General Critical Care, Temperature Control, and End-of-Life Decision Making in Patients Resuscitated from Cardiac Arrest

Cardiac arrest affects millions of people per year worldwide. Although advances in cardiopulmonary resuscitation and intensive care have improved outcomes over time, neurologic impairment and multiple organ dysfunction continue to be associated with a high mortality rate. The pathophysiologic mechanisms underlying the post-resuscitation disease are complex, and a coordinated, evidence-based approach to post-resuscitation care has significant potential to improve survival. Critical care management of patients resuscitated from cardiac arrest focuses on the identification and treatment of the underlying cause(s), hemodynamic and respiratory support, organ protection, and active temperature control. This review provides a state-of-the-art appraisal of critical care management of the post-cardiac arrest patient.


Introduction
For decades, the main focus of resuscitation research was the quality and effectiveness of cardiopulmonary resuscitation (CPR), which led to an increased probability of a return of spontaneous circulation (ROSC). In recent years, optimizing neurologically intact survival from cardiac arrest has been established as the primary aim of resuscitation scientists. Nevertheless, survival rates with good neurological outcomes vary substantially and depend on the location and circumstances of the arrest, the medical team's ability to restore perfusion to the heart, and the quality of post-resuscitation care [1][2][3].
The unique and complex pathophysiologic mechanisms underlying post-resuscitation disease increase the complexity of management. Patient-related factors, etiology of cardiac arrest, and initial rhythm can further complicate medical care and have a critical impact on outcomes. Unfortunately, the variation in post-cardiac arrest management and the lack of well-organized regional cardiac arrest centers remain important concerns and hamper international efforts toward the standardization of care.
Despite advances in resuscitation science, neurologic impairment, and multiple organ dysfunction cause considerable mortality and morbidity. High-quality post-resuscitation care has significant potential to reduce early mortality but requires major diagnostic and therapeutic resources and specific training. This review provides a state-of-the-art appraisal of critical care management of the post-cardiac arrest patient.

Pathophysiology
Post-cardiac arrest syndrome is a complex entity, including myocardial dysfunction, brain injury, the effects of ischemia and reperfusion, and the precipitating pathology or comorbidities as key features. Furthermore, this syndrome can involve injuries caused during the peri-arrest period and several systemic complications that may occur after * There are more α-1 adrenergic receptors in the veins than in the arteries, and they are abundant in the hepatic veins. In patients with severe hypovolemia, exogenous administration of pure α1-adrenergic receptor agonists will further constrict the already constricted hepatic veins, increasing the impedance of the outflow of blood from the splanchnic system into systemic circulation and leading to sequestration of blood within the liver (preventing auto-transfusion) [37]. In maximally vasoconstricted individuals, large doses of α1-adrenergic receptor agonists would not further increase venoconstriction and will constrict only arteries [37], aggravating the Fåhraeus effect (decaying of the relative hematocrit in small vessels as the vessel diameter decreases) and organ perfusion [38,39]. # Tidal volume should be set between 6 and 8 mL kg −1 predicted body weight (volume or pressure controlled); respiratory rate should be initially based on the underlying physiology and thereafter aimed at maintaining normocapnia or mild hypercapnia; plateau pressure should be maintained <30 cmH 2 O and corrected for intra-abdominal pressure when clinically indicated; avoid high positive end-expiratory pressures (set at ≤5 cmH 2 O in patients with ROSC and hemodynamic failure and then individualize); driving pressure should be maintained <14 cmH 2 O; mechanical power should be kept <17 J min −1 ; fraction of inspired oxygen should be titrated to maintain normoxia [40,41].

Airway and Anesthesia Management
Post-cardiac arrest patients have an increased risk of cardiovascular collapse and other complications during the peri-intubation period [42]. These physiological derangements may occur due to pre-existing circulatory failure, chronic diseases, effects of anesthetic agents, and transition to positive pressure ventilation. Although the role of skills in airway management and the presence of a second operator remain important [43,44], expertise in peri-intubation physiological optimization is crucial, and endotracheal intubation should be performed by well-trained physicians [45]. Standard periprocedural monitoring includes peripheral oxygen saturation, waveform capnography, blood pressure, electrocardiogram, heart rate, end-tidal oxygen concentration, and, whenever required, invasive monitoring [46].
Peri-intubation desaturation carries a four-fold increase in the adjusted odds of rearrest, and significant efforts should be made to improve denitrogenation and functional residual capacity and minimize shunting [42,47,48]. High-flow nasal cannula, non-invasive ventilation, and the 'ramped' position can be used to optimize the delivery of apneic oxygenation and first-pass success, provided that they do not impair hemodynamics and are tolerated by the patient [49][50][51]. In addition, endotracheal intubation is frequently associated with hemodynamic impairment, especially in post-cardiac arrest patients, and peri-procedural individualized hemodynamic optimization can decrease the risk of cardiovascular collapse [42,52,53]. Peri-intubation optimization of venous return and cardiac contractility in these individuals is crucial. Of note, fluid boluses may be ineffective, and early infusion of norepinephrine with or without an inotrope is usually required [54,55]. In patients with right ventricular failure and/or pulmonary hypertension who cannot tolerate further catecholamine-associated increases in pulmonary vascular resistance, vasopressin could be regarded as the vasopressor agent of choice.
The choice of sedative and induction agents should be also individualized based on the underlying hemodynamic status and comorbidities [42,52]. Although the merits of direct laryngoscopy vs. videolaryngoscopy for airway management in critically ill patients have been a matter of debate over the last few years, videolaryngoscopy may be of high value for overcoming anatomical difficulties but may also lead to life-threatening complications, including apnea and systemic hypotension. In clinical practice, there are several uncontrollable factors that may deteriorate a patient's physiology during videolaryngoscopy, such as secretions or blood in the airway, reflux of gastric contents, and failure to recognize profound desaturation/hypotension in the setting of clear laryngeal view [56,57].
Maintenance of anesthesia after intubation is usually achieved by a combination of hypnotics, opioids, and neuromuscular blocking agents. Although no definitive evidence exists regarding the best sedation regimen or optimal duration after cardiac arrest, patients must be adequately sedated using agents with short elimination half-life, the lowest possible doses, multimodal monitoring, and daily sedation breaks [58,59]. Midazolam and propofol are the most widely used sedatives, with the former allowing a faster neurological recovery; however, it is associated with a higher need for vasopressor therapy [60,61]. Of note, neuromuscular blocking agents are very useful and can facilitate mechanical ventilation, prevent shivering, and help achieve the target temperature quickly in comatose patients [62].
Patients treated with targeted temperature management (TTM) must be deeply sedated. However, TTM is associated with significant pharmacokinetic and pharmacodynamic alterations, which may result in sedative accumulation [63][64][65][66]. Therefore, the dosing of anesthetics may have to be reduced during TTM. The depth of anesthesia should always be monitored using frequent clinical assessment, continuous processed electroencephalogram monitoring and nerve stimulators, and should be carefully adjusted to tailoring drug administration to the individual patient.

Respiratory Management
The aim of positive pressure ventilation after ROSC is to improve oxygenation whilst minimizing circulatory impairment and other adverse events. Large tidal volumes, high respiratory rate and positive end-expiratory pressure (PEEP), and higher airway pressures may increase ventilator-induced lung injury and worsen hypotension and cerebral blood flow; therefore, they should be avoided, especially in the case of intravascular volume depletion or cardiogenic pulmonary edema [46,67]. Immediately after endotracheal intubation, patients should be mechanically ventilated using a lung-protective strategy with a tidal volume of 6-8 mL kg −1 , PEEP of ≤5 cmH 2 O, plateau pressure < 30 cmH 2 O, and driving pressure < 14 cmH 2 O. After hemodynamic optimization, PEEP levels can be titrated on an individualized basis, while mechanical power should be kept below 17 J min −1 , taking into account the driving pressure and respiratory rate [40,41,68].
Furthermore, an arterial blood gas sample should be obtained as soon as possible, and the fraction of inspired oxygen should be individualized and titrated upon ROSC to an arterial oxygen saturation of 94% to 96% or even lower in some patients [69]. Specifically, hyperoxia (i.e., arterial partial pressure of oxygen (PaO 2 ) > 100 mmHg) and hypoxemia (i.e., PaO 2 < 65 mmHg) should both be avoided. Although extreme hyperoxia is associated with poor neurologic outcomes [70][71][72][73], "critical" hypoxia does not equate to a specific oxygen concentration. Of note, many tissues function physiologically at levels equivalent to an atmosphere of 5% oxygen, and some at levels as low as 1% oxygen [74,75], and therefore, optimizing perfusion may be more important than setting a specific oxygenation target in anesthetized patients [76][77][78].
Whether oxygenation and ventilation targets should be modified in patients treated with TTM remains unknown. However, at a core temperature of 33 • C, the PaO 2 (and arterial partial pressure of carbon dioxide (PaCO 2 )) determined via an analysis of a warmed sample may be higher than the patient's actual PaO 2 (and PaCO 2 ); therefore, maintaining an arterial blood gas analysis PaO 2 between 70 and 100 mmHg may be reasonable in these patients. Furthermore, ventilation may be adjusted to maintain normocapnia (or even mild hypercapnia), especially in anesthetized patients treated with temperature control to 32-34 • C or in patients with decreased metabolism and carbon dioxide production [79].
Indeed, when the core temperature is 33 • C, the patient's actual PaCO 2 may be 6 to 7 mmHg lower than the value reported by the blood gas machine [80], while patients with chronic hypercapnia (e.g., chronic obstructive pulmonary disease) may require ventilator adjustments to achieve prehospitalization PaCO 2 values.
Notably, mild hypercapnia may improve cerebral perfusion and have anticonvulsant, anti-inflammatory, and anti-oxidant effects [81,82]. An observational cohort study of 16,542 patients found a greater likelihood of survival to discharge to home in the hypercapnic group and no difference in in-hospital mortality compared to normocapnic and hypocapnic patients [83]. Another study with 5258 cardiac arrest patients reported that unadjusted hospital mortality was the highest in the hypocapnic group (58.4%), compared with the hypercapnic (56.8%) and normocapnic (49.3%) group (p < 0.001). Analyses adjusted for age, lowest glucose, and PaO 2 revealed that hypocapnia (but not hypercapnia) was significantly associated with in-hospital mortality (p < 0.001) [84]. A post hoc analysis of the Japanese Association for Acute Medicine out-of-hospital cardiac arrest (OHCA) registry reported that severe hypocapnia, mild hypocapnia, severe hypercapnia, and exposure to both hypocapnia and hypercapnia were more likely to have a 1-month poor neurologic status compared with mild hypercapnia (reference: exposure to mild hypercapnia, respective adjusted odds ratios ( [85]. The clinical practice recommendations on the management of perioperative cardiac arrest (PERIOPCA) also recommend a lung-protective ventilation strategy (reducing tidal volume, plateau pressure, and driving pressure) and a PaCO 2 of 40-50 mmHg after perioperative cardiac arrest, especially in individuals with cerebral vasospasm or generalized atherosclerosis [86].

Circulatory Management
Myocardial stunning, vasoplegia, and capillary leaks are the main causes of circulatory failure after cardiac arrest and resuscitation, and advanced monitoring may be required in unstable patients to optimize oxygen delivery. Reversible post-cardiac arrest myocardial dysfunction, with a depressed left ventricular ejection fraction and an increased left ventricular end-diastolic pressure, may occur in the hours following ROSC, especially in patients with a longer duration of no-flow or CPR [5]. This impairment may persist for 48-72 h, and early echocardiography can quantify its extent, which may require the use of inotropes [5,[87][88][89]. Beta-adrenergic blockade may be necessary as well, as sustained catecholamine-induced β-adrenergic induction produces adverse effects relevant to postresuscitation management. Indeed, there is evidence suggesting that relative tachycardia is associated with poor neurological outcomes in post-cardiac arrest patients, independently of TTM, and with higher serum lactate levels and admission Sequential Organ Failure Assessment (SOFA) scores [90]. In another study, administration of beta-blocking agents (metoprolol i.v./per os or bisoprolol per os) during the first 72 h of post-resuscitation care was associated with survival at 6 months from the event in both the univariate (p < 0.001) and multiple logistic regression analyses (p = 0.002) [91]. However, classic β-blockers, such as metoprolol, are not easy to dose in such situations since they may lose their selectivity in the upper standard dose range or when given intravenously, while their longer duration of action may also lead to significant adverse events [92,93]. Ultra-short acting β-blockers, such as esmolol and landiolol, provide significant advantages in these circumstances since their effect can be terminated in a very short time [94,95]. Among these two i.v. agents, landiolol seems to be the most effective for decreasing heart rate in patients with acute heart failure and can be used alongside positive inotropic agents [96][97][98][99][100][101][102][103][104], e.g., in patients with left ventricular dysfunction and increased heart rate. Landiolol is also the only i.v. β-blocker with a specific dose recommendation for these patients [105]. Consequently, landiolol has been used in intensive care patients in conjunction with positive inotropic agents with positive outcomes (Figure 1) [106][107][108][109][110][111][112].

Figure 1.
Post-resuscitation care after intraoperative cardiac arrest. A 67-year-old man with a history of end-stage renal disease, heart failure, and pulmonary hypertension underwent emergency surgery (American Society of Anaesthesiologists physical status 5E) due to uncontrolled hemorrhage from an infected axillary-axillary synthetic loop graft. The patient was in hemorrhagic and septic shock upon arrival to the operating room. Two hours after induction of anesthesia, he suffered an intraoperative pulseless electrical activity cardiac arrest. The patient was resuscitated according to the PERSEUS treatment strategy (NCT04428060) [106] and received two i.v. bolus doses of epinephrine 500 μg in order to maintain a diastolic arterial pressure > 40 mmHg during CPR and hydrocortisone 200 mg. A systemic vascular resistance of 1000-1100 dynes s −1 cm −5 , end-tidal carbon dioxide of 15-18 mmHg, and central venous pressure (CVP) of 6-7 mmHg were recorded during CPR. Spontaneous circulation was restored after 2 min of CPR. Post-resuscitation cardiac ultrasound revealed left ventricular hypertrophy with severe systolic dysfunction (LVEF: ~30%, TR: 3 + /4 + ) and a systolic pulmonary artery pressure of 65-70 mmHg. Inferior vena cava diameter and CVP were 2 cm and 18 mmHg, respectively. The patient was initially supported with noradrenaline 1 μg kg −1 min −1 and dobutamine 4.17 μg kg −1 min −1 . Depth of anesthesia was adjusted to maintain bispectral index between 39 and 44 with full neuromuscular blockade. Lung-protective ventilation and targeted temperature management (35.2-35.5 °C) were applied. Esmolol infusion was started due to increased heart rate (sinus rhythm 138 beats min −1 ; noradrenaline 1 μg kg −1 min −1 , dobutamine 4.17 μg kg −1 min −1 , esmolol 14.58 μg kg −1 min −1 ). Subsequently, arginine vasopressin (AVP) was added to facilitate decatecholaminisation and mitigate the effects of noradrenaline on pulmonary vasculature (noradrenaline 0.4 μg kg −1 min −1 , AVP 0.05 IU min −1 , dobutamine 2.92 μg kg −1 min −1 , esmolol 14 μg kg −1 min −1 ). Thereafter, esmolol was replaced by landiolol in an effort to lower the ventricular rate without markedly deteriorating hemodynamics (noradrenaline 0.37 μg kg −1 min −1 , AVP 0.05 IU min −1 , dobutamine 2.7 μg kg −1 min −1 , landiolol 6 μg kg −1 min −1 ). Tranexamic acid was administrated, and the patient was transfused with a packed red blood cell/fresh frozen plasma/platelets ratio of 2:1:1 (total 6:3:3). Intraoperative time (skin-to-skin) was four hours. After another two-hour intensive care in the operating room, the patient was transferred to the intensive care unit, from which he was discharged 7 days later with a cerebral performance category score of 1. 1: fluid resuscitation and transfusion; 2: total intravenous anesthesia; 3: medical infusion pumps; 4: FloTrac/EV1000 clinical platform; 5: patient monitors providing information in numerical and waveform formats; 6: Post-resuscitation care after intraoperative cardiac arrest. A 67-year-old man with a history of end-stage renal disease, heart failure, and pulmonary hypertension underwent emergency surgery (American Society of Anaesthesiologists physical status 5E) due to uncontrolled hemorrhage from an infected axillary-axillary synthetic loop graft. The patient was in hemorrhagic and septic shock upon arrival to the operating room. Two hours after induction of anesthesia, he suffered an intraoperative pulseless electrical activity cardiac arrest. The patient was resuscitated according to the PERSEUS treatment strategy (NCT04428060) [106] and received two i.v. bolus doses of epinephrine 500 µg in order to maintain a diastolic arterial pressure > 40 mmHg during CPR and hydrocortisone 200 mg. A systemic vascular resistance of 1000-1100 dynes s −1 cm −5 , end-tidal carbon dioxide of 15-18 mmHg, and central venous pressure (CVP) of 6-7 mmHg were recorded during CPR. Spontaneous circulation was restored after 2 min of CPR. Post-resuscitation cardiac ultrasound revealed left ventricular hypertrophy with severe systolic dysfunction (LVEF:~30%, TR: 3 + /4 + ) and a systolic pulmonary artery pressure of 65-70 mmHg. Inferior vena cava diameter and CVP were 2 cm and 18 mmHg, respectively. The patient was initially supported with noradrenaline 1 µg kg −1 min −1 and dobutamine 4.17 µg kg −1 min −1 . Depth of anesthesia was adjusted to maintain bispectral index between 39 and 44 with full neuromuscular blockade. Lung-protective ventilation and targeted temperature management (35.2-35.5 • C) were applied. Esmolol infusion was started due to increased heart rate (sinus rhythm 138 beats min −1 ; noradrenaline 1 µg kg −1 min −1 , dobutamine 4.17 µg kg −1 min −1 , esmolol 14.58 µg kg −1 min −1 ). Subsequently, arginine vasopressin (AVP) was added to facilitate decatecholaminisation and mitigate the effects of noradrenaline on pulmonary vasculature (noradrenaline 0.4 µg kg −1 min −1 , AVP 0.05 IU min −1 , dobutamine 2.92 µg kg −1 min −1 , esmolol 14 µg kg −1 min −1 ). Thereafter, esmolol was replaced by landiolol in an effort to lower the ventricular rate without markedly deteriorating hemodynamics (noradrenaline 0.37 µg kg −1 min −1 , AVP 0.05 IU min −1 , dobutamine 2.7 µg kg −1 min −1 , landiolol 6 µg kg −1 min −1 ).
Tranexamic acid was administrated, and the patient was transfused with a packed red blood cell/fresh frozen plasma/platelets ratio of 2:1:1 (total 6:3:3). Intraoperative time (skin-to-skin) was four hours. After another two-hour intensive care in the operating room, the patient was transferred to the intensive care unit, from which he was discharged 7 days later with a cerebral performance category score of 1. 1: fluid resuscitation and transfusion; 2: total intravenous anesthesia; 3: medical infusion pumps; 4: FloTrac/EV1000 clinical platform; 5: patient monitors providing information in numerical and waveform formats; 6: mechanical ventilator parameters and waveforms; 7: temperature control; 8: bispectral index (BIS) monitoring; 9: regional cerebral oxygen saturation (rSO2); 10: internal jugular vein cannulation; 11: manual external defibrillator/pacemaker. Although dobutamine is the first choice for short-term intravenous inotropic support in patients with decreased contractility, the mechanism of action of levosimendan makes it an attractive alternative [113,114]. Levosimendan increases the sensitivity of myocytes to calcium and improves contractility without increasing intracellular calcium levels; the latter is a key pathophysiological mechanism of post-resuscitation myocardial stunning and ischemic contracture, and attenuating this phenomenon seems important [5,115]. Whether post-resuscitation stable dysrhythmias must be treated immediately after their diagnosis remains unknown. However, they are commonly caused by focal cardiac ischemia, and patients with a new-onset dysrhythmia must be evaluated for percutaneous coronary intervention.
In critically ill patients, mean arterial pressure (MAP) represents the entry pressure for the perfusion of most organs and should be maintained >65-70 mmHg [116][117][118]. Although higher MAP levels may be required in patients with brain injury or persistent hypoperfusion (e.g., progressing acute kidney injury or altered mental status) [119,120], adequate circulatory volume, absence of left ventricular outflow tract obstruction, and microcirculatory flow and responsiveness (if possible) should be ideally assessed before using a vasopressor challenge, especially in patients treated with TTM [77,[121][122][123]. Considering that organ perfusion pressure is influenced by MAP and venous pressure, maintaining an optimal central venous pressure may facilitate adequate oxygen delivery. Additionally, it is important to remember that diastolic arterial pressure is key for coronary perfusion pressure, and its evaluation is also crucial. Patients with significant or unstable coronary artery disease and those with chronic pulmonary hypertension at a risk of low coronary perfusion pressure may require higher diastolic pressures [124].
Vasodilation should be actively treated with vasopressors, initially targeting a MAP of >65-70 mmHg, followed by an individualized approach. Norepinephrine is recommended as a first-choice agent to increase stressed volume and systemic vascular resistance whilst decreasing inflammation-induced capillary permeability [125,126]. Vasopressin can be used as an adjunct to limit the side effects of catecholamines or when agents with a different mechanism of action are needed; however, vasopressin may be the preferred option in patients with pulmonary hypertension, right ventricular failure, and/or vasopressin deficiency [127,128]. Moreover, vasopressin decreases the risk of atrial fibrillation and may improve renal function in patients with vasodilatory shock [129]. Few data are available regarding angiotensin II, limiting its use as a third-or fourth-choice agent in patients with angiotensin II deficiency or altered expression of angiotensin receptors [130,131]. Of note, hypoxic hypercapnia significantly affects the intra-and extrasplanchnic vascular capacitance system, and the dose of exogenous vasopressors should be possibly decreased in these patients to maintain adequate venous return and afterload [132][133][134]. Whether TTM affects hemodynamics remains a topic of discussion as these patients often have diverse requirements in vasopressor support [87,135].
In the context of the vasopressin-steroids-epinephrine (VSE) protocol, early post-ROSC, stress dose steroids may contribute to hemodynamic stabilization, especially in patients requiring high doses of vasopressors (e.g., norepinephrine equivalent ≥ 0.25 µg kg −1 min −1 ) and who have multiple organ failure [136,137]. In addition, a re-analysis of combined data from the two randomized VSE trials reported that exposure to stress dose steroids was associated with a lower risk of post-resuscitation lethal septic shock [138]. Nevertheless, a more recent, two-center, randomized trial of stress dose steroids (alone) vs. placebo did not confirm any steroid-associated physiological benefit [139].
Optimizing preload during the post-resuscitation period may be difficult, and high doses of balanced crystalloids are often needed. However, avoiding congestion and the injurious effects of fluid over-resuscitation is imperative. An effective fluid resuscitation strategy may necessitate the adoption of a complex, multimodal cardiovascular model capable of primarily integrating both the arterial and venous sides of the circulation, including microcirculatory flow and oxygen extraction [140].
Perhaps the most important to recognize is the patient who is fluid responsive but not fluid tolerant because this patient will be harmed by a fluid responsiveness-based strategy [138]. As multiple factors can impact the ability of different organs and compartments to accommodate fluids and maintain their function, and different patient phenotypes exist, frequent multimodal and comprehensive clinical assessments of fluid responsiveness and tolerance are necessary. This assessment may include medical history and physical examination, radiographic evaluation, advanced hemodynamic monitoring, intraabdominal pressure measurement, point of care ultrasound (POCUS), and assessment of abnormalities in splanchnic venous flow patterns (i.e., venous excess ultrasound score-VexUS) [141,142].
In patients with refractory circulatory failure, treatment with assistive devices, such as Impella, intra-aortic balloon pump, or veno-arterial extracorporeal membrane oxygenator pumps, may be indicated. However, these devices are often associated with compilations and should therefore be used in selected individuals [143][144][145][146].

Antibiotic Therapy
Over one-third of adults with OHCA may be bacteremic upon presentation to the Emergency Department [147]. In addition, patients undergoing TTM may develop insulin resistance [148,149], which may impair tissue perfusion and increase the risk of infection. More specifically, clinical and experimental studies suggest that hyperglycemia induces excessive vasoconstriction, endothelial dysfunction, oxidative stress, and inflammatory response, which contribute to microcirculatory dysfunction [77,[150][151][152][153]. However, bacteremia and antibiotic administration during resuscitation has not been associated with key outcomes [154]. Therefore, routine prophylactic antibiotics are not recommended despite the possibly increased risk for the development of pneumonia and other infections after cardiac arrest and should be reserved for those with evidence of infection. If antibiotics are administered, significant efforts must be made to improve tissue perfusion and local flow, and thus antibiotic delivery to the potential source of infection.

Active Temperature Control
In most patients, the primary neurologic injury occurs during cardiac arrest and may continue after ROSC [13]. However, the complex pathophysiology, diverse population, and lack of standardized protocols are major limitations in optimizing neuroprotection. Consequently, post-resuscitation neurological management requires a coordinated multidisciplinary approach aiming at attenuating the progression of cerebral injury.
Targeted temperature management has been described as the most effective neuroprotective strategy, and current recommendations suggest that it improves neurologic outcomes [69,155,156]. However, the recently published "Hypothermia versus Normothermia after Out-of-Hospital Cardiac Arrest" trial, the largest trial to date, found no difference in survival at 6 months or in health-related quality of life between TTM at 33 • C followed by controlled rewarming or targeted normothermia at 36 • C with early treatment of fever (body temperature > 37.7 • C) [157]. Similar results were reported in the CAPITAL CHILL trial, in which patients were randomly assigned to TTM of 31 • C or 34 • C for a period of 24 h [158]. Nevertheless, that study was underpowered to detect a clinically important difference of ≤3%.
Although the international guidelines recommend a target core body temperature of 32 to 36 • C and avoiding fever for at least 72 h [69], considerable debate exists on the optimal timing and temperature target, including whether just avoiding fever is enough or whether TTM is also effective for non-shockable rhythms. A target temperature on the higher end of the aforementioned range may be appropriate for patients with mild brain injury, higher bleeding risk, trauma, recent surgery, or septic shock. On the other hand, patients who may benefit from a temperature target of 32-33 • C include those with severe brain injury, subarachnoid hemorrhage, or stroke [159][160][161][162][163][164][165][166]. In addition, the HYPERION trial reported that moderate therapeutic hypothermia at 33 • C for 24 h led to a higher percentage of patients who survived with a favorable neurologic outcome at day 90 compared to targeted normothermia at 37 • C [167]. In a recent randomized trial of nearly 800 patients who received TTM targeting 36 • C for 24 h after resuscitation from cardiac arrest, the composite outcome of death from any cause or hospital discharge at 90 days with either severe neurologic disability or coma was similar for patients who subsequently underwent fever prevention for an additional 12 h (36 h total TTM duration) versus 48 h (72 h total TTM duration) [168] (Table 2).   Pulmonary edema: n = 7 (intra-arrest), n = 8 (hospital) (p = 0.59); pneumonia: n = 7 (intra-arrest), n = 3 (hospital) (p = 0.24); hyperthermia: n = 9 (intra-arrest), n = 5 (hospital) (p = 0.36); bacteremia: n = 1 (intra-arrest), n = 0 (hospital) (p = 1); hemorrhage: n = 3 (intra-arrest), n = 3 (hospital) (p = 0.88); arrhythmia: n = 5 (intra-arrest), n = 7 (hospital) (p = 0.39); convulsion: n = 8 (intra-arrest), n = 2     Based on current evidence and recommendations, the target temperature should be maintained stable for at least 24 h, avoiding variations and shivering, while fever should be avoided for at least 72 h after cardiac arrest [178,184,187,188]. Additionally, patients should be rewarmed at a slow rate (i.e., <0.5 • C h −1 ) [189,190].
Interestingly, several recent systematic reviews and meta-analyses of randomized trials do not support the use of TTM. Elbadawi et al. analyzed eight randomized studies with a total of 2927 patients and a weighted follow-up period of 4.9 months and reported that TTM was not associated with improved survival or neurological outcomes compared with normothermia in comatose patients after cardiac arrest [191]. Another systematic review and network meta-analysis of temperature targets found that mild, moderate, or deep hypothermia may not improve survival or functional outcome after OHCA and may be associated with more harm than benefit [192]. Granfeldt et al. assessed all aspects of TTM, including timing, temperature, duration, method of induction and maintenance, and rewarming in 32 trials and reported that the use of TTM at 32-34 • C, when compared to normothermia, did not result in improved outcomes [193]. In another recent systematic review and Bayesian meta-analysis of seven adult cardiac arrest trials, TTM at 32-34 • C for ≥12 h versus normothermia with active control of fever had a chance of ≤53% to ≤78% to reduce the risk of death or unfavorable neurological outcome by 2-4% [194]. Consequently, more high-quality, large, randomized studies are warranted to further clarify the value of targeted hypothermia versus targeted normothermia.

Prognostication
The overall prognosis of patients following cardiac arrest remains poor, with only half of them surviving to discharge [173,195]. Early prognostication can be difficult, and clinical examination should be initially performed after ROSC and thereafter on a daily basis to assess the neurological status and guide decision making. However, assessments may be confounded by physiological derangements such as hypoxia, hypothermia, circulatory failure, and metabolic acidosis. Most in-hospital deaths in comatose patients are caused by hypoxic-ischemic brain injury [69]. Therefore, the overall prognosis depends on the no-flow time, the quality and duration of CPR, and the quality of post-resuscitation care. Furthermore, the combination of patient characteristics, e.g., age and frailty, components of medical care, anesthesia, TTM, and organ injury mandate that prognosis be determined in most patients only after the first five to seven days after ICU admission [69,[196][197][198]. Notably, late awakening may be due to ongoing cardiovascular instability or multiple organ failure and does not preclude full neurological recovery [199][200][201].
As accurate prognostication is essential, a multimodal approach should be used in all comatose patients. Brain-computed tomography, measurements of biomarkers such as protein S100B or neuron-specific enolase, evoked potentials, electroencephalogram, and frequent clinical examination are important tools [202]. However, several factors can limit prognostication; for example, TTM may affect the predictive value of computed tomography, while continuous electroencephalogram may have a limited predictive value for a good outcome [203][204][205][206]. Additionally, no clear cut-off has been identified for neuronspecific enolase, and serial sampling at 24, 48, and 72 h after ROSC are necessary to assess trends.
In general, the neurological outcome depends on the prompt restoration of the systemic circulation and adequate oxygen delivery to meet cerebral oxygen demands [207,208]. Until recently, it was assumed that under normal circumstances, autoregulation maintains a constant cerebral blood flow, and changes in mean blood pressure within a range of 50-150 mmHg have a minor influence on cerebral blood flow [209,210]. However, recent evidence suggests that autoregulation maintains cerebral blood flow within a smaller range above baseline MAP [16,211]. After cardiac arrest, the evidence is conflicting, with several studies showing that cerebral autoregulation is preserved after cardiac arrest [212,213], while other studies reported the absence of autoregulation [214].
Monitoring cerebral perfusion with transcranial Doppler sonography (TCD) may enhance clinicians' ability to optimize individual cerebral perfusion, minimize secondary brain damage, and improve prognostication among patients admitted to the ICU after cardiac arrest. This technology allows the measurement of key parameters, e.g., cerebral blood flow velocity and pulsatility index, that allow ongoing, real-time assessments of patients' autoregulatory indices, intracranial pressure, compliance, and cerebral blood flow, and can identify potentially treatable derangements [215,216]. Various studies have provided conflicting results concerning the association between initial TCD values and neurological outcomes [208,212,[217][218][219][220][221].
Although the interpretation of an elevated pulsatility index is complex, values > 1.19 are typically associated with increased downstream cerebrovascular resistance [222]. However, the pulsatility index may increase in the context of decreasing cerebrovascular resistance [217,223]. Transcranial Doppler sonography parameters may complement other available neuromonitoring tools, such as intracranial pressure monitors and near-infrared spectroscopy [224,225]. Of note, TCDs are non-invasive, and their validity may be superior to near-infrared spectroscopy [226]. Real-time data interpretation requires substantial bioinformatic infrastructure and clinician expertise. The GOODYEAR trial (NCT04000334) is anticipated to shed more light on the feasibility of an early goal-directed hemodynamic management with TCD during the first 12 h after ROSC.

Ethics of Critical Care and End-of-Life Decisions following Cardiac Arrest
Maximizing the benefit of critical care for patients and their families is a key aspect of post-cardiac arrest management. Apart from high-quality organ support, preventing pertinent harm and early end-of-life care decisions are tightly related to the application of patient-centered care. In this context, discussions with the patient or family members following cardiac arrest may affect the quality of care and should rely on a shared decisionmaking process [227]. The latter can support and optimize the appropriate allocation of resources, decrease ICU/hospital length of stay, aid in the selection of palliative care pathways, and reduce health care costs.
Effective communication is very important for patient relatives, who may be severely impacted by the illness and critical care stay of their loved ones, experiencing various phycological disorders, such as anxiety, acute stress, post-traumatic stress disorder, and depression [228,229]. Communication in the context of shared decision-making is associated with higher patient/family satisfaction and increased decisional confidence [227,230,231]. However, communication with families is not always easy, and structured communication tools may improve shared decision-making and patient/family satisfaction [232,233]. Consequently, family support interventions that can help reduce these psychological impacts and family-centered communication and care should be key objectives of post-resuscitation management [230][231][232][234][235][236]. Indeed, post-ICU admission-focused discussions with relatives can increase documentation of patient preferences and facilitate advance care planning and end-of-life care [232,[237][238][239].
Of note, advance care planning is associated with improvements in symptom control and quality of life, decreases in family carers' decisional conflict, improvements in ICU care and post-resuscitation suffering, lower caregiver burden, and higher patient/family satisfaction [240]. However, an important limitation in the implementation and research in advanced care planning is the lack of a standardized approach; this is also a main cause for the inconsistent findings between studies [238,[241][242][243][244][245]. In addition, advanced care planning and shared decision-making may increase organ donation pathways and rates after ensuring family members that donation will be considered only when ongoing treatment cannot improve outcome.
On the other hand, specific and adequate training of healthcare professionals is imperative for improving critical care and end-of-life decisions following cardiac arrest [246]. However, this type of training is often inadequate during medical training or time of specialization/subspecialization. Easily accessible relevant training programs or workshops to improve the delivery of end-of-life care must be available amongst all hospital staff. For example, the 'End-of-life Care for All (e-ELCA)' program is an e-learning library that provides resources to enhance the training and education of the health and social care workforce. The e-ELCA has been highlighted as a resource to help with the implementation of the NICE Guidelines on improving care for people who are in their last days of life [247]. Similarly focused programs created for cardiac arrest patients and their families may enhance the quality of post-resuscitation care.

Conclusions
A substantial proportion of cardiac arrest deaths can be attributed to the development of post-cardiac arrest syndrome, and post-resuscitation care is the fourth link in the chain of survival. Critical care management requires highly specialized resources and should be based on a multidisciplinary approach ensuring best-practice critical care.

Conflicts of Interest:
The authors declare no conflict of interest.