Library › Emergencies and critical care › Cardiac intensive care

Cardiac intensive care: ventilation, sepsis, and organ support

Care of adults in the cardiac intensive care unit: organization and staffing, heart–lung interactions and ventilation, sedation and delirium, vascular access, sepsis, and organ support from transfusion and nutrition to renal replacement and palliative care.

From coronary care unit to cardiac intensive care unit

The coronary care unit of the 1960s was built around one idea: continuous electrocardiographic monitoring of patients with acute myocardial infarction by nurses trained to defibrillate. Ventricular fibrillation in the first days after infarction became a treatable event, and the unit became the place where reperfusion, antithrombotic therapy, and the early management of infarction developed. As primary percutaneous coronary intervention shortened hospital stays and moved uncomplicated infarction to step-down beds, the population of the unit changed.

In an academic coronary care unit followed from 1989 through 2006 (29,275 admissions), admission severity rose, ST-elevation infarction declined, non-ST-elevation infarction increased, and noncardiovascular diagnoses such as sepsis, acute kidney injury, and respiratory failure grew fastest. Mechanical ventilation and central venous catheterization were used increasingly often. Medicare data for 3.4 million admissions with a cardiac intensive care stay between 2003 and 2013 showed that the share of patients with a primary noncardiac discharge diagnosis rose from 38.0% to 51.7%. Infectious diseases doubled as a principal diagnosis (7.8% to 15.1%), coronary artery disease fell from 32.3% to 19.0%, and comorbid heart failure (13.9% to 34.4%) and renal failure (7.1% to 19.6%) became far more common. Patients with primary noncardiac diagnoses received more noncardiac procedures and had higher risk-adjusted mortality.

The American Heart Association (AHA) recognized this shift in a 2012 scientific statement that adopted the term cardiac intensive care unit (CICU) and described its patients as having both cardiovascular and multisystem critical illness. The statement called for staffing and training models that combine cardiology and critical care expertise. A 2020 AHA statement then addressed the preventable complications of CICU care (ventilator-associated pneumonia, catheter infection, delirium, venous thromboembolism, bleeding, and others) and noted that most of the evidence comes from general intensive care populations and is applied to cardiac patients by extrapolation.

Case mix of the contemporary cardiac intensive care unit

The Critical Care Cardiology Trials Network registry described consecutive admissions to 16 advanced tertiary CICUs in the United States and Canada in 2017 and 2018. Of 3,310 admissions, 77.3% were for primary cardiac problems, 10.2% for postprocedural care, 7.7% for mixed general and cardiac problems, and 4.9% for overflow from general medical units. The median age was 65 years, and noncardiovascular comorbidities were common.

Critical Care Cardiology Trials Network registry, 16 tertiary CICUs, 2017–2018. Respiratory insufficiency and shock were the most common indications for CICU care and carried a poor prognosis.

The CICU physician manages the ventilator, sedation, sepsis, renal replacement therapy, nutrition, and end-of-life care as routinely as hemodynamics and arrhythmia. Risk in a CICU population is concentrated in a minority: in a Mayo Clinic cohort of 9,898 CICU patients, the 32% with a critical care diagnosis (cardiac arrest, shock, respiratory failure, or sepsis) accounted for more than 85% of hospital deaths.

Older adults are a growing share of this population. Age changes the response to every CICU drug: beta-adrenergic responsiveness declines, arterial stiffness rises, diastolic relaxation slows, and creatinine-based estimates overstate kidney function in patients with low muscle mass. Frailty, measured with instruments such as the Clinical Frailty Scale, predicts mortality and dependence at discharge independently of age. Age alone does not exclude revascularization or mechanical support; frailty, cognition, and the patient's stated goals determine whether intensive treatment serves the patient.

Levels of care, staffing models, and the cardiac intensivist

The 2012 AHA statement proposed three descriptive levels of CICU. Level 1 units provide the most advanced care, and for them the writing group favored care directed by a cardiac intensivist who combines cardiovascular and critical care expertise. Level 2 units fall between the others and may use a high-intensity staffing model or intensivist staffing for selected patients. In Level 3 units, cardiologists care for most patients, and intensivists are available for consultation, for example when new respiratory failure requires mechanical ventilation. The 2025 AHA update retained three levels: Level 1 destination centers for comprehensive care, Level 2 secondary referral centers, and Level 3 community units. It replaced the term "closed unit" with high-intensity, CICU-based staffing and named that model best practice for Level 1 units. It listed nurse-to-patient ratios of 1:1 to 1:2 in Level 1 units and 1:1 to 1:3 in Level 2 and 3 units.

Staffing categories described in the 2012 and 2020 AHA statements. High-intensity staffing is defined by a closed unit or mandatory intensivist consultation.

Direct evidence for cardiac intensivist staffing is observational. When a Korean tertiary CICU changed in 2013 from low-intensity staffing to a high-intensity model led by a dedicated cardiac intensivist, CICU mortality fell from 8.9% (55 of 616 patients) to 4.1% (74 of 1,815). In a propensity-matched cohort the rates were 7.5% and 3.7% (adjusted odds ratio 0.53; 95% confidence interval [CI] 0.32 to 0.86), and among the 135 patients treated with extracorporeal membrane oxygenation (ECMO), mortality fell from 54.5% to 22.5%. Length of stay and readmission did not change. Before-and-after designs cannot separate the staffing change from concurrent changes in practice.

In the United States, the traditional route to dual expertise is sequential fellowship training in cardiovascular disease and critical care medicine, which requires at least 4 years of total fellowship training for board eligibility in both specialties. Shorter exploratory pathways in cardiac critical care that do not lead to board certification are also offered, and the 2025 AHA statement considers them better suited to Level 2 than Level 1 units. Training competencies are defined in the Core Cardiovascular Training Statement (COCATS-4); curricula vary widely across programs, and a dedicated competency statement for critical care cardiology is in development.

The unit runs on more than physicians. Nurses with cardiac critical care competence, critical care pharmacists, respiratory therapists, perfusionists for extracorporeal support, dietitians, physical and occupational therapists, and palliative care specialists each own parts of the bundles described below. The 2022 compendium on central line infection prevention classifies an adequate nurse-to-patient ratio, with limited use of float nurses, as an essential practice (high-quality evidence). Structured multidisciplinary rounds with a daily goals checklist deliver these elements.

Shock teams. Multidisciplinary shock teams, their activation criteria, and regional transfer to Level 1 centers are covered on the cardiogenic shock page.

Severity scoring and mortality prediction in the CICU

General intensive care scores perform reasonably in CICU patients. In 9,961 unselected Mayo Clinic CICU patients (hospital mortality 9.0%), the day 1 Sequential Organ Failure Assessment (SOFA) score discriminated hospital mortality with an area under the curve of 0.83, similar to the Acute Physiology and Chronic Health Evaluation (APACHE) III and IV. The mean and maximum SOFA score over several days discriminated better than the day 1 score, a rising score from day 1 to day 2 identified higher risk, and a day 1 score below 2 identified low risk. Discrimination differed by diagnosis: the area under the curve was 0.85 to 0.86 in acute coronary syndromes, 0.76 to 0.79 in heart failure, and only 0.68 to 0.78 in patients with critical care diagnoses such as shock, cardiac arrest, respiratory failure, or sepsis.

Scores describe populations. None is accurate enough to decide whether an individual patient receives or continues life support.

Heart–lung interactions: pleural pressure, venous return, and ventricular afterload

The heart sits inside the thorax, so every breath changes the pressure around it. Three mechanisms account for most hemodynamic effects of ventilation: changes in pleural pressure alter venous return and left ventricular afterload; changes in lung volume alter pulmonary vascular resistance and autonomic tone; and the work of breathing itself consumes cardiac output.

Venous return. Systemic venous return is driven by the gradient between mean systemic filling pressure and right atrial pressure. Spontaneous inspiration lowers pleural and right atrial pressure and increases venous return. Positive-pressure inspiration raises pleural and right atrial pressure, narrows the gradient, and reduces venous return. The effect is largest in hypovolemia, vasodilation, and tamponade, and when positive end-expiratory pressure (PEEP) or intrinsic PEEP keeps intrathoracic pressure high throughout the cycle.

Left ventricular afterload. Left ventricular afterload is wall stress, which depends on transmural pressure: intracavitary systolic pressure minus pleural pressure. A large negative swing in pleural pressure during labored spontaneous breathing raises transmural pressure and afterload for the same arterial pressure; a patient with pulmonary edema who generates large negative pleural swings loads a failing ventricle with every breath. Positive pressure reverses this: raising pleural pressure lowers transmural pressure and unloads the left ventricle. A failing, dilated left ventricle, whose stroke volume is afterload-sensitive and preload-insensitive, therefore often gains output with positive pressure, while a normal, preload-dependent ventricle loses output.

Lung volume and the pulmonary circulation. Pulmonary vascular resistance is lowest near functional residual capacity. At low lung volumes, collapsed alveoli cause hypoxic vasoconstriction and kinking of extra-alveolar vessels; at high lung volumes, distended alveoli compress the alveolar capillaries. Resistance therefore has a U-shaped relation to lung volume, and both atelectasis and overdistension raise right ventricular afterload. Hypoxemia, hypercapnia, and acidemia add pulmonary vasoconstriction.

Work of breathing. Respiratory muscles can consume a large fraction of cardiac output during respiratory distress. Sedation and controlled ventilation transfer that output to other organs, one reason intubation can improve lactate clearance in shock.

Ventricular interdependence. The ventricles share a septum and a pericardium. A pressure- or volume-loaded right ventricle shifts the septum leftward and limits left ventricular filling; positive pressure that raises right ventricular afterload therefore lowers left ventricular output both by reducing right ventricular output and by septal shift.

The net hemodynamic effect depends on which ventricle limits output and on volume status. Hemodynamic response is reassessed after each change in PEEP or mode.

PEEP, the right ventricle, and measured filling pressures

PEEP recruits collapsed alveoli and improves oxygenation, and moderate PEEP that restores lung volume toward functional residual capacity can lower pulmonary vascular resistance. Beyond that point, PEEP and high plateau pressure overdistend alveoli, raise right ventricular afterload, and reduce venous return. A right ventricle with limited contractile reserve dilates, tricuspid regurgitation worsens, the septum shifts, and left ventricular filling falls. Acute cor pulmonale is a recognized complication of acute respiratory distress syndrome (ARDS) ventilation, and in patients with right ventricular infarction, pulmonary embolism, or pulmonary hypertension the margin is small.

Ventilator settings that protect the right ventricle in these patients are a low tidal volume, a plateau pressure kept as low as possible, the least PEEP that maintains recruitment and oxygenation, correction of hypercapnia and acidemia when lung protection allows, and avoidance of hypoxemia. Echocardiographic findings that indicate right ventricular strain from ventilation include right ventricular dilation relative to the left ventricle, septal flattening, and a falling tricuspid annular plane systolic excursion after PEEP is raised.

PEEP also changes the numbers used to manage the heart. Pressures measured in the thorax (right atrial, pulmonary artery, and wedge pressure) are referenced to atmosphere and include part of the pleural pressure. Positive pressure and high PEEP raise measured wedge and central venous pressure without an equal rise in transmural filling pressure, and measurement at end-expiration minimizes, but does not remove, the error. The fraction of airway pressure transmitted to the pleural space depends on lung and chest wall compliance: stiff lungs transmit less, a stiff chest wall or abdominal hypertension more. Temporary disconnection from PEEP to measure a "true" wedge is avoided because it causes derecruitment and hypoxemia. Wedge pressure measurement and its sources of error are covered on the right heart catheterization page.

Positive pressure and the failing left ventricle

In acute cardiogenic pulmonary edema and in decompensated systolic heart failure, positive pressure works as both a respiratory and a hemodynamic therapy. Continuous positive airway pressure (CPAP) or bilevel ventilation recruits flooded alveoli, increases functional residual capacity, reduces work of breathing, lowers venous return, and lowers left ventricular transmural pressure. Removal of positive pressure reverses all of these effects at once, which is the physiology of weaning-induced pulmonary edema.

Positive pressure becomes harmful when the left ventricle is not the limiting chamber. In dynamic left ventricular outflow obstruction (hypertrophic cardiomyopathy, some patients after aortic valve replacement, apical ballooning with basal hyperkinesis), lower preload enlarges the gradient. In severe aortic or mitral stenosis, output depends on preload and heart rate, and both can fall at induction. In tamponade and constriction, raised intrathoracic pressure further limits filling; spontaneous breathing is maintained until drainage when possible.

Pulse pressure variation, passive leg raising, and fluid responsiveness

During controlled positive-pressure ventilation, each breath transiently reduces right ventricular filling and, a few beats later, left ventricular filling. When both ventricles operate on the steep part of the Frank-Starling curve, stroke volume and arterial pulse pressure vary with the respiratory cycle; when they operate on the flat part, they vary little. Pulse pressure variation (PPV) is calculated as the difference between maximal and minimal pulse pressure over one breath divided by their mean. In 40 mechanically ventilated patients with acute circulatory failure related to sepsis, a PPV of 13% separated patients whose cardiac index rose by 15% or more after volume expansion from those whose index did not. Sensitivity was 94% and specificity was 96%.

The test is valid only under the conditions in which it was derived, and these are often absent in the CICU. Spontaneous breathing efforts, atrial fibrillation and frequent ectopy, tidal volumes below about 8 mL/kg predicted body weight, low lung compliance, open chest, and intra-abdominal hypertension all make PPV unreliable. In right ventricular failure and acute cor pulmonale, PPV can be high because of right ventricular afterload, not preload reserve, and a fluid bolus may worsen the condition.

Passive leg raising transfers venous blood from the legs and splanchnic circulation to the heart, and the effect reverses when the legs are lowered. In a meta-analysis of 21 studies (991 adults), a rise in cardiac output of about 10% or more during passive leg raising predicted fluid responsiveness with a pooled sensitivity of 85% and specificity of 91% (area under the curve 0.95). The change in pulse pressure performed less well (sensitivity 56%, specificity 83%), so output is measured in real time by echocardiographic velocity-time integral, pulse contour analysis, or another continuous method. The test remains usable with spontaneous breathing and arrhythmia. The Surviving Sepsis Campaign favors dynamic measures (passive leg raise or fluid challenge with stroke volume, stroke volume variation, pulse pressure, or PPV) over physical examination or static measures alone. Fluid responsiveness indicates only that output will rise with fluid; whether the patient needs more output, and whether the lungs and kidneys tolerate the volume, is a separate judgment. Preload assessment and the order of resuscitation are covered on the vasopressors and inotropes page.

Ventilation calculator: predicted body weight, tidal volume, driving pressure, and pulse pressure variation

Tidal volume is set by predicted body weight, which depends on sex and height and not on actual weight. Driving pressure represents the tidal volume relative to the compliance of the aerated lung.

Predicted body weight (kg) = 50 (men) or 45.5 (women) + 0.91 × (height in cm − 152.4), as in the ARDS Network trial. Driving pressure = plateau pressure − PEEP, measured with an inspiratory hold in a passive patient. PPV = (PPmax − PPmin) ÷ [(PPmax + PPmin) ÷ 2] × 100. The 13% threshold comes from 40 ventilated patients with acute circulatory failure related to sepsis. PPV is interpreted only in passive patients in sinus rhythm ventilated with at least 8 mL/kg. Values from 9% to 13% form a gray zone in which fluid responsiveness is not predicted reliably.

In the ARDS Network trial, 861 patients with acute lung injury or ARDS were ventilated with an initial tidal volume of 6 mL/kg predicted body weight and a plateau pressure of 30 cm H₂O or less, or 12 mL/kg with a plateau pressure of 50 cm H₂O or less. Mortality was 31.0% and 39.8% (P = 0.007), and ventilator-free days increased. In an individual-patient analysis of 3,562 patients from nine ARDS trials, driving pressure was the ventilation variable most strongly associated with survival: each increase of about 7 cm H₂O carried a relative risk of death of 1.41 (95% CI 1.31 to 1.51), even when plateau pressure and tidal volume were within protective limits. Changes in tidal volume or PEEP were associated with survival only when they lowered driving pressure.

Oxygen, high-flow nasal oxygen, and noninvasive ventilation

Respiratory insufficiency is the most common reason for CICU care, and most of it is cardiogenic pulmonary edema, pneumonia, or both. The choice of support depends on the cause, the work of breathing, carbon dioxide retention, mental status, and hemodynamics.

Conventional oxygen. Oxygen corrects hypoxemia but does nothing for work of breathing or left ventricular afterload. Hyperoxia causes coronary and systemic vasoconstriction. In a Mayo Clinic cohort of 3,368 CICU patients without hypoxemia, admission arterial PaO₂ had a J-shaped relation with hospital mortality, with a nadir near 100 mm Hg; a PaO₂ above 300 mm Hg was associated with higher mortality (adjusted odds ratio 2.37). Oxygen is therefore titrated to saturation, and patients who are not hypoxemic do not receive it routinely. Oxygenation targets after cardiac arrest are covered on the post-cardiac arrest care page.

High-flow nasal oxygen. Heated, humidified oxygen delivered at high flow through nasal prongs supplies a precise inspired oxygen fraction, washes carbon dioxide out of the anatomic dead space, generates a small, flow-dependent positive airway pressure, and is better tolerated than a mask. In FLORALI, 310 patients with nonhypercapnic acute hypoxemic respiratory failure (most with pneumonia) were assigned to high-flow oxygen, standard oxygen, or noninvasive ventilation. Intubation rates did not differ significantly (38%, 47%, and 50%), but 90-day mortality was lower with high-flow oxygen than with standard oxygen (hazard ratio 2.01 for standard oxygen) or noninvasive ventilation (hazard ratio 2.50). For sepsis with acute hypoxemic respiratory failure, the 2026 Surviving Sepsis Campaign guideline suggests high-flow nasal oxygen over conventional oxygen and over noninvasive ventilation as the initial device (conditional recommendations, very low to low certainty). In cardiogenic pulmonary edema, its small and variable positive pressure provides less left ventricular unloading than CPAP, and comparative trials are small.

Noninvasive positive-pressure ventilation. CPAP applies one pressure throughout the cycle; bilevel ventilation adds inspiratory pressure support above an expiratory pressure, which further reduces work of breathing and lowers carbon dioxide.

In the 3CPO trial (1,069 patients with acute cardiogenic pulmonary edema), noninvasive ventilation did not lower 7-day mortality compared with standard oxygen (9.5% and 9.8%) but relieved dyspnea, acidosis, and hypercapnia faster, and CPAP and bilevel ventilation performed similarly. A 2019 Cochrane review of 24 trials found that noninvasive ventilation may reduce hospital mortality and probably reduces intubation. CPAP and bilevel ventilation are therefore interchangeable for avoiding intubation in cardiogenic pulmonary edema, and bilevel ventilation is preferred when hypercapnia or respiratory muscle fatigue is present. The trial evidence is detailed on the acute decompensated heart failure page.

Settings from the 3CPO trial protocol. Failure criteria reflect common practice; a trial of noninvasive ventilation is reassessed at the bedside early and repeatedly.

Noninvasive ventilation has a limited role in cardiogenic shock. Patients with hypoperfusion and rising lactate benefit from controlled ventilation, which removes the oxygen cost of breathing and allows safe sedation for procedures and mechanical support.

Intubation of the patient with cardiac disease

Emergency intubation is one of the most dangerous procedures in intensive care. The international INTUBE cohort (2,964 critically ill adults in 29 countries) recorded at least one major adverse peri-intubation event in 45.2% of patients: cardiovascular instability in 42.6%, severe hypoxemia in 9.3%, and cardiac arrest in 3.1%. Patients with cardiac disease are at particular risk because three changes occur within minutes: induction agents remove sympathetic tone and reduce vascular resistance and contractility; positive pressure reduces venous return and raises right ventricular afterload; and apnea causes hypoxemia and hypercapnia that raise pulmonary vascular resistance.

Randomized trials in critically ill adults define several elements of the procedure. Preoxygenation with noninvasive ventilation reduced hypoxemia during intubation compared with an oxygen mask (9.1% and 18.5%) in PREOXI (1,301 patients); cardiac arrest occurred in 0.2% and 1.1%. Bag-mask ventilation between induction and laryngoscopy reduced severe hypoxemia (oxygen saturation below 80%) compared with no ventilation in PreVent (401 patients; 10.9% and 22.8%), without a significant increase in operator-reported aspiration. A routine 500-mL fluid bolus did not reduce cardiovascular collapse in PREPARE II (21.0% with the bolus and 18.2% without). Video laryngoscopy increased first-attempt success compared with direct laryngoscopy in DEVICE (85.1% and 70.8%) without increasing severe complications. In the RSI trial (2,365 patients), ketamine and etomidate produced similar in-hospital mortality by day 28 (28.1% and 29.1%), but cardiovascular collapse during intubation was more frequent with ketamine (22.1% and 17.0%).

Phenotype-specific preparation based on the physiology above; trial data come from undifferentiated critically ill adults.

The first hour after intubation. Hypotension is most frequent in the minutes after intubation, when sedation, positive pressure, and the loss of sympathetic drive combine. A post-intubation plan prepared before induction includes lung-protective initial settings with PEEP chosen for the ventricle at risk, an analgesic infusion started before the induction agent wears off (to prevent awareness and the hypertensive, tachycardic response to an unsedated tube), confirmation of tube position by waveform capnography and chest radiograph, an arterial line if not already placed, and an orogastric tube. Patients with cardiogenic pulmonary edema often become hypotensive once positive pressure removes the work of breathing and the sympathetic response it generated.

Mechanical ventilation strategy in cardiac patients

Lung-protective ventilation. Low tidal volume (6 mL/kg predicted body weight), plateau pressure of 30 cm H₂O or less, and attention to driving pressure are the standard for ARDS and a reasonable default for most ventilated CICU patients. The 2026 Surviving Sepsis Campaign guideline suggests tidal volumes of 6 to 8 mL/kg in patients with sepsis without lung injury.

ARDS in cardiac patients. The Berlin definition of ARDS requires onset within 1 week of a known clinical insult or new or worsening respiratory symptoms; bilateral opacities not fully explained by effusions, lobar or lung collapse, or nodules; a PaO₂/FiO₂ ratio of 300 mm Hg or less with PEEP or CPAP of at least 5 cm H₂O; and respiratory failure not fully explained by cardiac failure or fluid overload. Hydrostatic and permeability edema coexist in many CICU patients: pneumonia or sepsis in a patient with heart failure, aspiration during cardiac arrest, transfusion-related lung injury, or the inflammatory response to cardiopulmonary bypass or extracorporeal circulation. Echocardiography and invasive filling pressures help decide how much of the edema is hydrostatic and therefore responds to diuresis or afterload reduction. Prone positioning for at least 16 hours a day reduced 28-day mortality in severe ARDS in PROSEVA (16.0% and 32.8%), and it also relieves right ventricular afterload. Venovenous extracorporeal membrane oxygenation is used for refractory hypoxemia, and venoarterial support for combined cardiac and respiratory failure.

Modes. Volume-controlled and pressure-controlled modes deliver equivalent protection when tidal volume and plateau pressure are controlled. Pressure support ventilation is used once the patient breathes spontaneously. High respiratory drive in spontaneous modes can generate large tidal volumes and negative pleural swings that reproduce the afterload problem of labored breathing.

Ventilation during mechanical circulatory support. With venoarterial extracorporeal membrane oxygenation, the native lungs oxygenate blood ejected by the left ventricle. Poor native lung function with recovering left ventricular ejection can deliver deoxygenated blood to the coronary and cerebral circulation (differential hypoxemia), so right radial arterial saturation is monitored. Left ventricular distension on peripheral support raises pulmonary venous pressure and causes pulmonary edema. Device management and unloading are covered on the cardiogenic shock page.

Liberation from mechanical ventilation

Liberation proceeds through daily screening for readiness, a spontaneous awakening trial, a spontaneous breathing trial, and extubation. In the Awakening and Breathing Controlled trial, pairing a daily spontaneous awakening trial with a spontaneous breathing trial increased days breathing without assistance (14.7 and 11.6 days), shortened intensive care and hospital stay, and was associated with lower 1-year mortality (hazard ratio 0.68; 95% CI 0.50 to 0.92), at the cost of more self-extubation. The spontaneous breathing trial can be performed with low-level pressure support or a T-piece. In a Spanish trial of 1,153 patients, a 30-minute breathing trial with 8 cm H₂O of pressure support led to more successful extubation than a 2-hour T-piece breathing trial (82.3% and 74.0%) without more reintubation (11.1% and 11.9%).

Cardiac patients fail breathing trials more often, and the breathing trial is a cardiac stress test. A T-piece or zero-PEEP trial reproduces the loading conditions after extubation and unmasks weaning-induced cardiac dysfunction most reliably; a trial on pressure support with PEEP may hide it. Patients older than 65 years or with underlying cardiac or respiratory disease are at high risk of extubation failure, and in HIGH-WEAN (641 such patients), high-flow nasal oxygen alternating with noninvasive ventilation immediately after extubation reduced reintubation at day 7 compared with high-flow oxygen alone (11.8% and 18.2%).

Weaning-induced pulmonary edema

Weaning-induced pulmonary edema is acute cardiogenic pulmonary edema provoked by the transition from positive-pressure to spontaneous breathing. It is a leading cause of failed weaning in patients with heart disease and is frequently unrecognized because the patient appears to fail for respiratory reasons.

Physiology. The original physiologic study (15 patients with chronic obstructive pulmonary disease and cardiovascular disease) recorded the changes during 10 minutes of spontaneous breathing after mechanical ventilation. Mean esophageal pressure fell from +5 to −2 mm Hg, cardiac index rose from 3.2 to 4.3 L/min/m², systolic pressure and heart rate rose, and left ventricular end-diastolic volume increased. Transmural pulmonary artery occlusion pressure rose from 8 to 25 mm Hg, forcing return to ventilation. After a mean of 10 days of diuresis, which reduced blood volume and lowered body weight by about 5 kg, 9 of the 15 patients were weaned without a rise in occlusion pressure.

Mechanisms by which the switch from positive-pressure to spontaneous breathing loads the heart.

Systolic dysfunction, left ventricular diastolic dysfunction with high filling pressures, mitral regurgitation that worsens with afterload, myocardial ischemia, positive fluid balance, chronic obstructive pulmonary disease (which amplifies pleural pressure swings), and obesity predispose to the syndrome. In a cohort of 81 patients undergoing 283 T-piece trials, 45% of trials failed and weaning-induced pulmonary edema caused 59% of the failures. Chronic obstructive pulmonary disease (odds ratio 8.7), structural heart disease (odds ratio 4.5), and obesity (odds ratio 3.6) were independent risk factors. Diastolic dysfunction is now recognized as a mechanism as important as systolic failure.

Diagnosis. The reference standard was a pulmonary artery occlusion pressure of 18 mm Hg or more at the end of a failed trial, and less invasive markers can substitute for the catheter.

Hemoconcentration and lung water data from a single-center study of patients who had failed a first trial; lung ultrasound data from a prospective study against expert diagnosis. Both need confirmation in larger cohorts.

Treatment. Diuresis to reduce volume is the most common intervention: in the original study it allowed most patients to wean, and in the cohort study the passive leg raising test tracked its effect. Among patients treated with diuretics after weaning-induced edema, the edema recurred in 7 of 9 patients whose leg raising test stayed negative (no preload reserve), whereas 6 of 7 whose test became positive before the next trial were extubated without recurrence. Afterload reduction with nitrates, control of hypertension and tachycardia during the trial, treatment of ischemia, rate or rhythm control of atrial fibrillation, and management of severe mitral regurgitation complete the approach. Extubation to noninvasive ventilation allows the positive pressure that unloads the left ventricle to continue for patients at high risk. Persistent failure prompts a search for an unrecognized valve lesion, ischemia, or constrictive physiology.

Ventilator-associated pneumonia and ventilator-associated events

Ventilator-associated pneumonia is pneumonia that develops more than 48 hours after intubation. It is difficult to diagnose in the CICU because cardiogenic edema, atelectasis, and post-arrest inflammation produce fever, infiltrates, and secretions that mimic infection. Ventilator-associated events, a surveillance definition based on sustained worsening of oxygenation after a period of stability, capture pneumonia together with pulmonary edema, atelectasis, and ARDS. Many of the measures that prevent ventilator-associated events act by shortening ventilation and avoiding fluid overload.

The 2022 compendium of the Society for Healthcare Epidemiology of America, the Infectious Diseases Society of America, and partner organizations classifies prevention practices for adults by strength of evidence.

Adult recommendations of the 2022 compendium on ventilator-associated pneumonia and events. Chlorhexidine oral care moved to "generally not recommended" because meta-analyses raised the possibility of higher mortality.

Aspiration during cardiac arrest and emergency intubation makes early pneumonia common in arrest survivors; the trial evidence on prophylactic antibiotics after arrest is reviewed on the post-cardiac arrest care page. Diagnosis of pneumonia in ventilated patients relies on a new or progressive infiltrate with clinical signs and a lower respiratory tract culture, and empirical antibiotics are narrowed or stopped according to culture results and the clinical course.

Analgesia-first sedation and sedation targets

Pain is the most common source of distress in ventilated patients. It comes from the endotracheal tube, suctioning, turning, lines, sternotomy and access-site wounds, and immobility, and untreated pain produces tachycardia, hypertension, and agitation that raise myocardial oxygen demand. The 2018 guideline of the Society of Critical Care Medicine on pain, agitation and sedation, delirium, immobility, and sleep disruption (PADIS) states as a good practice statement that pain is treated before a sedative is considered. It suggests an assessment-driven, protocol-based, stepwise approach to pain and sedation (conditional recommendation), light over deep sedation in ventilated adults (conditional), and propofol or dexmedetomidine over benzodiazepines (conditional).

This analgesia-first, or analgosedation, approach uses an opioid such as fentanyl or hydromorphone at the lowest effective dose, adds nonopioid analgesics (acetaminophen; neuropathic agents for neuropathic pain), and adds a sedative only when pain control alone does not achieve comfort and ventilator synchrony. Pain is measured with the patient's self-report when possible and with the Behavioral Pain Scale or Critical-Care Pain Observation Tool in patients who cannot report it; PADIS identifies these as the most valid and reliable tools. Sedation depth is measured with the Richmond Agitation-Sedation Scale (RASS), a 10-level scale from +4 (combative) to −5 (unarousable), validated in medical, surgical, cardiac surgery, coronary, and neuroscience units.

The evidence for less sedation is consistent. Daily interruption of sedative infusions in 128 medical intensive care patients shortened the median duration of ventilation from 7.3 to 4.9 days and intensive care stay from 9.9 to 6.4 days. In a Danish trial, a protocol of no sedation with bolus morphine increased days without ventilation compared with propofol and midazolam with daily interruption (13.8 and 9.6 days). Deep sedation remains appropriate for specific indications: severe ARDS with neuromuscular blockade, refractory intracranial hypertension, status epilepticus, early temperature control after cardiac arrest, and some patients on extracorporeal support.

Hemodynamic effects summarized from PADIS 2018 and the trials cited in this section. Doses are titrated to validated pain and RASS targets.

Dexmedetomidine or propofol. In SPICE III, 4,000 patients ventilated early in critical illness were assigned to dexmedetomidine as the sole or primary sedative or to usual care, with a target of RASS −2 to +1. Mortality at 90 days was identical (29.1% in both groups); bradycardia and hypotension were more common with dexmedetomidine, and 64% of its recipients needed supplemental propofol in the first 2 days. In MENDS2, 422 ventilated adults with sepsis receiving light sedation had similar days alive without delirium or coma, ventilator-free days, 90-day mortality, and 6-month cognition with dexmedetomidine (0.2 to 1.5 µg/kg/h) or propofol (5 to 50 µg/kg/min). The 2025 focused update of PADIS nevertheless suggests dexmedetomidine over propofol for ventilated adults in whom light sedation or reduction of delirium is the priority (conditional recommendation, moderate certainty), with alternatives for patients who need deep sedation or are at high risk of bradycardia. In the CICU, the bradycardia risk weighs more heavily than in general units: patients with conduction disease, beta-blockade, or recent inferior infarction may not tolerate dexmedetomidine.

Delirium in the cardiac intensive care unit

Delirium is an acute, fluctuating disturbance of attention and awareness caused by medical illness. It is classified as hyperactive (agitation, pulling at lines), hypoactive (withdrawal, inattention, somnolence), or mixed. Hypoactive delirium predominates and is missed without screening: of 566 patients who developed delirium in MIND-USA, 89% had the hypoactive form.

In 590 patients in an academic CICU with a primary cardiovascular diagnosis who had a documented Confusion Assessment Method for the ICU (CAM-ICU) score, delirium occurred in 20.3%. Delirious patients were older and sicker, stayed longer (median 5 and 2 days), and had higher mortality (27% and 3%); delirium remained independently associated with death after adjustment. In mechanically ventilated patients, delirium is far more frequent: in the CAM-ICU validation cohort from medical and coronary intensive care units, it occurred in 83.3% of patients, and it persisted at hospital discharge in 10.4%.

Risk factors specific to cardiac critical illness. Low cardiac output and cerebral hypoperfusion; microemboli from catheters, cardiac surgery, and extracorporeal circuits; hypoxemia; atrial fibrillation; post-arrest brain injury; heart failure in older patients with baseline cognitive impairment; sleep disruption from continuous monitoring and procedures; and drugs used in the CICU (benzodiazepines, opioids, anticholinergic agents, high-dose corticosteroids, and some antiarrhythmics) all contribute.

PADIS 2018 states as a good practice statement that critically ill adults are regularly assessed for delirium with a valid tool.

Prevention: the ABCDEF bundle. The ABCDEF bundle groups the elements that prevent delirium and shorten ventilation: Assess, prevent, and manage pain; Both spontaneous awakening and breathing trials; Choice of analgesia and sedation; Delirium assessment, prevention, and management; Early mobility and exercise; and Family engagement and empowerment. In the ICU Liberation Collaborative (15,226 adults in 68 units), complete bundle performance was associated with lower hospital death within 7 days (adjusted hazard ratio 0.32), less next-day mechanical ventilation, coma, and delirium, and less physical restraint use. These associations are observational, and sicker patients are less able to receive every element. Nonpharmacologic measures include reorientation, glasses and hearing aids, and a day–night light cycle.

PADIS 2018 suggests not using haloperidol, atypical antipsychotics, dexmedetomidine, statins, or ketamine to prevent delirium in all critically ill adults.

Treatment. The first step is a search for causes: hypoxemia, hypoperfusion, infection, drugs (especially benzodiazepines and anticholinergics), urinary retention, electrolyte and glucose disturbances, withdrawal, pain, and sleep deprivation. Antipsychotics do not shorten delirium. In MIND-USA, intravenous haloperidol (up to 20 mg daily) or ziprasidone (up to 40 mg daily) did not change days alive without delirium or coma compared with placebo in patients with respiratory failure or shock (7.9 days with haloperidol, 8.7 with ziprasidone, and 8.5 with placebo). In AID-ICU, haloperidol (2.5 mg three times daily plus as-needed doses, maximum 20 mg daily) did not increase days alive and out of hospital at 90 days (35.8 and 32.9 days; adjusted mean difference 2.9 days, 95% CI −1.2 to 7.0). The 2025 PADIS update made no recommendation for or against antipsychotics for delirium. Short-term antipsychotic use for distressing hyperactive symptoms or hallucinations remains common; in cardiac patients it requires a baseline and follow-up corrected QT interval (QTc) and correction of potassium and magnesium, as described on the long QT syndrome and QT prolongation page. Dexmedetomidine is suggested for ventilated adults whose agitation prevents weaning or extubation (PADIS 2018, conditional). Physical restraints increase agitation and are minimized.

Mobilization, sleep, and family presence

Immobility in the CICU begins with bed rest after infarction, femoral sheaths, intra-aortic balloon pumps, and sedation, and it continues through prolonged ventilation. Muscle mass falls quickly in critical illness, and intensive care-acquired weakness prolongs ventilation and slows recovery. PADIS 2018 suggests rehabilitation or mobilization for critically ill adults (conditional), and the 2025 focused update suggests enhanced mobilization or rehabilitation over usual care (conditional, moderate certainty), while noting that the best dose, timing, and form are not established. Mobilization proceeds through passive range of motion, sitting at the edge of the bed, standing, and walking, advanced daily according to hemodynamic stability, sedation level, and the safety of lines and devices. Large-bore femoral arterial devices limit hip flexion, and the position of every cannula is checked before and after each change of position. Mobilization is paused for new arrhythmia, rising vasopressor requirements, active ischemia, or unstable airway.

Sleep in the CICU is fragmented by alarms, light, vital sign checks, laboratory draws, imaging, and procedures at night, and by drugs that suppress rapid eye movement sleep, including benzodiazepines and opioids. Sleep deprivation contributes to delirium. PADIS 2018 suggests noise and light reduction strategies (conditional, low quality), and the 2025 update suggests melatonin (conditional, low certainty). Nighttime care is clustered, routine laboratory tests are timed for the morning, and nonurgent imaging and baths are moved out of the night.

Family presence is the "F" of the ABCDEF bundle. Open visiting, participation in rounds, and inclusion in comfort care reduce anxiety for patients and help families understand the course of illness, and family members often first detect the change in attention or behavior that signals delirium. Family members are also at risk of anxiety, depression, and post-traumatic stress after a relative's critical illness, and regular, consistent communication from the team reduces that burden.

Central venous and arterial access: indications, sites, and ultrasound guidance

Indications in the CICU. Central venous catheters deliver vasoactive drugs, inotropes, hypertonic or irritant solutions, and parenteral nutrition; provide access for transvenous pacing, renal replacement therapy, and pulmonary artery catheterization; and allow central venous pressure and saturation measurement. Arterial catheters provide beat-to-beat pressure in shock and during titration of vasoactive drugs, frequent blood gas sampling, and pulse pressure variation. Vasoactive drugs can be started through a well-functioning peripheral catheter without waiting for central access, a practice addressed with its extravasation precautions on the vasopressors and inotropes page.

Site selection. In 3SITES, 3,471 nontunneled catheters in 3,027 adult intensive care patients were randomized to the subclavian, jugular, or femoral vein. Catheter-related bloodstream infection or symptomatic deep vein thrombosis occurred at 1.5, 3.6, and 4.6 per 1,000 catheter-days. Risk was higher with femoral than subclavian catheters (hazard ratio 3.5) and with jugular than subclavian catheters (hazard ratio 2.1), but pneumothorax requiring a chest tube complicated 1.5% of subclavian and 0.5% of jugular insertions. The 2022 compendium therefore prefers the subclavian site to reduce infection when catheters are placed in the intensive care unit (high-quality evidence). In cardiac patients, other factors often decide: anticoagulation and antiplatelet therapy favor compressible sites (jugular or femoral, not subclavian); the femoral veins may be needed for mechanical support or a future procedure; and the right internal jugular vein provides the straightest route for a pacing wire or pulmonary artery catheter and is preserved when possible in candidates for endomyocardial biopsy or transplantation.

Central venous access sites compared. Infection and thrombosis rates for the jugular, subclavian, and femoral sites come from the 3SITES trial.

Ultrasound guidance. A Cochrane review of 35 trials (5,108 participants) compared two-dimensional ultrasound with landmarks for internal jugular cannulation. Ultrasound reduced total complications by 71% (risk ratio 0.29), inadvertent arterial puncture by 72% (risk ratio 0.28), and hematoma by 73%. It increased overall success by 12% and first-attempt success by 57%, and it reduced the number of attempts and the time to cannulation. Real-time guidance during the puncture was better supported than marking the vein before puncture. The American Society of Echocardiography and the Society of Cardiovascular Anesthesiologists recommend real-time ultrasound guidance for internal jugular cannulation. The 2022 compendium lists ultrasound guidance as an essential practice (high-quality evidence) and states that the probe and gel must not breach sterile technique. For arterial lines, a meta-analysis of seven trials (546 patients) found that ultrasound guidance increased first-attempt success of radial cannulation (risk ratio 1.55) and reduced hematoma. For femoral arterial access, the FAUST trial (1,004 patients) found that real-time ultrasound improved first-pass success (83% and 46%), reduced inadvertent venipuncture (2.4% and 15.8%), and reduced vascular complications (1.4% and 3.4%) compared with fluoroscopic guidance; the technique matters most for large-bore arterial access for mechanical circulatory support.

Technique and verification. Ultrasound identifies the vein, its patency, its relation to the artery, and the needle tip in short- and long-axis views. Guidewire position in the vein is confirmed by ultrasound before dilation, because inadvertent arterial dilation with a large catheter is the most serious mechanical complication. The catheter tip is positioned at the cavoatrial junction for jugular and subclavian catheters and confirmed by chest radiograph or ultrasound, and transduced pressure or blood gas confirms venous placement when there is doubt.

Infection prevention. The 2022 compendium on central line-associated bloodstream infection lists as essential practices an all-inclusive insertion cart or kit, maximal sterile barrier precautions, alcoholic chlorhexidine skin preparation (containing at least 2% chlorhexidine gluconate) allowed to dry before puncture, chlorhexidine-containing dressings, transparent dressing changes at least every 7 days or when soiled, disinfection of hubs before access, and daily review with prompt removal of catheters that are no longer essential.

Arterial catheter complications. In a review of more than 25,000 radial, femoral, and axillary arterial catheterizations, major complications such as permanent ischemic damage, sepsis, and pseudoaneurysm occurred in fewer than 1%, with similar rates at the three sites. Temporary radial occlusion is more common and usually silent. Arterial catheters are a source of bloodstream infection and are managed with the same aseptic standards as venous catheters.

Hemodynamic monitoring and point-of-care ultrasound

Pulmonary artery catheter. Randomized trials in heart failure (ESCAPE), general critical illness (PAC-Man), and high-risk surgery showed no benefit from routine catheterization in patients whom clinicians were willing to manage without one. Its value lies in patients whose hemodynamics cannot be inferred at the bedside: mixed or unclear shock, cardiogenic shock with escalating support, right ventricular failure, and evaluation for advanced therapies. Trials, indications, and waveform interpretation are covered on the right heart catheterization page, and its use to classify and manage shock on the cardiogenic shock page.

Echocardiography and point-of-care ultrasound. Focused cardiac ultrasound performed by the treating physician answers a limited set of questions at the bedside: global left and right ventricular function, pericardial effusion and tamponade physiology, gross valve dysfunction, inferior vena cava size and variation, and response to an intervention. Lung ultrasound distinguishes interstitial edema (multiple bilateral B-lines), consolidation, effusion, and pneumothorax (absent lung sliding with a lung point), and serial B-line counts track decongestion and weaning-induced edema. Comprehensive transthoracic or transesophageal echocardiography by an echocardiographer is needed for valve quantification, prosthetic valve function, endocarditis, mechanical complications of infarction, and cannula position during mechanical support. Venous congestion assessment with Doppler of the hepatic, portal, and intrarenal veins is discussed on the acute decompensated heart failure page. Normal values are listed on the transthoracic echocardiography normal values page.

Arterial pressure monitoring. Noninvasive cuff measurements are reliable at moderate pressures but less accurate in profound hypotension, arrhythmia, obesity, and vasoconstriction. The 2026 Surviving Sepsis Campaign (SSC) guideline accepts invasive or noninvasive monitoring in septic shock and favors an arterial catheter with intermediate-to-high or escalating vasopressor doses, multiple vasopressors, frequent blood sampling, or inconsistent noninvasive readings. In patients with intra-aortic balloon pumps, the augmented diastolic pressure is read from the arterial waveform. Mean pressure is the value used for targets because it is the least affected by catheter damping and site.

Laboratory perfusion markers. Lactate, central or mixed venous oxygen saturation, the venous-to-arterial carbon dioxide gap, urine output, and capillary refill describe the adequacy of perfusion. Lactate rises with hypoperfusion but also with epinephrine (beta-2 stimulation of aerobic glycolysis), liver failure, seizures, and metformin, so its trend is interpreted with the clinical context.

Sepsis and septic shock in the cardiac intensive care unit

Infection is one of the fastest-growing principal diagnoses in CICU patients, and sepsis enters the unit in three ways: as the admitting problem in a patient with heart disease, as a complication of cardiac critical care (pneumonia, catheter infection, device infection, endocarditis), and as a mimic or companion of cardiogenic shock.

Definitions. The Third International Consensus Definitions (Sepsis-3) define sepsis as life-threatening organ dysfunction caused by a dysregulated host response to infection, operationalized as an acute increase in SOFA score of 2 points or more, which is associated with in-hospital mortality above 10%. Septic shock is a subset with profound circulatory, cellular, and metabolic abnormalities, identified by a vasopressor requirement to maintain a mean arterial pressure (MAP) of 65 mm Hg or more and a serum lactate above 2 mmol/L despite adequate volume resuscitation; this combination carries hospital mortality above 40%. "Severe sepsis" and the systemic inflammatory response syndrome criteria are no longer part of the definition.

Initial management. The 2026 SSC guideline, developed by the Society of Critical Care Medicine and the European Society of Intensive Care Medicine, retains the core elements of early care: screening; blood cultures before antimicrobials whenever this does not delay therapy; prompt empirical antimicrobials (with coverage for multidrug-resistant or anaerobic organisms only when risk is high); source control; an initial crystalloid bolus for sepsis-induced hypotension followed by early vasopressors if hypotension persists; and an initial MAP target of 65 mm Hg (60 to 65 mm Hg at age 65 years or older). New recommendations in the 2026 guideline include antimicrobial administration during ambulance or air transport for definite or probable septic shock, when in-hospital evaluation is expected to take more than 60 minutes and a structured prehospital sepsis screening process is in place. They also include pathogen-specific rapid diagnostic tests in selected patients according to local epidemiology. After an initial 30 mL/kg, either a liberal or a restrictive fluid strategy is acceptable, guided by dynamic measures of fluid responsiveness. The 2026 SSC guideline suggests balanced crystalloids over 0.9% saline for initial resuscitation (conditional recommendation, moderate certainty), and 0.9% saline for patients with traumatic brain injury.

In cardiac patients, the fixed-volume bolus is applied with judgment. A patient with a dilated, poorly contracting left ventricle, severe diastolic dysfunction, or right ventricular failure may develop pulmonary edema or worse right ventricular dilation before reaching the nominal volume, and early vasopressors, smaller boluses with reassessment, and echocardiography to define the cardiac phenotype protect these patients.

Vasopressors. Norepinephrine is the first vasopressor; vasopressin is added as norepinephrine requirements escalate, and epinephrine is added as a third agent. In septic shock with cardiac dysfunction, the 2026 SSC guideline suggests norepinephrine or epinephrine as the first vasopressor and, for persistent hypoperfusion, adding dobutamine to norepinephrine or using epinephrine alone (conditional, very low certainty); it suggests against levosimendan. Corticosteroids are suggested for septic shock with ongoing vasopressor requirement. The full table of SSC vasoactive recommendations and their evidence is on the vasopressors and inotropes page.

Mixed septic and cardiogenic shock. The treatments diverge: septic shock with preserved output needs vasoconstriction and source control, whereas cardiogenic shock may need inotropes, mechanical support, and revascularization. Echocardiography and, when the picture remains unclear, a pulmonary artery catheter separate a low systemic vascular resistance with high output from a low output with high filling pressures, and identify the combination. Prolonged cardiogenic shock acquires a vasodilatory, inflammatory component, and septic shock can unmask or cause ventricular failure.

Infection prevention and stewardship. The CICU's major hospital-acquired infections (ventilator-associated pneumonia, catheter-related bloodstream infection, catheter-associated urinary tract infection, and Clostridioides difficile infection) are prevented by the bundles above, daily removal of unnecessary devices, and antibiotic stewardship: cultures before therapy, narrowing to the organism, defined treatment durations, and stopping antibiotics when infection is excluded. The 2026 SSC guideline suggests selective decontamination of the digestive tract for ventilated patients with sepsis in units with low antimicrobial resistance, a recommendation debated because of its potential ecological effect. Infection of cardiac devices, prosthetic valves, and mechanical support cannulas is a specific CICU hazard; endocarditis is covered on the infective endocarditis page.

Sepsis-induced cardiomyopathy

Sepsis-induced cardiomyopathy is acute, reversible dysfunction of one or both ventricles caused by sepsis, in the absence of an alternative cause such as acute coronary occlusion. No single accepted definition exists, which explains the wide range of reported incidence: estimates depend on whether dysfunction is defined by ejection fraction, strain, diastolic indices, or right ventricular measures, and on when echocardiography is performed relative to resuscitation and vasopressors.

Natural history. The classic radionuclide study of 20 patients with septic shock found that 10 had an ejection fraction below 40% despite a normal or high cardiac index, with dilated ventricles maintaining stroke volume. Paradoxically, 10 of 13 survivors but none of 7 nonsurvivors had an initial ejection fraction below 40%, and in survivors the ejection fraction and volumes returned to normal within about 10 days. Ventricular dilation that preserves stroke volume was interpreted as an adaptive response; nonsurvivors with normal-sized ventricles could not use it.

Mechanisms. Proposed mechanisms include circulating cytokines (tumor necrosis factor-α, interleukin-1β, interleukin-6) and damage-associated molecular patterns, inducible nitric oxide synthase and peroxynitrite, mitochondrial dysfunction with impaired oxidative phosphorylation, abnormal calcium handling, downregulation of beta-adrenergic receptors and signaling, microvascular dysfunction, and myocardial edema. Coronary blood flow is preserved or increased, and myocyte necrosis is limited, which fits a functional and largely reversible injury. Some features may protect the cell by reducing energy expenditure, resembling myocardial hibernation.

Echocardiographic phenotypes. Sepsis-induced cardiomyopathy appears as left ventricular systolic dysfunction (low ejection fraction, although low afterload can keep ejection fraction normal while contractility is reduced), reduced global longitudinal strain (more sensitive than ejection fraction), diastolic dysfunction with a low e′ velocity, and right ventricular dysfunction or dilation, often worsened by ventilation, ARDS, and pulmonary vasoconstriction. Ejection fraction depends heavily on loading conditions and vasopressor dose. In a meta-analysis of 7 studies (585 patients), new left ventricular systolic dysfunction defined by low ejection fraction was neither sensitive (52%) nor specific (63%) for 30-day mortality. Diastolic dysfunction and right ventricular dysfunction have been more consistently associated with mortality.

Biomarkers. Troponin is elevated in about 61% of patients with sepsis, and elevation was associated with mortality in a meta-analysis of 13 studies (relative risk 1.91; adjusted odds ratio 1.92). The pattern usually reflects supply-demand mismatch, direct myocardial injury, and reduced clearance, but an acute coronary syndrome is excluded when the electrocardiogram, regional wall motion, or clinical history suggests it. Natriuretic peptides rise with ventricular stretch, inflammation, and kidney dysfunction and do not separate septic from cardiogenic dysfunction.

Features summarized from the review literature on sepsis-induced cardiomyopathy and the Fourth Universal Definition of Myocardial Infarction. Echocardiography is repeated after resuscitation because loading conditions change the appearance.

Treatment. Treatment is source control, antimicrobials, and restoration of perfusion. Inotropes are added only for low cardiac output with persistent hypoperfusion despite adequate volume and MAP, and they are titrated to perfusion, not to an ejection fraction. In LeoPARDS (516 patients), levosimendan did not reduce organ dysfunction or 28-day mortality (34.5% and 30.9%), made successful weaning from ventilation less likely (hazard ratio 0.77), and caused more supraventricular tachyarrhythmia (3.1% and 0.4%); in a subgroup analysis of patients with biochemical evidence of cardiac dysfunction (elevated troponin I or N-terminal pro-B-type natriuretic peptide), it also produced no benefit.

Beta-blockade. Persistent tachycardia shortens diastolic filling and raises myocardial oxygen demand. In a single-center phase 2 trial of 154 patients with septic shock and a heart rate of 95/min or more on high-dose norepinephrine, esmolol titrated to a heart rate of 80 to 94/min reduced norepinephrine requirements and was associated with lower 28-day mortality (49.4% and 80.5%). This result came from an open-label trial with an unusually high control mortality. The multicenter STRESS-L trial of landiolol in 126 patients with septic shock and tachycardia on norepinephrine for more than 24 hours was stopped early for possible harm: SOFA scores did not improve, and 28-day mortality was 37.1% with landiolol and 25.4% with standard care (difference not significant). The 2026 SSC guideline suggests against beta-blockers as a treatment for septic shock (conditional recommendation, very low certainty). Tachycardia in sepsis is treated by correcting its causes: hypovolemia, fever, pain, anemia, and excessive catecholamine doses.

Mechanical circulatory support. Venoarterial extracorporeal membrane oxygenation is a rescue option for sepsis-induced cardiogenic shock with severe myocardial dysfunction. In a retrospective international cohort, 82 patients treated at five ECMO centers (cardiac index 3 L/min/m² or less or ejection fraction 35% or less, with an inotrope score of at least 75 µg/kg/min or lactate of at least 4 mmol/L) were compared with 130 similar controls from septic shock databases. Survival at 90 days was 60% with ECMO and 25% without (risk ratio for mortality 0.54), and the association persisted after propensity weighting (51% and 14%). These data come from expert centers and a nonrandomized comparison, and they apply only to patients with severe ventricular failure, not to vasoplegic septic shock with preserved cardiac output. Device selection and complications are covered on the cardiogenic shock page.

Arrhythmia, myocardial injury, and stress cardiomyopathy in critical illness

Atrial fibrillation. New-onset atrial fibrillation is common in critical illness and reflects atrial stretch, inflammation, catecholamines, electrolyte disturbance, and underlying atrial disease. In a California population cohort of 49,082 patients with severe sepsis, new-onset atrial fibrillation occurred in 5.9% and was associated with in-hospital ischemic stroke (2.6% and 0.6%; adjusted odds ratio 2.70) and with higher mortality (56% and 39%; adjusted relative risk 1.07). Treatment focuses on the trigger, potassium and magnesium repletion, and minimizing catecholamines. Rate control with beta-blockers is limited by hypotension, amiodarone is used for rate or rhythm control in shock, and urgent cardioversion is reserved for instability, though recurrence is frequent while the trigger persists. Anticoagulation decisions weigh a stroke risk that is uncertain for atrial fibrillation provoked by acute illness against a high bleeding risk; the general approach is on the atrial fibrillation page.

Myocardial injury and type 2 infarction. Troponin elevation in critically ill patients is classified by the Fourth Universal Definition of Myocardial Infarction. Myocardial injury is a troponin above the 99th percentile upper reference limit; infarction requires injury with clinical evidence of ischemia. Type 2 infarction results from an imbalance between oxygen supply and demand (hypotension, hypoxemia, anemia, tachyarrhythmia, coronary spasm) without acute atherothrombosis. Distinguishing it from type 1 infarction determines whether antithrombotic therapy and angiography are indicated; the classification is discussed on the non-ST-elevation acute coronary syndrome page.

Stress cardiomyopathy. Critical illness, surgery, and high-dose catecholamines can produce takotsubo syndrome, with regional dysfunction that extends beyond a single coronary territory. Dynamic left ventricular outflow obstruction in the apical form worsens with inotropes and hypovolemia. Its management in shock is covered on the cardiogenic shock page.

Acute kidney injury, cardiorenal syndrome, and renal replacement therapy

Acute kidney injury is defined by the Kidney Disease: Improving Global Outcomes (KDIGO) criteria as a rise in serum creatinine of 0.3 mg/dL or more within 48 hours, a rise to 1.5 times baseline or more within 7 days, or urine output below 0.5 mL/kg/h for 6 hours. In the CICU it results from low cardiac output, venous congestion, nephrotoxins (iodinated contrast, aminoglycosides, vancomycin with piperacillin-tazobactam, nonsteroidal anti-inflammatory drugs), cardiopulmonary bypass, rhabdomyolysis, hemolysis from mechanical support, and atheroembolism after catheterization. Venous congestion is as important as low flow: a raised central venous pressure lowers renal perfusion pressure and increases renal interstitial pressure, so creatinine can fall with decongestion. Worsening creatinine during effective decongestion is handled differently from worsening during hypoperfusion, as discussed on the acute decompensated heart failure page.

Timing. Two randomized trials showed that starting renal replacement therapy early in the absence of urgent indications does not improve survival, and a third defined the limit of deferral. In AKIKI, 620 patients with KDIGO stage 3 acute kidney injury who were receiving mechanical ventilation, catecholamines, or both were assigned to immediate therapy or to a delayed strategy. In the delayed strategy, therapy began only for severe hyperkalemia, metabolic acidosis, pulmonary edema, blood urea nitrogen above 112 mg/dL, or oliguria for more than 72 hours. Mortality at 60 days was 48.5% and 49.7%; 49% of the delayed group never received therapy, and catheter-related bloodstream infection was more frequent with early therapy (10% and 5%). In STARRT-AKI, 3,019 critically ill patients with severe acute kidney injury were assigned to accelerated therapy (within 12 hours of eligibility) or a standard strategy in which therapy was deferred until conventional indications developed or injury persisted beyond 72 hours. Mortality at 90 days was 43.9% and 43.7%; 38.2% of the standard group never needed therapy; and among survivors, dependence on dialysis at 90 days was more frequent with the accelerated strategy (10.4% and 6.0%), as were adverse events (23.0% and 16.5%). Deferral has limits. In AKIKI 2 (278 patients), the patients had stage 3 injury and either oliguria for more than 72 hours or blood urea nitrogen above 112 mg/dL. Waiting further until a mandatory indication or a blood urea nitrogen of 140 mg/dL gave no more days alive and free of renal replacement therapy by day 28 (median 10 and 12 days). The longer wait was associated with higher 60-day mortality in a multivariable analysis (hazard ratio 1.65; 95% CI 1.09 to 2.50).

KDIGO 2012 acute kidney injury guideline and the trials cited.

Interactions with the heart. Renal replacement therapy removes potassium, magnesium, calcium, and phosphate and can provoke arrhythmia; dialysate and replacement fluid composition and frequent electrolyte measurement prevent this. Intermittent hemodialysis causes hypotension through rapid fluid removal and osmotic shifts. Many CICU drugs are removed by continuous therapy or accumulate in kidney failure: dosing of antibiotics, anticoagulants, antiarrhythmics, and digoxin is adjusted to the delivered clearance. Ultrafiltration and continuous therapy for congestion refractory to diuretics are discussed on the acute decompensated heart failure page, and diuretic resistance on the diuretics page.

Liver and gut. Hepatic congestion and low output produce cholestatic or hypoxic hepatitis (sharp transaminase rise after a hypotensive episode), impair metabolism of drugs such as lidocaine, amiodarone, and warfarin, and cause coagulopathy. The gut mucosa is sensitive to splanchnic hypoperfusion and vasopressors; nonocclusive mesenteric ischemia presents with rising lactate, abdominal distension, and feeding intolerance in a patient on high-dose vasoconstrictors.

Stress ulcer prophylaxis in ventilated and shocked patients

Critically ill patients develop stress-related mucosal injury from splanchnic hypoperfusion. In REVISE, 4,821 invasively ventilated adults in 68 units received intravenous pantoprazole 40 mg daily or placebo. Clinically important upper gastrointestinal bleeding occurred in 1.0% and 3.5% (hazard ratio 0.30; 95% CI 0.19 to 0.47), and 90-day mortality was 29.1% and 30.9% (hazard ratio 0.94; not significant); ventilator-associated pneumonia and Clostridioides difficile infection did not increase. The 2024 guideline of the Society of Critical Care Medicine and the American Society of Health-System Pharmacists identifies coagulopathy, shock, and chronic liver disease as factors that likely increase the risk of stress-related bleeding. It states that mechanical ventilation alone is not a firm risk factor and that enteral nutrition probably lowers risk, and it suggests a proton pump inhibitor or histamine-2 receptor antagonist at low dose (for example, pantoprazole 40 mg or less daily or famotidine 40 mg or less daily) for patients with these risk factors (conditional recommendations). Prophylaxis is stopped when the risk factors resolve and before transfer out of the unit. Many CICU patients meet these criteria through shock or coagulopathy.

Transfusion thresholds in cardiac critical care

Anemia lowers arterial oxygen content and coronary oxygen delivery, and transfusion carries risks of volume overload, transfusion-related lung injury, infection, and immunomodulation. Restrictive thresholds of 7 to 8 g/dL are safe for most critically ill adults; patients with acute myocardial infarction are the exception that the evidence has not resolved.

Thresholds are triggers for a single unit followed by reassessment, not targets.

MINT did not meet its primary end point, but the confidence interval included a clinically important benefit of the liberal strategy, and harm from a restrictive strategy could not be excluded. The 2025 ACC/AHA recommendation for acute coronary syndrome accordingly departs from the restrictive thresholds of other hospitalized patients. Patients with heart failure receive each unit with attention to volume and, when needed, a diuretic. Anemia and bleeding management in acute coronary syndrome is covered on the non-ST-elevation acute coronary syndrome page.

Blood conservation reduces the need for transfusion: limiting the number and volume of diagnostic samples (small-volume tubes, closed arterial sampling systems), treating iron deficiency, and avoiding unnecessary anticoagulant and antiplatelet intensity.

Nutrition support in cardiac critical illness

Malnutrition and cardiac cachexia are common in patients with advanced heart failure, and critical illness adds catabolism, immobility, and interrupted feeding for procedures. The 2022 guideline of the American Society for Parenteral and Enteral Nutrition (ASPEN) reviewed 36 trials (20,578 participants) and found no differences in clinical outcomes between higher and lower energy intake, higher and lower protein intake, isocaloric parenteral and enteral nutrition, or the addition of supplemental parenteral nutrition. Because harm from high early energy delivery could not be excluded, it lowered the energy prescription to 12 to 25 kcal/kg/day and kept the 2016 protein suggestion of 1.2 to 2.0 g/kg/day.

Feeding in shock. Two French trials in ventilated adults receiving vasopressors for shock define the risks of feeding a hypoperfused gut. In NUTRIREA-2 (2,410 patients), early enteral and early parenteral nutrition targeting 20 to 25 kcal/kg/day produced similar 28-day mortality (37% and 35%), but enteral feeding caused more vomiting, diarrhea, and bowel ischemia (2% and less than 1%). In NUTRIREA-3 (3,044 patients), low calorie and protein targets for the first 7 days (6 kcal/kg/day and 0.2 to 0.4 g/kg/day of protein) produced similar 90-day mortality to standard targets (25 kcal/kg/day and 1.0 to 1.3 g/kg/day; 41.3% and 42.8%), with faster readiness for discharge and fewer gastrointestinal complications, bowel ischemia, and liver dysfunction. High protein delivery (2.2 g/kg/day or more) did not hasten recovery in EFFORT Protein and appeared harmful in patients with acute kidney injury.

Energy and protein ranges from the 2022 ASPEN guideline; feeding in shock from the NUTRIREA trials.

Glycemic control in cardiac critical illness

Stress hyperglycemia is frequent in CICU patients through catecholamines (endogenous and infused), cortisol, inflammation, corticosteroid therapy, and dextrose-containing infusions, and hyperglycemia on admission is associated with higher mortality in acute myocardial infarction and heart failure. Tight control, however, increases hypoglycemia without improving survival. In NICE-SUGAR, 6,104 adults expected to need at least 3 days of intensive care were assigned to a target of 81 to 108 mg/dL or 180 mg/dL or less. Mortality at 90 days was higher with intensive control (27.5% and 24.9%; odds ratio 1.14; 95% CI 1.02 to 1.28), and severe hypoglycemia (40 mg/dL or less) occurred in 6.8% and 0.5%.

The 2024 Society of Critical Care Medicine guideline on glycemic control states as a good practice statement that glycemic management protocols are started for persistent hyperglycemia of 180 mg/dL or more in critically ill adults. It suggests against titrating insulin to a lower target of 80 to 139 mg/dL in favor of a conventional range of 140 to 200 mg/dL (conditional, moderate certainty). It also suggests continuous intravenous insulin over intermittent subcutaneous insulin, with glucose monitoring every hour or more often during instability (conditional). Insulin infusions require protocols with explicit decision support; hypoglycemia below 70 mg/dL is treated at once, and patients receiving insulin while enteral feeding is interrupted for procedures are at particular risk. Glucose after cardiac arrest is addressed on the post-cardiac arrest care page.

Venous thromboembolism prophylaxis

Immobility, central venous catheters, heart failure, infection, and mechanical support place CICU patients at high risk of venous thromboembolism, while antiplatelet therapy, procedures, thrombocytopenia, and coagulopathy place them at high risk of bleeding. The 2018 American Society of Hematology guideline for hospitalized medical patients recommends pharmacologic prophylaxis with unfractionated or low-molecular-weight heparin over none for critically ill patients (strong recommendation, moderate certainty). It suggests low-molecular-weight heparin over unfractionated heparin (conditional, moderate certainty) and pharmacologic over mechanical prophylaxis (conditional, very low certainty). Mechanical prophylaxis is used when anticoagulation is contraindicated. Adding intermittent pneumatic compression to pharmacologic prophylaxis did not reduce proximal deep vein thrombosis in PREVENT (2,003 patients; 3.9% and 4.2%).

Many CICU patients need no separate prophylaxis because they are already receiving therapeutic anticoagulation for atrial fibrillation, acute coronary syndrome, mechanical support, or a mechanical valve. Kidney function determines the choice between low-molecular-weight and unfractionated heparin, and heparin-induced thrombocytopenia is considered when the platelet count falls by half 5 to 10 days after heparin exposure; its diagnosis and management are covered on the parenteral anticoagulants and heparin-induced thrombocytopenia page. Prophylaxis is interrupted around procedures and resumed promptly afterward.

Fluid balance, electrolytes, and deresuscitation

Fluid in critical illness passes through phases: rescue resuscitation for shock, optimization while perfusion is restored, stabilization, and evacuation of the excess once the patient is stable. Cardiac patients tolerate the first phases poorly and need the last one most. A positive cumulative balance raises filling pressures, worsens gas exchange, lengthens ventilation, and contributes to weaning-induced pulmonary edema and to congestive kidney injury.

Trials in septic shock show that less fluid after initial resuscitation is safe. In CLASSIC, 1,554 intensive care patients with septic shock who had received at least 1,000 mL of fluid were assigned to restricted or standard intravenous fluid. The restrictive group received a median of 1,798 mL in the unit and the standard group 3,811 mL; 90-day mortality was 42.3% and 42.1%, and serious adverse events and days alive without life support were similar. In the CICU, the default after resuscitation is to stop maintenance infusions that are not needed, concentrate drug infusions, count the volume of drug carriers and flushes, and begin diuresis or ultrafiltration as soon as vasopressor needs allow.

Potassium. Hypokalemia from loop diuretics, catecholamines, insulin, alkalosis, and renal replacement therapy lowers the threshold for ventricular and atrial arrhythmia and prolongs the QT interval. Hyperkalemia accompanies kidney injury, acidosis, rhabdomyolysis, hemolysis on mechanical support, and drugs (mineralocorticoid antagonists, angiotensin inhibitors, heparin, trimethoprim). Supplementation to a high-normal concentration has long been routine after cardiac surgery. In TIGHT K, 1,690 patients without prior atrial arrhythmia undergoing isolated coronary bypass surgery were assigned to supplementation when serum potassium fell below 4.5 mEq/L (tight) or below 3.6 mEq/L (relaxed). New-onset atrial fibrillation in the first 120 hours occurred in 26.2% and 27.8%, a difference within the noninferiority margin, with no difference in other arrhythmias or in-hospital mortality. Patients with ischemia, prolonged QT, or ventricular arrhythmia remain a separate case, in which hypokalemia is corrected promptly.

Magnesium. Hypomagnesemia accompanies diuretics, refeeding, renal replacement therapy, and alcohol use; it causes refractory hypokalemia and predisposes to torsades de pointes, for which intravenous magnesium is the first treatment regardless of the serum level. Magnesium is repleted with care in kidney failure.

Sodium and chloride. Hyponatremia in heart failure reflects neurohormonal water retention and diuretics and is a marker of poor prognosis; its management during decongestion is covered on the acute decompensated heart failure page. Hypernatremia appears during aggressive diuresis without free water replacement and in sedated patients without access to water. Large volumes of 0.9% saline produce hyperchloremic metabolic acidosis, which is avoided by balanced crystalloids and by monitoring chloride and bicarbonate.

Phosphate and calcium. Hypophosphatemia follows refeeding, insulin, and continuous renal replacement therapy and impairs myocardial and diaphragmatic contractility, a cause of weaning failure. Ionized hypocalcemia follows massive transfusion and citrate anticoagulation of renal replacement circuits and reduces contractility and vascular tone; ionized calcium is measured, not estimated from total calcium, in these settings.

Acid–base. Lactic acidosis is treated by restoring perfusion. Sodium bicarbonate does not improve hemodynamics in lactic acidosis and adds sodium, volume, and carbon dioxide; it is reserved for specific indications such as hyperkalemia, sodium-channel blocker toxicity, and selected patients with severe acidemia and acute kidney injury. Metabolic alkalosis from loop diuretics (contraction alkalosis) blunts respiratory drive and can delay weaning; it is treated with chloride and potassium repletion, and acetazolamide in selected patients.

Patients after cardiac surgery and structural procedures

Many CICUs also receive patients after cardiac surgery, transcatheter valve procedures, complex percutaneous coronary intervention, ablation, and device implantation. Most recover without complication; the task is to recognize the few who deteriorate.

After cardiopulmonary bypass. Cardiopulmonary bypass causes a systemic inflammatory response, hemodilution, coagulopathy from heparin, platelet dysfunction, fibrinogen consumption, myocardial stunning, and vasodilation. Vasoplegic syndrome, a low systemic vascular resistance with normal or high output after bypass, is treated with norepinephrine and vasopressin, and refractory cases with methylene blue or hydroxocobalamin, as described on the vasopressors and inotropes page. Low output from stunning, graft failure, valve dysfunction, or right ventricular failure is distinguished from vasoplegia by echocardiography and pulmonary artery catheter measurements, and postcardiotomy shock that does not respond to drugs leads to temporary mechanical support, as covered on the cardiogenic shock page.

Bleeding and tamponade. Chest tube output is recorded hourly. Persistent high drainage, a falling hemoglobin, or hemodynamic deterioration prompts correction of residual heparin with protamine, of fibrinogen with cryoprecipitate, of platelet dysfunction with platelets, and of hypothermia and hypocalcemia, and surgical re-exploration when bleeding persists. Tamponade after cardiac surgery is often caused by a localized posterior or loculated clot, so the classic signs and a circumferential effusion may be absent. Sudden reduction of chest tube drainage, rising filling pressures, falling output, and rising vasopressor need suggest it, and transesophageal echocardiography is often needed when transthoracic windows are limited by dressings and air.

Other postoperative issues. Postoperative atrial fibrillation peaks on the second and third days; prophylaxis and management are covered on the atrial fibrillation page. Phrenic nerve injury paralyzes a hemidiaphragm and impairs weaning. Stroke from emboli of atheroma, air, or particulate matter is screened for with neurologic examination at each sedation interruption. Epicardial pacing wires provide temporary pacing for conduction disease after valve surgery; atrioventricular block after transcatheter aortic valve replacement is covered on the bradycardia and atrioventricular block page.

After transcatheter and percutaneous procedures. Large-bore femoral access for valve replacement or mechanical support requires frequent assessment of the access site and distal pulses, and retroperitoneal hemorrhage presents as unexplained hypotension, flank or back pain, and a falling hemoglobin before visible bleeding. New hypotension after a structural procedure raises the question of tamponade from wire or pacing lead perforation or annular injury, and of a new valve or paravalvular lesion; bedside echocardiography answers both quickly.

Special populations in the CICU

Pregnancy and the postpartum period. Cardiac critical illness in pregnancy (peripartum cardiomyopathy, pulmonary edema in valve disease, aortic dissection, pulmonary embolism, and acute coronary syndromes, often from spontaneous coronary artery dissection) is managed by a team that includes maternal–fetal medicine, obstetric anesthesia, and neonatology. Aortocaval compression is relieved by left lateral uterine displacement after about 20 weeks. Angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and angiotensin receptor–neprilysin inhibitors are avoided during pregnancy, as is warfarin in the first trimester. Positive-pressure ventilation is complicated by reduced functional residual capacity, faster desaturation, and airway edema.

Kidney failure on dialysis. Patients receiving maintenance dialysis have limited tolerance of volume, a high burden of coronary and valvular calcification, and arteriovenous fistulas that are protected from blood pressure cuffs, venipuncture, and catheterization in that arm; central venous catheters on the side of a fistula or future access are avoided. Many drugs used in the CICU require adjustment or avoidance.

Obesity. Tidal volume is set from predicted body weight, not actual weight; reduced chest wall compliance raises plateau pressure without the same risk to the lung, and higher PEEP is often needed to prevent atelectasis. Echocardiographic windows are limited, and invasive monitoring is used more often. Medication dosing uses total, ideal, or adjusted weight depending on the drug.

Durable left ventricular assist devices and implanted devices. Patients with continuous-flow left ventricular assist devices often have no palpable pulse; MAP is measured by Doppler or an arterial line. Low flow alarms raise the questions of hypovolemia, right ventricular failure, tamponade, arrhythmia, and pump thrombosis, and device management involves the implanting team. Patients with defibrillators and resynchronization devices have the device interrogated on admission for arrhythmia history, sensing, and lead integrity. Tachycardia therapies are suspended with a magnet or programming during procedures that use electrocautery, and at the end of life when the patient chooses.

Related reference pages and the daily CICU review

Cardiogenic shock, vasoactive drugs, acute decompensated heart failure, post-cardiac arrest care, and invasive hemodynamics have their own reference pages: cardiogenic shock, vasopressors and inotropes, acute decompensated heart failure, post-cardiac arrest care, and right heart catheterization and hemodynamics. The daily review below applies to every CICU patient.

A daily goals checklist that combines the elements of the 2020 AHA statement on prevention of complications in the CICU, the ABCDEF bundle, and infection prevention compendia.

Palliative care, goals of care, and withdrawal of life-sustaining treatment

Death in the CICU usually follows a decision to limit or withdraw treatment, and many patients who survive face serious functional and cognitive limitation. Palliative care is therefore part of CICU care from admission, not an alternative introduced when treatment fails, and a 2025 AHA scientific statement on critical cardiovascular illness describes strategies to integrate its principles into CICU practice. The AHA and American Stroke Association define palliative care as patient- and family-centered care that optimizes quality of life by anticipating, preventing, and treating suffering. They state that it should be integrated into the care of all patients with advanced cardiovascular disease early in the course, with attention to communication, shared decision making, advance care planning, and physical, emotional, spiritual, and psychological distress.

Communication. Structured family meetings early in the stay, repeated as the course changes, establish who speaks for the patient, what the patient valued, the prognosis with its uncertainty, and the treatment options. Prognosis in cardiac critical illness is often uncertain because ventricular recovery, transplant candidacy, and neurologic recovery after arrest may take days to become clear.

Time-limited trials. A time-limited trial is an agreement between clinicians and the patient or surrogate to use a defined therapy (for example, mechanical ventilation, temporary mechanical support, or renal replacement therapy) for a defined period with defined markers of improvement, after which the plan is reviewed. Time-limited trials fit the CICU well because many supports are bridges: to recovery, to a durable device, to transplantation, or to a decision.

Disagreement. The 2015 multisociety policy statement of critical care organizations addresses requests for treatment that clinicians believe will not achieve its goals. It recommends the term "potentially inappropriate" for treatments that have some chance of achieving the patient's goal but that clinicians believe are not justified by competing ethical considerations. It reserves "futile" for treatments that cannot achieve their physiologic goal, and it describes a fair process of dispute resolution with second opinions, ethics consultation, and the option of transfer.

Withdrawal of cardiac support. Withdrawal of mechanical ventilation follows a protocol with anticipatory treatment of dyspnea and distress with opioids and sedatives titrated to symptoms. Withdrawal of vasoactive drugs, temporary mechanical support, or a ventricular assist device usually leads to death within minutes to hours, and the family is prepared for its timing and appearance. Deactivation of implantable cardioverter-defibrillator therapies prevents painful shocks at the end of life and is ethically and legally the same as withdrawing any other life-sustaining treatment. The 2010 Heart Rhythm Society expert consensus statement specifies that patients or their surrogates may request deactivation and that a clinician with a conscientious objection transfers the task to another clinician. The same principles apply to pacing in a pacemaker-dependent patient, although death may follow deactivation quickly, so the expected course is discussed in advance with the patient or surrogate. Donation after circulatory death is discussed with the organ procurement organization before withdrawal when the patient is a potential donor. Neurologic determination of death after cardiac arrest is covered on the post-cardiac arrest care page, and palliative care and advanced therapy decisions in shock on the cardiogenic shock page.

Symptom control. Dyspnea is treated with opioids, positioning, a fan, and noninvasive ventilation when it relieves distress and matches the patient's goals; anxiety with reassurance and short-acting sedatives; and thirst with oral care. Monitoring that no longer guides treatment, laboratory testing, and alarms are stopped once comfort is the goal.

Post-intensive care syndrome and recovery after CICU care

Post-intensive care syndrome is new or worse impairment in physical, cognitive, or mental health that arises after critical illness and persists beyond the hospital stay; the term was adopted at a 2010 Society of Critical Care Medicine stakeholder conference and applies to family members (post-intensive care syndrome–family) as well as survivors. Weakness from critical illness polyneuropathy and myopathy, cognitive impairment after delirium, hypoperfusion, or cardiac arrest, and anxiety, depression, and post-traumatic stress are common, and they overlap with the deconditioning and depression of advanced heart disease.

Prevention begins in the unit with the ABCDEF bundle. After discharge, cardiac rehabilitation addresses physical recovery, and screening for cognitive impairment, depression, and post-traumatic stress identifies patients who need specific treatment. Medication reconciliation at each transition removes drugs started for critical illness (antipsychotics, sedatives, stress ulcer prophylaxis, high-dose insulin) and restarts guideline-directed cardiac therapy that was held. Survivorship after cardiac arrest is covered on the post-cardiac arrest care page.

Sources of error in cardiac intensive care

• Treating hypoxemia in a patient with labored breathing and pulmonary edema with oxygen alone, leaving the large negative pleural swings that load the failing left ventricle.
• Continuing a failing trial of noninvasive ventilation until the patient is exhausted, acidotic, and hemodynamically unstable at the time of intubation.
• Inducing anesthesia for intubation in cardiogenic shock, right ventricular failure, severe aortic stenosis, or tamponade without a vasopressor running and a plan for post-intubation hypotension.
• Raising PEEP for oxygenation in a patient with right ventricular failure or pulmonary embolism without reassessing right ventricular size and systemic pressure.
• Reading wedge or central venous pressure on positive pressure as transmural filling pressure, or disconnecting PEEP to measure it.
• Using pulse pressure variation in a patient who triggers the ventilator, is in atrial fibrillation, receives tidal volumes below 8 mL/kg, or has acute cor pulmonale.
• Calculating tidal volume from actual body weight in an obese patient.
• Attributing repeated failure of breathing trials to lung disease without testing for weaning-induced pulmonary edema, and extubating a volume-overloaded cardiac patient without a plan for postextubation support.
• Starting a sedative before treating pain, sedating to an unspecified depth, or using benzodiazepine infusions as the default sedative.
• Missing hypoactive delirium because only agitated patients are screened, or giving an antipsychotic without a QTc and electrolytes.
• Placing a central catheter by landmark technique when ultrasound is available, or dilating a vessel before confirming the guidewire is in the vein.
• Leaving central venous, arterial, and urinary catheters in place after they are no longer needed.
• Giving a fixed 30 mL/kg fluid bolus to a patient with septic shock and severe ventricular dysfunction without reassessment.
• Treating sepsis-related tachycardia with beta-blockers, or adding an inotrope because the ejection fraction is low when perfusion is adequate.
• Starting renal replacement therapy in the absence of urgent indications on the expectation that earlier is better, or deferring it indefinitely in persistent oliguria.
• Applying a 7 g/dL transfusion threshold to a patient with acute myocardial infarction, or transfusing a heart failure patient without attention to volume.
• Targeting glucose of 80 to 110 mg/dL, or stopping insulin without adjusting for interrupted feeding.
• Omitting venous thromboembolism prophylaxis after a procedure, or doubling anticoagulation in a patient already receiving a therapeutic dose.
• Continuing stress ulcer prophylaxis, antipsychotics, or sedatives after transfer out of the unit.
• Postponing discussion of goals of care until after a crisis, or describing withdrawal of support as stopping care.

References

1. Morrow DA, Fang JC, Fintel DJ, et al. Evolution of critical care cardiology: transformation of the cardiovascular intensive care unit and the emerging need for new medical staffing and training models: a scientific statement from the American Heart Association. Circulation. 2012;126:1408–1428. doi:10.1161/CIR.0b013e31826890b0
2. Fordyce CB, Katz JN, Alviar CL, et al. Prevention of complications in the cardiac intensive care unit: a scientific statement from the American Heart Association. Circulation. 2020;142:e379–e406. doi:10.1161/CIR.0000000000000909
3. Bohula EA, Katz JN, van Diepen S, et al. Demographics, care patterns, and outcomes of patients admitted to cardiac intensive care units: the Critical Care Cardiology Trials Network prospective North American multicenter registry of cardiac critical illness. JAMA Cardiol. 2019;4:928–935. doi:10.1001/jamacardio.2019.2467
4. Sinha SS, Sjoding MW, Sukul D, et al. Changes in primary noncardiac diagnoses over time among elderly cardiac intensive care unit patients in the United States. Circ Cardiovasc Qual Outcomes. 2017;10:e003616. doi:10.1161/CIRCOUTCOMES.117.003616
5. Na SJ, Chung CR, Jeon K, et al. Association between presence of a cardiac intensivist and mortality in an adult cardiac care unit. J Am Coll Cardiol. 2016;68:2637–2648. doi:10.1016/j.jacc.2016.09.947
6. Wiley BM, Zern EK. Defining training in critical care cardiology: what is the “gold standard?”. JACC Adv. 2024;3:100849. doi:10.1016/j.jacadv.2024.100849
7. Jentzer JC, Bennett C, Wiley BM, et al. Predictive value of the Sequential Organ Failure Assessment score for mortality in a contemporary cardiac intensive care unit population. J Am Heart Assoc. 2018;7:e008169. doi:10.1161/JAHA.117.008169
8. Alviar CL, Miller PE, McAreavey D, et al. Positive pressure ventilation in the cardiac intensive care unit. J Am Coll Cardiol. 2018;72:1532–1553. doi:10.1016/j.jacc.2018.06.074
9. Michard F, Boussat S, Chemla D, et al. Relation between respiratory changes in arterial pulse pressure and fluid responsiveness in septic patients with acute circulatory failure. Am J Respir Crit Care Med. 2000;162:134–138. doi:10.1164/ajrccm.162.1.9903035
10. Monnet X, Marik P, Teboul JL. Passive leg raising for predicting fluid responsiveness: a systematic review and meta-analysis. Intensive Care Med. 2016;42:1935–1947. doi:10.1007/s00134-015-4134-1
11. Acute Respiratory Distress Syndrome Network; Brower RG, Matthay MA, Morris A, et al. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med. 2000;342:1301–1308. doi:10.1056/NEJM200005043421801
12. Amato MB, Meade MO, Slutsky AS, et al. Driving pressure and survival in the acute respiratory distress syndrome. N Engl J Med. 2015;372:747–755. doi:10.1056/NEJMsa1410639
13. Gray A, Goodacre S, Newby DE, et al. Noninvasive ventilation in acute cardiogenic pulmonary edema. N Engl J Med. 2008;359:142–151. doi:10.1056/NEJMoa0707992
14. Berbenetz N, Wang Y, Brown J, et al. Non-invasive positive pressure ventilation (CPAP or bilevel NPPV) for cardiogenic pulmonary oedema. Cochrane Database Syst Rev. 2019;4:CD005351. doi:10.1002/14651858.CD005351.pub4
15. Russotto V, Myatra SN, Laffey JG, et al. Intubation practices and adverse peri-intubation events in critically ill patients from 29 countries. JAMA. 2021;325:1164–1172. doi:10.1001/jama.2021.1727
16. Casey JD, Seitz KP, Driver BE, et al. Ketamine or etomidate for tracheal intubation of critically ill adults. N Engl J Med. 2026;394:1608–1620. doi:10.1056/NEJMoa2511420
17. Lemaire F, Teboul JL, Cinotti L, et al. Acute left ventricular dysfunction during unsuccessful weaning from mechanical ventilation. Anesthesiology. 1988;69:171–179. doi:10.1097/00000542-198808000-00004
18. Liu J, Shen F, Teboul JL, et al. Cardiac dysfunction induced by weaning from mechanical ventilation: incidence, risk factors, and effects of fluid removal. Crit Care. 2016;20:369. doi:10.1186/s13054-016-1533-9 (erratum Crit Care. 2017;21:50)
19. Dres M, Teboul JL, Anguel N, et al. Extravascular lung water, B-type natriuretic peptide, and blood volume contraction enable diagnosis of weaning-induced pulmonary edema. Crit Care Med. 2014;42:1882–1889. doi:10.1097/CCM.0000000000000295
20. Ferré A, Guillot M, Lichtenstein D, et al. Lung ultrasound allows the diagnosis of weaning-induced pulmonary oedema. Intensive Care Med. 2019;45:601–608. doi:10.1007/s00134-019-05573-6
21. Klompas M, Branson R, Cawcutt K, et al. Strategies to prevent ventilator-associated pneumonia, ventilator-associated events, and nonventilator hospital-acquired pneumonia in acute-care hospitals: 2022 update. Infect Control Hosp Epidemiol. 2022;43:687–713. doi:10.1017/ice.2022.88
22. Devlin JW, Skrobik Y, Gélinas C, et al. Clinical practice guidelines for the prevention and management of pain, agitation/sedation, delirium, immobility, and sleep disruption in adult patients in the ICU. Crit Care Med. 2018;46:e825–e873. doi:10.1097/CCM.0000000000003299
23. Lewis K, Balas MC, Stollings JL, et al. A focused update to the clinical practice guidelines for the prevention and management of pain, anxiety, agitation/sedation, delirium, immobility, and sleep disruption in adult patients in the ICU. Crit Care Med. 2025;53:e711–e727. doi:10.1097/CCM.0000000000006574
24. Shehabi Y, Howe BD, Bellomo R, et al. Early sedation with dexmedetomidine in critically ill patients. N Engl J Med. 2019;380:2506–2517. doi:10.1056/NEJMoa1904710
25. Ely EW, Inouye SK, Bernard GR, et al. Delirium in mechanically ventilated patients: validity and reliability of the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU). JAMA. 2001;286:2703–2710. doi:10.1001/jama.286.21.2703
26. Pauley E, Lishmanov A, Schumann S, et al. Delirium is a robust predictor of morbidity and mortality among critically ill patients treated in the cardiac intensive care unit. Am Heart J. 2015;170:79–86. doi:10.1016/j.ahj.2015.04.013
27. Pun BT, Balas MC, Barnes-Daly MA, et al. Caring for critically ill patients with the ABCDEF bundle: results of the ICU Liberation Collaborative in over 15,000 adults. Crit Care Med. 2019;47:3–14. doi:10.1097/CCM.0000000000003482
28. Girard TD, Exline MC, Carson SS, et al. Haloperidol and ziprasidone for treatment of delirium in critical illness. N Engl J Med. 2018;379:2506–2516. doi:10.1056/NEJMoa1808217
29. Parienti JJ, Mongardon N, Mégarbane B, et al. Intravascular complications of central venous catheterization by insertion site. N Engl J Med. 2015;373:1220–1229. doi:10.1056/NEJMoa1500964
30. Brass P, Hellmich M, Kolodziej L, Schick G, Smith AF. Ultrasound guidance versus anatomical landmarks for internal jugular vein catheterization. Cochrane Database Syst Rev. 2015;1:CD006962. doi:10.1002/14651858.CD006962.pub2
31. Troianos CA, Hartman GS, Glas KE, et al. Guidelines for performing ultrasound guided vascular cannulation: recommendations of the American Society of Echocardiography and the Society of Cardiovascular Anesthesiologists. J Am Soc Echocardiogr. 2011;24:1291–1318. doi:10.1016/j.echo.2011.09.021
32. Buetti N, Marschall J, Drees M, et al. Strategies to prevent central line-associated bloodstream infections in acute-care hospitals: 2022 update. Infect Control Hosp Epidemiol. 2022;43:553–569. doi:10.1017/ice.2022.87
33. Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315:801–810. doi:10.1001/jama.2016.0287
34. Prescott HC, Antonelli M, Alhazzani W, et al. Surviving Sepsis Campaign: international guidelines for management of sepsis and septic shock 2026. Crit Care Med. 2026;54:725–812. doi:10.1097/CCM.0000000000007075 (also Intensive Care Med. 2026;52:863–936; publisher correction Intensive Care Med. 2026;52:1411–1413)
35. Parker MM, Shelhamer JH, Bacharach SL, et al. Profound but reversible myocardial depression in patients with septic shock. Ann Intern Med. 1984;100:483–490. doi:10.7326/0003-4819-100-4-483
36. Hollenberg SM, Singer M. Pathophysiology of sepsis-induced cardiomyopathy. Nat Rev Cardiol. 2021;18:424–434. doi:10.1038/s41569-020-00492-2
37. Gordon AC, Perkins GD, Singer M, et al. Levosimendan for the prevention of acute organ dysfunction in sepsis. N Engl J Med. 2016;375:1638–1648. doi:10.1056/NEJMoa1609409
38. Whitehouse T, Hossain A, Perkins GD, et al. Landiolol and organ failure in patients with septic shock: the STRESS-L randomized clinical trial. JAMA. 2023;330:1641–1652. doi:10.1001/jama.2023.20134
39. Bréchot N, Hajage D, Kimmoun A, et al. Venoarterial extracorporeal membrane oxygenation to rescue sepsis-induced cardiogenic shock: a retrospective, multicentre, international cohort study. Lancet. 2020;396:545–552. doi:10.1016/S0140-6736(20)30733-9
40. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int Suppl. 2012;2:1–138. doi:10.1038/kisup.2012.1
41. STARRT-AKI Investigators; Bagshaw SM, Wald R, Adhikari NKJ, et al. Timing of initiation of renal-replacement therapy in acute kidney injury. N Engl J Med. 2020;383:240–251. doi:10.1056/NEJMoa2000741
42. Gaudry S, Hajage D, Martin-Lefevre L, et al. Comparison of two delayed strategies for renal replacement therapy initiation for severe acute kidney injury (AKIKI 2): a multicentre, open-label, randomised, controlled trial. Lancet. 2021;397:1293–1300. doi:10.1016/S0140-6736(21)00350-0
43. Meyhoff TS, Hjortrup PB, Wetterslev J, et al. Restriction of intravenous fluid in ICU patients with septic shock. N Engl J Med. 2022;386:2459–2470. doi:10.1056/NEJMoa2202707
44. O’Brien B, Campbell NG, Allen E, et al. Potassium supplementation and prevention of atrial fibrillation after cardiac surgery: the TIGHT K randomized clinical trial. JAMA. 2024;332:979–988. doi:10.1001/jama.2024.17888
45. Cook D, Deane A, Lauzier F, et al. Stress ulcer prophylaxis during invasive mechanical ventilation. N Engl J Med. 2024;391:9–20. doi:10.1056/NEJMoa2404245
46. MacLaren R, Dionne JC, Granholm A, et al. Society of Critical Care Medicine and American Society of Health-System Pharmacists guideline for the prevention of stress-related gastrointestinal bleeding in critically ill adults. Crit Care Med. 2024;52:e421–e430. doi:10.1097/CCM.0000000000006330
47. Carson JL, Stanworth SJ, Guyatt G, et al. Red blood cell transfusion: 2023 AABB international guidelines. JAMA. 2023;330:1892–1902. doi:10.1001/jama.2023.12914
48. Carson JL, Brooks MM, Hébert PC, et al. Restrictive or liberal transfusion strategy in myocardial infarction and anemia. N Engl J Med. 2023;389:2446–2456. doi:10.1056/NEJMoa2307983
49. Rao SV, O’Donoghue ML, Ruel M, et al. 2025 ACC/AHA/ACEP/NAEMSP/SCAI guideline for the management of patients with acute coronary syndromes: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2025;151:e771–e862. doi:10.1161/CIR.0000000000001309
50. Compher C, Bingham AL, McCall M, et al. Guidelines for the provision of nutrition support therapy in the adult critically ill patient: the American Society for Parenteral and Enteral Nutrition. JPEN J Parenter Enteral Nutr. 2022;46:12–41. doi:10.1002/jpen.2267 (corrigendum JPEN J Parenter Enteral Nutr. 2022;46:1458–1459)
51. Reignier J, Plantefeve G, Mira JP, et al. Low versus standard calorie and protein feeding in ventilated adults with shock: a randomised, controlled, multicentre, open-label, parallel-group trial (NUTRIREA-3). Lancet Respir Med. 2023;11:602–612. doi:10.1016/S2213-2600(23)00092-9
52. NICE-SUGAR Study Investigators; Finfer S, Chittock DR, Su SY, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med. 2009;360:1283–1297. doi:10.1056/NEJMoa0810625
53. Honarmand K, Sirimaturos M, Hirshberg EL, et al. Society of Critical Care Medicine guidelines on glycemic control for critically ill children and adults 2024. Crit Care Med. 2024;52:e161–e181. doi:10.1097/CCM.0000000000006174
54. Schünemann HJ, Cushman M, Burnett AE, et al. American Society of Hematology 2018 guidelines for management of venous thromboembolism: prophylaxis for hospitalized and nonhospitalized medical patients. Blood Adv. 2018;2:3198–3225. doi:10.1182/bloodadvances.2018022954 (correction Blood Adv. 2023;7:1671)
55. Braun LT, Grady KL, Kutner JS, et al. Palliative care and cardiovascular disease and stroke: a policy statement from the American Heart Association/American Stroke Association. Circulation. 2016;134:e198–e225. doi:10.1161/CIR.0000000000000438
56. Bosslet GT, Pope TM, Rubenfeld GD, et al. An official ATS/AACN/ACCP/ESICM/SCCM policy statement: responding to requests for potentially inappropriate treatments in intensive care units. Am J Respir Crit Care Med. 2015;191:1318–1330. doi:10.1164/rccm.201505-0924ST
57. Needham DM, Davidson J, Cohen H, et al. Improving long-term outcomes after discharge from intensive care unit: report from a stakeholders’ conference. Crit Care Med. 2012;40:502–509. doi:10.1097/CCM.0b013e318232da75
58. Cannesson M, Le Manach Y, Hofer CK, et al. Assessing the diagnostic accuracy of pulse pressure variations for the prediction of fluid responsiveness: a "gray zone" approach. Anesthesiology. 2011;115:231–241. doi:10.1097/ALN.0b013e318225b80a
59. Sinha SS, Geller BJ, Katz JN, et al. Evolution of critical care cardiology: an update on structure, care delivery, training, and research paradigms: a scientific statement from the American Heart Association. Circulation. 2025;151:e687–e707. doi:10.1161/CIR.0000000000001300
60. Jentzer JC, van Diepen S, Alviar C, et al. Arterial hyperoxia and mortality in the cardiac intensive care unit. Curr Probl Cardiol. 2024;49:102738. doi:10.1016/j.cpcardiol.2024.102738
61. Bessière F, Khenifer S, Dubourg J, et al. Prognostic value of troponins in sepsis: a meta-analysis. Intensive Care Med. 2013;39:1181–1189. doi:10.1007/s00134-013-2902-3
62. ARDS Definition Task Force; Ranieri VM, Rubenfeld GD, Thompson BT, et al. Acute respiratory distress syndrome: the Berlin Definition. JAMA. 2012;307:2526–2533. doi:10.1001/jama.2012.5669
63. Lampert R, Hayes DL, Annas GJ, et al. HRS expert consensus statement on the management of cardiovascular implantable electronic devices (CIEDs) in patients nearing end of life or requesting withdrawal of therapy. Heart Rhythm. 2010;7:1008–1026. doi:10.1016/j.hrthm.2010.04.033
64. Bohula EA, Landzberg MJ, Menon V, et al. Palliative and end-of-life care during critical cardiovascular illness: a scientific statement from the American Heart Association. Circulation. 2025;151:e1075–e1090. doi:10.1161/CIR.0000000000001334
65. Gibbs KW, Semler MW, Driver BE, et al. Noninvasive ventilation for preoxygenation during emergency intubation. N Engl J Med. 2024;390:2165–2177. doi:10.1056/NEJMoa2313680
66. Russell DW, Casey JD, Gibbs KW, et al. Effect of fluid bolus administration on cardiovascular collapse among critically ill patients undergoing tracheal intubation: a randomized clinical trial. JAMA. 2022;328:270–279. doi:10.1001/jama.2022.9792
67. Prekker ME, Driver BE, Trent SA, et al. Video versus direct laryngoscopy for tracheal intubation of critically ill adults. N Engl J Med. 2023;389:418–429. doi:10.1056/NEJMoa2301601
68. Casey JD, Janz DR, Russell DW, et al. Bag-mask ventilation during tracheal intubation of critically ill adults. N Engl J Med. 2019;380:811–821. doi:10.1056/NEJMoa1812405
69. Frat JP, Thille AW, Mercat A, et al. High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure. N Engl J Med. 2015;372:2185–2196. doi:10.1056/NEJMoa1503326
70. Guérin C, Reignier J, Richard JC, et al. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med. 2013;368:2159–2168. doi:10.1056/NEJMoa1214103
71. Thille AW, Muller G, Gacouin A, et al. Effect of postextubation high-flow nasal oxygen with noninvasive ventilation vs high-flow nasal oxygen alone on reintubation among patients at high risk of extubation failure: a randomized clinical trial. JAMA. 2019;322:1465–1475. doi:10.1001/jama.2019.14901
72. Andersen-Ranberg NC, Poulsen LM, Perner A, et al. Haloperidol for the treatment of delirium in ICU patients. N Engl J Med. 2022;387:2425–2435. doi:10.1056/NEJMoa2211868
73. Hughes CG, Mailloux PT, Devlin JW, et al. Dexmedetomidine or propofol for sedation in mechanically ventilated adults with sepsis. N Engl J Med. 2021;384:1424–1436. doi:10.1056/NEJMoa2024922
74. Holst LB, Haase N, Wetterslev J, et al. Lower versus higher hemoglobin threshold for transfusion in septic shock. N Engl J Med. 2014;371:1381–1391. doi:10.1056/NEJMoa1406617
75. Reignier J, Boisramé-Helms J, Brisard L, et al. Enteral versus parenteral early nutrition in ventilated adults with shock: a randomised, controlled, multicentre, open-label, parallel-group study (NUTRIREA-2). Lancet. 2018;391:133–143. doi:10.1016/S0140-6736(17)32146-3
76. Morelli A, Ertmer C, Westphal M, et al. Effect of heart rate control with esmolol on hemodynamic and clinical outcomes in patients with septic shock: a randomized clinical trial. JAMA. 2013;310:1683–1691. doi:10.1001/jama.2013.278477
77. Gaudry S, Hajage D, Schortgen F, et al. Initiation strategies for renal-replacement therapy in the intensive care unit. N Engl J Med. 2016;375:122–133. doi:10.1056/NEJMoa1603017
78. RENAL Replacement Therapy Study Investigators; Bellomo R, Cass A, Cole L, et al. Intensity of continuous renal-replacement therapy in critically ill patients. N Engl J Med. 2009;361:1627–1638. doi:10.1056/NEJMoa0902413
79. VA/NIH Acute Renal Failure Trial Network; Palevsky PM, Zhang JH, O’Connor TZ, et al. Intensity of renal support in critically ill patients with acute kidney injury. N Engl J Med. 2008;359:7–20. doi:10.1056/NEJMoa0802639