensatory factors fail. You often get tachycardia that is reactive, which can lead to a feedback loop of increasing myocardial demand, increased ischemia, and increased infarction. The treatment of cardiogenic shock includes the treatment of the underlying cause. For acute MIs, this includes the basic treatments of aspirin, rate control if needed, opiates. Opiates are frequently given, but they can exacerbate the situation. As previously noted, the SOAP-2 trial showed increased mortality with the use of dopamine. You may need combination therapy of inotropes and vasopressors. The primary treatment, if feasible, is opening up the vessel with either a PCI or thrombolytics. For dysrhythmias, you want to treat the underlying cause, correct electrolytes, and give antiarrhythmics or electrical therapy. For RV infarcts, fluids and dobutamine, and acute MR require vasopressors and inotropes. The primary treatment of patients with shock is reperfusion, if feasible. Mechanical devices may help bridge the gap prior to reperfusion, including an intra-aortic balloon pump and an impeller. The shock trial that compared initial medical therapy in patients with cardiogenic shock with initial revascularization showed an absolute risk reduction of 9.3% in 30-day mortality. This was not statistically significant, but at 6 months, a secondary endpoint, mortality was 50% compared with 63.1%, which was statistically significant. Of note, age had a strong interaction with treatment effect, both at 30 days and 6 months, with less benefit in patients older than 75 years. As noted, there was a significant mortality benefit to early revascularization that persisted long-term, including the secondary endpoint of 13 months, with an absolute risk reduction in mortality at 13% at 6 months, giving a number needed to treat to save one life of roughly 7. How about data for mechanical support in patients with cardiogenic shock? This is a common practice, but the data supporting is less clear. This trial randomized 600 patients post-myocardial infarction with cardiogenic shock if revascularization was planned. Shock care was defined as a systolic blood pressure of less than 90 for more than 30 minutes, or a needed infusion of catecholamines to maintain a systolic blood pressure above 90. The intra-aortic balloon pump here was inserted either prior to the PCI or immediately afterwards, with the timing of insertion at the discretion of the investigator. The primary endpoint of the study was 30-day all-cause mortality. Note that 95% of the patients received primary PCI. Shown here on the y-axis is mortality. You don't need to be a statistician to see that these curves essentially overlap. There is no difference between groups in the primary outcome measure of 30-day mortality. In addition, none of the secondary outcomes were different between the groups, including renal function, creatinine, reinfarction, or stroke. More patients in the control group received a ventricular assist device, and those patients who received a ventricular assist device had a higher mortality. There was also some crossover between groups, but in sensitivity analyses, including a PERB protocol analysis, again, there was no difference between those who received the balloon pump and those that did not. The only group of patients that seemed to do better with the IABP were patients who were under 50, but this was a subgroup analysis that was not corrected for multiple comparisons. What about the impella? It comes in two sizes, a smaller one that can be placed percutaneously femorally, or a larger one that can be placed surgically in the axillary artery or the descending aorta. It is potentially an option for acute LV dysfunction that is non-valvular and does not include a VSD or free wall rupture, and additionally in which there is no aortic emergency. Here is a small study that suggests that patients treated with impella have a greater decrease in lactate compared with those treated with an intraaortic balloon pump. Of note, this is a small study with 25 patients. The PROTECT-2 trial in high-risk PCI patients randomized to the smaller impella versus an intraaortic balloon pump did not show a mortality difference between the groups or a difference in 30-day major adverse events. Let's move on to the third case, a 59-year-old male returning to the U.S. with a long flight, who develops acute dyspnea on picking up his bag. On clearing customs, he is taken to the local safety net hospital. On arrival, he is tachycardic, tachypneic, has on lab an elevated BMP intraponin, and his EKG shows sinus tachycardia with a Q and an inverted T wave in LEAD3, a normal chest radiograph. What kind of shock does this patient have? Well, he would have obstructive shock. Obstructive shock is caused by physical obstruction of the great vessels of the heart itself. Which impedes adequate filling and ejection. Obstructive shock is the grab bag of the shock categories. It includes pulmonary embolism, cardiac tamponade, tension pneumothorax, aortic stenosis. Imaging is often quite helpful to diagnose massive pulmonary embolism. On the right, you can see a proximal filling defect. And on the left, you can see a large right ventricle bowing into the left ventricle. There are a number of useful signs of pulmonary embolism on echocardiography, and here you can see four of them, including an underfilled left ventricle caused again by the RV bowing into the LV. This can sometimes show a left ventricle that has an appearance of a D or the D sign. You can see tricuspid regurgitation and a dilated RV with so-called McConnell sign. Or free wall akinesis with sparing of the apex. To treat obstructive shock, you must treat the underlying cause. In the previous example of a PE causing obstructive shock, one would want to give lytics either systemically or locally. Making the diagnosis of the cause of obstructive shock is key. Bedside sonography as well as imaging can be diagnostic. While you are preparing for definitive treatment, it is also useful to treat the patient supportively with fluids, vasopressors, or inoprote as necessary. Let's move on to the fourth and final case, an 18-year-old who crashed his motorcycle into a tree without a helmet. He is intubated in the field for GCSF8. Given a liter of fluids, he has an obvious left femoral fracture. In the emergency room, he is tachycardic, hypotensive, and intubated. His initial labs show a white count of 12, a hemoglobin of 6.7, platelets of 72, and an INR of 2.1, and a CT scan has a grade 3 liver laceration. This patient meets criteria for hypovolemic shock from blood loss. Hypovolemic shock results from decreased preload. In the most recent case, this is due to blood loss, which leads to decreased cardiac output. There are a number of potential causes of hypovolemic shock that may vary according to what type of practice you have. Trauma, GI bleeding, pancreatitis, decreased oral intake, diarrhea, nausea and vomiting, and burns are the most common causes. In hypovolemic shock, the systemic vascular resistance is high. With volume resuscitation, you generally want to replace what has been lost. Crystalloids and blood products are the most commonly used replacement fluids. If one is having blood loss, it is important to control further blood loss prior to aggressive volume resuscitation. Tourniquets, surgical, or IR or GI intervention all can help depending on the cause of hypovolemic shock. It is beyond the scope of this talk to discuss in detail the types of blood products or crystalloids, but it is worth noting that massive transfusion protocols include additional blood components other than packed cells. Again, there will be a separate talk on the treatment of trauma during this review. In conclusion, the four common forms of shock include distributive, cardiogenic, hypovolemic, and obstructive shock. Making the correct diagnosis using history, physical exam, labs, and timely imaging will help you deliver timely and appropriate care to your patients. I appreciate your listening to this session. I wish that it could have been in person. If you have any questions or suggestions for improvement, please feel comfortable sending me an email at jsevran at emory e-m-o-r-y dot e-d-u. Thank you and good luck on your exam.

shock syndromes

diagnosis

management

types of shock

septic shock

cardiogenic shock

hypovolemic shock

obstructive shock