Cardiogenic shock (CS) is estimated to complicate approximately 5–10% of all admissions for acute myocardial infarction or decompensated heart failure, with in-hospital mortality ranging from 30–40%.1–6 Given advances in early revascularization, 7 temporary mechanical circulatory support (tMCS), 8 and multidisciplinary shock teams, 9–11 the majority of patients with CS can now survive to hospital discharge. 5,12–15 After discharge, these survivors face the long-term sequelae of CS and the challenges of recovery. This review summarizes the current state of knowledge of survivorship after CS, identifies lessons learned from recovery following other types of critical illness that are relevant to CS survivors, and explores opportunities to improve long-term outcomes among this vulnerable patient population.Clinically, CS is characterized by sustained systemic hypoperfusion with compromised end-organ function due to inadequate cardiac output. 16 CS exists across a continuum of severity as defined by the Society of Coronary Angiography and Intervention (SCAI) shock stages. 17 Multiple etiological subtypes of CS exist, including CS due to an acute myocardial infarction (AMI-CS), CS due to decompensated heart failure (HF-CS), and CS following cardiac surgery (post-cardiotomy CS). 16 Despite HF-CS being the most common form of CS in the modern era, 13,14 most randomized trials of CS management have focused on AMI-CS, 7,8,18 Patients with these distinct phenotypes of CS have important differences in their underlying pathophysiology, pre-admission health status, and potential for myocardial recovery, resulting in differences in their in-hospital outcomes. 6,13While CS across phenotypes is a highly morbid condition, contemporary data consistently demonstrate that the majority of patients with CS survive to hospital discharge, with survival rates of approximately 60–70% across multiple cohorts. 5,12–15 Recent evidence also shows that survival in CS has improved while the prevalence of CS has increased over the past two decades. 5,15 Unfortunately, a large proportion of patients with CS develop multi-organ failure during admission, 13,19 while many others experience complications of their critical illness such as delirium or vascular complications. 14,20,21 The use of invasive critical care interventions, particularly tMCS (including veno-arterial extracorporeal membrane oxygenation [VA-ECMO], micro-axial percutaneous left ventricular assist devices, and intra-aortic balloon pumps [IABP]), as well as invasive mechanical ventilation and renal replacement therapy, are common among patients with CS. 5,12–15 Consequently, while most patients with CS will survive to hospital discharge, survivors will continue to face the sequelae of CS long after their index hospitalization.
In other forms of critical illness, many survivors have severe and persistent impairments in multiple recovery domains, collectively referred to as Post-Intensive Care Syndrome, or PICS. 22–25 This multi-faceted syndrome includes cognitive impairments, limitations in physical function, and challenges involving mental health. 25 Together, these factors can lead to diminished quality of life. 26 While these features are the core aspects of PICS, it is important to remember that PICS can include multiple other changes in a patient’s medical and physical conditions, such as the development of new comorbidities or worsening of pre-existing ones, in addition to physical sequelae of procedures such as scarring or vocal cord dysfunction. 27 Furthermore, caregivers and family members of intensive care unit (ICU) survivors have also been found to face the risk of burnout and health challenges themselves, with high rates of depression, anxiety, and post-traumatic stress among caregivers. 27,28
Importantly, PICS may have unique manifestations following different forms of critical illness. 23,29 Survivors of CS represent a distinct population with key differences from other groups of ICU survivors in terms of their underlying pathophysiology of shock, pre-morbid health status, and advanced cardiac care treatments received while hospitalized (Figure 1). 5,12,24 The use of tMCS in particular sets patients with CS apart as a unique population of survivors of critical illness. In addition, the persistent nature of cardiac illness, including both the possibility of recurrent cardiac events as well as ongoing fear surrounding recurrence, has the potential to shape CS survivorship much differently than among critically ill patients who survive curable conditions such as sepsis. 30 Thus, while much can be learned from lessons of PICS due to non-cardiac illness, post-ICU survivorship specifically among patients with CS is a distinct entity and thus these patients must be studied as a separate group.
Cardiogenic shock and its associated therapies likely have significant effects on the brain, ranging from augmented yet non-pulsatile perfusion with some tMCS devices to changes in microvascular tone with vasoactive medications to hemolysis and inflammation. Understanding these acute changes in the brain and its vasculature during CS may identify opportunities to improve short- and long-term cognitive outcomes. Chronic heart failure is associated with cognitive impairment, particularly in older adults, and may be mediated by mechanisms including dysregulated cerebral blood flow, chronic hypoperfusion, and inflammatory changes. 31 Cognitive impairment has important clinical implications on self-care, medication management, executive and physical function, and independence. There is also a strong association between cognitive impairment, physical frailty, and other forms of cardiovascular disease (CVD) beyond heart failure. 32 Importantly, cognitive status in chronic CVD may be able to be improved through CVD therapies, as illustrated by research showing a subset of patients have improved cognitive function following transcatheter aortic valve replacement. 33 Durable LVAD implantation has also been demonstrated to improve long-term cerebral hemodynamics, 34 which could in turn improve cognitive function.
Cognitive impairment is common following critical care illnesses in general. The BRAIN-ICU study is one of the largest investigations in this field, enrolling 467 survivors from the medical or surgical ICU who then underwent detailed neuropsychological testing at 3 months and 12 months after discharge. 35 For the entire cohort, median cognitive testing scores were approximately 1.5 standard deviations below age-adjusted population norms, scores comparable to those seen in patients with mild dementia. These cognitive impairments were consistent across age groups; remarkably, even among patients <50 years old, approximately one-fifth of this youngest group of patients had cognitive status comparable to patients with mild Alzheimer’s dementia. In contrast to Alzheimer’s disease, where memory is the primary cognitive domain affected, the BRAIN-ICU cohort of patients had more global dysfunction encompassing multiple cognitive sub-domains.
Delirium is one of the strongest predictors of long-term cognitive impairment following critical illness, with the risk of residual cognitive impairment increasing with a longer duration of delirium. 35 Delirium can occur in either hypoactive or hyperactive forms; while hyperactive delirium is the more easily recognized classic subtype, hypoactive delirium occurs more frequently in critically ill patients and has a stronger association with long-term cognitive impairment. 36 Limited data exists regarding delirium in patients with CS, with one single-center study demonstrating a 24% rate of delirium among patients receiving tMCS. 21 Benzodiazepine use has been shown in randomized trials to significantly increase the incidence of delirium, 37 although a direct link between benzodiazepine use and long-term cognitive dysfunction is not well-established. 35 Other risk factors which have been associated with long-term cognitive impairment after critical illness include older age, higher comorbidity burden (including pre-existing cognitive dysfunction), and longer duration of invasive mechanical ventilation. 38 Studies evaluating cognitive rehabilitation to improve brain function among ICU survivors who develop cognitive dysfunction have been extremely small with inconsistent results across these studies. 38,39
A large proportion of ICU survivors will experience long-term physical limitations which can result in persistent dependence on others for daily activities. 23,24,40 Age and greater burden of pre-existing comorbidities at the time of ICU admission are both associated with worse long-term functional status. 23,40 ICU survivors can develop post-ICU muscular weakness, which can be due to myopathy, polyneuropathy, or a combination; demonstrated risk factors for ICU myopathy/neuropathy include female sex, immobility, multi-organ dysfunction, and duration of invasive mechanical ventilation. 41, 42 In the post-ICU setting, the diagnosis of myopathy/neuropathy is generally made clinically, although there may be a role for nerve conduction studies, electromyography, or muscle biopsy for patients in the ICU with persistent muscular weakness which limits progression of their care. 41 The physiologic changes that result in muscular weakness occur within days of ICU admission and involve both changes in muscle composition as well as in muscle mass. 43,44 This suggests that interventions that occur in the ICU to minimize the development of ICU-acquired weakness may have the greatest chance of improving long-term functional status among survivors of critical illness. Multiple studies have tested early mobilization in the ICU with promising results in some but not all studies. 45–47 These protocols generally consisted of a stepwise progression in mobility to the maximum extent of a patient’s abilities on a given day, ranging from passive range of motion performed by therapists for unresponsive patients to exercises for awake patients as tolerated either in bed, seated, standing, or walking. This remains an area in need of further study, particularly in patients with CS where different therapeutic strategies (such as axillary versus femoral tMCS, or rapid versus prolonged weaning of tMCS) can greatly affect patients’ ability to mobilize in the ICU and thus potentially modify the risk of ICU-acquired weakness.
Physical limitations after critical illness are likely multifactorial, encompassing muscular weakness, cardiopulmonary dysfunction, and deconditioning. However, it may be that muscle dysfunction is the primary driver of impaired physical function in PICS in general. On non-invasive cardiopulmonary exercise testing among a cohort of ICU survivors on a five-year follow-up, 38% had impaired exercise capacity with muscular limitation as the most common contributing factor. 48 Baseline and long-term cardiac function were not described in this cohort, although 40% had undergone cardiac surgery. Herridge et al evaluated a population of acute respiratory distress syndrome survivors five years after discharge and found that overall patients had normal to near-normal pulmonary function but continued to have significantly reduced six-minute walk distance. 22 In patients who survive cardiogenic shock, presumably cardiac function plays a greater role in determining long-term functional capacity, although this has not been characterized. Procedural sequelae can contribute to physical dysfunction as well, 27 an important consideration for patients who have received tMCS.
Physical frailty is also common among ICU survivors and is most often newly acquired over the course of critical illness. 49 While various definitions of frailty exist, the concept refers to an age-associated syndrome of diminished physiologic reserve across multiple health domains which increases vulnerability to adverse health outcomes. 50 Frailty often co-exists with other sequelae of critical illness including disability and cognitive impairment. 49 Frailty and sarcopenia have been strongly associated with worse outcomes in multiple forms of CVD, including AMI-CS. 51,52 Importantly, multiple small studies have demonstrated the potential of interventions such as cardiac rehabilitation to modify markers of frailty and improve outcomes. 50 A subgroup analysis of patients with ejection fraction ≤40% from the TARGET-EFT study of a multi-component intervention targeting physical weakness, cognitive impairment, malnutrition, and anemia among adults ≥65 years with signs of frailty suggested the benefit of these interventions may be attenuated with reduced EF. 53 Likewise, there have been multiple relatively small studies evaluating post-ICU interventions such as physical rehabilitation or nutritional supplementation to improve physical function in debilitated ICU survivors in general with mixed results. 27 The type and timing of interventions to modify frailty and physical function, as well as optimal patient selection for these interventions, remains an area of significant ongoing research which will need to include CS survivors specifically.
Mental health outcomes following critical illness in general are complex and likely heavily reflective of both a patient’s pre-morbid condition as well as the etiology of their critical illness. Jackson et al in an analysis of medical and surgical ICU survivors in the BRAIN-ICU study found that one-third of patients reported depressive symptoms at 1-year follow-up, with only 7% reporting post-traumatic stress disorder (PTSD). 24 The rate of depression was higher among patients with a history of depression preceding their critical illness. Depression in the BRAIN-ICU cohort was primarily related to self-perceived physical limitations and somatic complaints rather than affective (mood) symptoms, suggesting that efforts at improving physical function (such as through rehabilitation) may be more effective at improving mood symptoms than psychotropic medications in PICS in general. 24
Mental health conditions are common in patients with CVD and are associated with an increased risk of adverse outcomes. 54 There is increasing evidence establishing mechanistic links between psychological health and CVD outcomes both indirectly through health behavioral changes as well as directly through biological pathways such as autonomic dysfunction and inflammation. 54 Three of the most common mental health conditions include depression, anxiety, and PTSD. Patients with CVD are estimated to have rates of depression two to three times higher than the general population, 55 and depression is an independent risk factor for mortality following acute coronary syndrome or decompensated heart failure, even after adjusting for risk factors including LV dysfunction and age. 56,57 Sex-based differences in prevalence of depression have been reported across many CVD conditions. 58,59
Patients with CVD may experience ongoing fear of recurrent cardiac events which predisposes to PTSD symptoms. 60 These stress symptoms have been observed to affect patients across multiple forms of cardiac disease including chronic heart failure, 61 heart transplant, 62 and after myocardial infarction. 63 Notably, rates of stress-associated mental health disorders vary significantly based on the specific form of CVD and may be particularly accentuated in patients who survive cardiac arrest. Agarwal et al found that over one-third of arrest survivors developed PTSD, with a three-fold higher risk of adverse cardiac events and mortality within one-year after hospital discharge among these patients. 30 Fernando et al recently published an analysis demonstrating that new mental health diagnoses are common following AMI-CS, particularly among women and among patients in a rehabilitation facility or long-term hospital. 64
Mental health represents a domain where CS survivors may differ significantly from other populations of ICU survivors. One example of the differing manifestations of PICS across forms of critical illness is the significantly higher rate of PTSD in post-arrest patients (36%) relative to the BRAIN-ICU cohort (7%).24,30 In contrast to curable and potentially isolated etiologies such as infection that lead to critical illness in a large proportion of medical and surgical ICU patients, CS survivors face the persistent risk of future events due to chronic heart failure or coronary artery disease, as well as ever-present reminders of their illness such as cardiac implantable electronic devices (CIEDs) and daily cardiac medications. As a result, mental health, in particular cardiac-specific anxiety, 65 may play a larger role in CS survivorship than in other forms of PICS. Given the high prevalence of mental health problems in patients with CVD overall as well as in ICU survivors in general, all CS survivors should be considered for brief screening for mental health conditions. There is some data that antidepressant medical therapy can improve cardiac outcomes in patients with established CVD, while psychotherapy has been shown to improve depressive symptoms but not cardiac outcomes. 54 The BRAIN-ICU study would also suggest that physical rehabilitation may benefit ICU survivors to a greater extent than interventions directly targeted at mental health. 24 Ultimately further research is necessary to first establish the incidence and types of mental health problems most common among CS survivors and then test interventions to improve outcomes specifically among patients who survive CS.
While self-reported quality of life (QoL) may be the most important outcome for patients who survive a critical illness, it may also be the most challenging endpoint to modify directly given the many factors that influence QoL. Patient-reported QoL varies widely across ICU survivorship studies, 66 and is heavily impacted by patients’ pre-morbid status and age, as well as other factors such as sociodemographic and family background. 67 Quality of life is also a multifaceted construct that is measured in different ways across studies. In addition, such metrics can be dynamic over time, are influenced by various personal and situational factors, and may not correlate well with objective measures of functional status. 68 In a study of cardiac arrest survivors, for example, depressive symptoms were strongly correlated with patients’ self-perception of their recovery status, while there was no association between objective cognitive function and recovery self-perception. 69 Likewise in ICU survivors, physical performance measures have been found to show stronger correlations with patient-reported outcomes centered on specific tasks but limited correlation with QoL. 70 Cuthbertson et al examined QoL over 5 years in a cohort of ICU survivors; they observed an initial significant decline in patient-reported physical QoL at 3 months follow-up, and an increase over the next several years (although still below population norms), followed by a decline at 5 years of follow up. 26 This study also found that QoL was lower in some aspects among younger patients, particularly in terms of mental health-related scores. QoL is affected by the complex interaction between the cognitive, physical, and mental health domains which are not easily untangled. Patients report cognitive impairment to be their greatest fear when asked questions regarding hypothetical scenarios, 71 while despite the high prevalence of mental health conditions among ICU survivors, physical limitations appear to be self-reported more commonly than emotional problems on QoL assessments. 24,26 Overall, QoL can be influenced by a variety of factors and ultimately may reflect a patient’s self-perception of their illness trajectory.
The concept of illness trajectories has been used to characterize the health course of patients with chronic heart failure, 72 with chronic progressive illness in general, 73 and following critical illness. 74 In general, illness trajectories describe a patient’s functional/health status across three time periods: pre-admission, acute hospitalization, and post-discharge. While in the course of clinical care we often implicitly consider a patient’s illness trajectory, explicitly reflecting on these trajectories may allow us to more fully work with our patients to help them achieve the best possible outcome in alignment with their goals of care.
There is currently relatively limited data on long-term trajectories among survivors of CS, particularly regarding functional status and QoL. Sterling et al studied 9,789 patients with AMI-CS in Ontario (excluding patients with out-of-hospital cardiac arrest), of whom 69.8% survived to hospital discharge. 5 Of these survivors, 86% had been living independently prior to admission with CS, and only 49% were able to return home without assistance at the time of discharge. Sinha et al studied a population of 520 patients at Inova with HF-CS or AMI-CS, with 69.6% of patients surviving to discharge. 12 Less than two-thirds of CS survivors in this cohort were able to be discharged to home. Across both of these studies, the rates of 30-day readmission (19.2% and 22.7%, respectively) and one-year survival (84.7% and 77.9%) were similar among those who survived to hospital discharge. The majority of 30-day readmissions in the Inova cohort were for a non-cardiac etiology. 12 Shah et al observed a similar predominance of non-cardiac hospitalizations in an analysis of long-term outcomes of Medicare patients ≥65yo with AMI-CS. 75
Heart replacement therapies (heart transplant or durable left ventricular assist device [LVAD]) represent an important path to survival for patients with CS who fail to experience native heart recovery. In the Inova cohort, 8.8% of patients with CS received heart replacement therapies while admitted. 12 However, patients with CS who undergo durable LVAD implantation from tMCS (Interagency Registry for Mechanical Assisted Circulatory Support [INTERMACS] profile 1) have worse short- and long-term outcomes than patients without CS undergoing LVAD placement. 76 Referral for advanced therapies following index hospitalization for CS represents an important unmet need given the risk of poor outcomes among CS survivors.
It is highly likely that patients with acute-on-chronic HF-CS have on average very different illness trajectories than those with AMI-CS or de-novo HF-CS. Sinha et al did not find a statistically significant difference in 1-year all-cause mortality among CS survivors with HF-CS versus those with AMI-CS (HR 0.73, 95% CI 0.45–1.18, p=0.20), but the difference in post-discharge survival curves suggests that survivors of HF-CS are a vulnerable cohort. 12 Pre-admission functional status is strongly related to post-ICU functional trajectory among ICU survivors in general, 77 and the typical differences in baseline health status between patients with AMI-CS compared to those with HF-CS likely are also reflected in their post-discharge course. Additional research characterizing functional outcomes among CS survivors will be necessary to verify typical trajectories among the different subgroups of CS.
When considering illness trajectories, it is important also to consider where the inflection points are which allow for that trajectory to be modified. Some patients may present with CS that is clearly not survivable, or at least not survivable with a QoL that would be acceptable to that particular patient (such as where the best-case scenario would entail lifelong dialysis in a patient who finds that to be unacceptable). In these patients, a comfort-focused approach would be best after the goals of care have been established. Interventions in the ICU can determine both a patient’s survival as well as influence their functional ability for years to come. Inflection points may (or may not) also exist at other time points in a patient’s course following the ICU. In PICS in general, post-ICU rehabilitation appears to be more effective in patients with a lower burden of pre-admission comorbidities. 78 Cardiac rehabilitation has been shown to improve outcomes in patients following an AMI and potentially modify frailty. 50 Future research will be necessary to identify inflection points that allow for improved survival, functional status, and QoL in CS.
The concept of illness trajectories has important implications for the long-term management of patients with CS. Worse functional status prior to admission likely augurs a lower probability of functional recovery after CS, although this has not yet been fully described in the literature. Resource limitations, particularly for heart replacement therapies, make identifying patients’ probability of recovery an important component of decision-making in CS. These complex decisions must often be made quickly with limited information at hand. Shock teams have been shown to be associated with improved outcomes in CS; 9–11 little research has yet been done into how shock teams make decisions, but one potential mechanism of benefit through a multidisciplinary team-based approach is that teams may be able to better identify a patient’s trajectory than individual physicians alone.
Throughout this process, understanding a patient’s goals of care and sharing decision-making with the patient is essential. A quality of life that one patient may find completely unacceptable could be fully acceptable to a different patient. In turn, physicians can only relate the estimated probabilities of particular outcomes, and interventions that may offer the greatest chance of survival (such as VA ECMO) may also have the highest risk of complications and subsequent survival with significantly diminished quality of life. Complicating these time-sensitive discussions is the fact that patients with CS may not be able to speak for themselves. Incorporating palliative care, both primary palliative care delivered by the Cardiac ICU team as well as secondary palliative care provided by dedicated specialists, is a key aspect of patient-centered care. 79 It is important to note that it is incumbent upon members of the Cardiac ICU team to have these conversations with the patient themselves in concert with palliative care specialists, rather than viewing goals of care discussions as merely something to be outsourced to consultants.
The importance of survivorship following cardiogenic shock is beginning to be recognized. The Shock Academic Research Consortium (SHARC), an international multi-stakeholder collaboration which includes clinicians, investigators, and representatives from the Food and Drug Administration and industry, recently published consensus definitions on clinical trials in CS. 16 The members of this panel emphasized the need to include patient-centered outcomes in CS trials. The SHARC panel recommended that outcomes across at least four domains be assessed in future CS trials: health-related QoL, activities of daily living, disability, and days alive out of the hospital. Two recent American Heart Association (AHA) Scientific Statements focused on CS in older adults and the care of older adults in the Cardiac ICU have also discussed the importance of considering long-term outcomes holistically. 80,81 Several forthcoming studies will shed light on CS survivorship. The HALO-Shock trial (NCT #04419480) is enrolling 40 CS survivors and randomizing them to implantable pulmonary artery pressure monitoring with a hierarchical primary endpoint assessed at 6 months which includes heart failure-related QoL. The Cardiogenic Shock Working Group is also assessing heart failure-related QoL after discharge. The VANQUISH Shock Collaborative is measuring QoL and functional disability at 6 and 12 months after discharge. 82 Our own research group is conducting the Comprehensive Outcomes After Cardiogenic Shock (COACS) study evaluating cognitive ability, functional status, and QoL among CS survivors.
The field of CS survivorship is still in its infancy, with significant work remaining to 1) characterize patient-centered outcomes among CS survivors, 2) develop and test interventions (both during admission and after discharge) to improve functional outcomes, and 3) translate these lessons into routine clinical use, including in cardiology clinics. One of the challenges in both clinical care and research involving ICU survivors is the lack of standardized outcome measures. Assessment tools that are highly useful from a research perspective may be impractical and too time-consuming for routine clinical care, while other metrics may require specialized domain expertise to administer and interpret.
In Table 1, we present our recommendations for brief outcome measures that can be used routinely in cardiology clinics for ICU survivors. These outcome assessments take approximately 5 minutes and can be partially completed by the patient as part of pre-clinic forms. The use of these tools allows for rapid systematic screening for impairments across the spectrum of core manifestations of PICS, including cognitive function and mental health. Abnormalities in these metrics should prompt consideration of referral to domain experts (such as neuropsychologists) for definitive diagnosis and treatment. As the vast majority of CS survivors will follow in cardiology clinic, we have a tremendous opportunity as cardiologists to facilitate helping our patients receive care that reflects the multi-domain nature of PICS.
We also present recommendations for outcome measures that can be used for research purposes or clinically in dedicated post-ICU clinics in Table 1, building on the recommendations of the SHARC panel while incorporating measures to assess cognitive function and mental health as well. 16 The brief research tools are those that can be performed quickly by telephone or in-person, taking approximately 15–20 minutes in total to encompass all major aspects of PICS. These could be used as secondary outcome measures in observational or interventional trials of patients with CS and may be especially valuable in trials that test strategies of care in CS. In addition, Table 1 also includes recommendations for more detailed research tools. Some of these metrics require in-person administration by trained professionals with content expertise such as neuropsychologists and can be used in studies focused on defining specific aspects of PICS in detail among survivors of CS. These assessments have also been translated into languages other than English, facilitating their use for both clinical and research purposes among diverse populations.
Post-ICU survivorship begins on admission to the ICU. While the exact timing of organ disruption that leads to the long-term cognitive, physical, and mental health problems that comprise PICS has not been well-described, particularly in patients with CS, the majority of this injury likely occurs fairly early in the ICU course and may be in large part iatrogenic. This means that the greatest opportunities to improve long-term CS survivorship may present themselves in the early phases of admission. There have been tremendous efforts over the past 30 years, especially in the medical ICU population, that have demonstrated improvements in holistic patient outcomes including survival. These interventions include protocolized spontaneous awakening and breathing trials (SATs/SBTs) for patients receiving invasive mechanical ventilation, which shortened the duration of mechanical ventilation and improved mortality in a randomized trial. 83 Similarly, light sedation has been associated with a shorter duration of mechanical ventilation and lower rates of tracheostomy, 47 while early mobilization is associated with improved muscle strength. 47 It is important to note the common theme among many of these interventions of minimizing iatrogenic harm.
Together these interventions have been collectively termed the ICU liberation bundle, also known as the ABCDEF bundle, by the Society of Critical Care Medicine (SCCM). 84,85 This bundle consists of Assessing/preventing/managing pain, Both SAT and SBTs for intubated patients, Choice of analgesia/sedation (namely avoidance of benzodiazepines), Delirium (assessing, preventing, and managing), Early mobility, and Family engagement/empowerment. 47,84,85 Implementation of the components of the bundle has been associated with significant reductions in mortality and discharge to a facility in a quality improvement study of >15,000 patients across 68 ICUs, of which the vast majority were non-cardiac ICUs. 86 The SCCM does not distinguish across types of ICUs in recommending implementation of the liberation bundle, and an AHA Scientific Statement on the management of CS notes that while the bundle has not specifically been validated in patients with CS, its use should be considered in this patient population. 87 There have been several important initial studies examining components of the bundle in the Cardiac ICU such as protocolized SATs, 88 family engagement, 89 and early mobilization90 which suggest these interventions are safe and effective among patients with cardiac critical illness in general.
The development of multi-disciplinary shock teams has been one of the major advances in CS care over the past decade. These teams have been associated with increased escalation of tMCS and improved survival. 9–11 While much of the data regarding the benefit of shock teams concerns care of patients with refractory or worsening CS, there is also increasing recognition that team-based care is important through the longitudinal course of a patient’s hospitalization. 81,91 Weaning tMCS requires complex decision-making with a paucity of existing evidence, and the medical ICU data would suggest that the de-escalatory phase of critical illness is a period that can significantly impact a patient’s long-term functional status. 91 Ongoing team-based reassessment, potentially with the addition of other specialists such as rehabilitation therapists, geriatricians, palliative care specialists, or mental health practioners, may be helpful to increase cognizance of the multi-dimensional nature of surviving CS and mitigate the potential for iatrogenic harm due to immobility and delayed weaning, among other complications. 81
There are also interventions following the resolution of critical illness which may improve PICS. Post-ICU clinics have been developed at many institutions, including some focused on cardiac ICU survivors. 92,93 These clinics are generally staffed by intensivists and include an interdisciplinary team, particularly pharmacists, social workers, rehabilitation therapists, and mental health specialists. 93 This multidisciplinary approach allows post-ICU clinics to provide a broader spectrum of care than would be found in any other single clinic. Typical care in a post-ICU clinic includes a comprehensive recovery assessment for the primary components of PICS (using validated metrics such as those described in Table 1), medication review, and care coordination, with other interventions such as post-ICU debriefing, advanced care planning, and provision of educational material also performed in many clinics. 93 These clinics may also have established referral pathways to assist patients in establishing with specialists for specific post-ICU problems such as vocal cord dysfunction, dysphagia, or urological problems. Evidence is mixed regarding the efficacy of these resource-intensive clinics in improving patient outcomes, in part due to variability in patients’ pre-ICU health status, type of critical illness, and ICU interventions. 27
Specialized forms of rehabilitation have also been evaluated among ICU survivors, such as intensive physical therapy as well as cognitive rehabilitation, which have shown promise in small studies, albeit with overall inconsistent evidence. 27,39 One of the major challenges of interventions after the ICU is the significant influence of pre-existing comorbidities on a patient’s ability to benefit from these interventions, particularly intensive rehabilitation. 78 Patients who survived respiratory failure and/or sepsis comprise most of the data regarding post-ICU clinics and rehabilitation programs, 93 and these patients have very different pre-admission comorbidity profiles as well as ICU therapies than patients with CS. It is possible that CS survivors may have more potential to benefit from post-ICU interventions than other populations of ICU survivors who have been studied previously due to these differences. Until further evidence emerges to identify how CS survivors specifically should be managed, we suggest that the SCCM consensus recommendations should be followed. This statement regards survivors of shock in general as a high-risk cohort and recommends early screening (within 2–4 weeks of discharge) for impairments using brief metrics such as those in Table 1. 94 Serial assessment may also be important given the evolving nature of post-ICU recovery. Abnormalities in these screening assessments at any point during the post-ICU course should then prompt referral to specialized care to address those domains. There is not sufficient evidence at this time to know whether CS survivors would require specific CS or post-cardiac ICU survivorship clinics rather than general post-ICU clinics. There may be benefit in having such clinics that are staffed by cardiologists, such as the ability to titrate cardiac medications in conjunction with the patient’s primary cardiologist or greater familiarity with specific procedural sequelae related to tMCS devices.
There are numerous gaps in our understanding of CS survivorship, some of which are highlighted in Table 2. First, little is known regarding multi-domain functional status and quality of life among CS survivors. Characterizing the components of PICS among CS survivors will be key to identifying which functional domains are most in need of addressing. PICS following cardiogenic shock may be quite different than PICS following other forms of critical illness due to the unique nature of CS physiology and management. Beyond this, PICS likely manifests differently among patients with different phenotypes of CS, in particular AMI-CS compared to HF-CS, and these subtypes will need to be evaluated individually as part of larger studies. It will also be important to identify the impact of social determinants of health on CS survivorship. 95–97 (Figure 2)
Second, work will need to be done to define the mechanisms and time course of the development of functional impairments following CS. This will include identifying changes in cognition, physical status, and mental health over time, starting within the Cardiac ICU and continuing longitudinally. These mechanisms may very well be different in CS survivors than in survivors of other forms of critical illness. Using cognitive function as an example, when an acute decline in cardiac output due to CS is combined with cardiology-specific interventions such as manipulation of catheters/devices in the aorta, non-pulsatile flow with some tMCS devices, and potential hemolysis from tMCS, it would not be surprising if the mechanisms leading to cognitive dysfunction following CS are distinct from those seen in other post-ICU populations.
Third, observational research will be necessary to first identify inflection points in the post-ICU illness trajectories of CS survivors, followed by testing interventions at these points to improve outcomes in this patient population. It is not known what is the optimal type or timing of rehabilitation among CS survivors, nor is there data regarding the efficacy of intensive follow-up interventions such as post-ICU clinics. Given that most readmissions after CS are for non-cardiac reasons, 12,75 interventions to improve outpatient care following CS will most likely need to take a multi-dimensional approach. It will also be important to define best practices for family engagement across the time spectrum of CS care as well as the impact of CS survivorship on family and caregivers. 98
Finally, as noted by Van Diepen et al, interventions that have been shown to improve outcomes in non-cardiac ICUs have generally not been evaluated broadly in patients with CS. 87 An AHA Scientific Statement on preventing complications in the Cardiac ICU demonstrates the complexities in sedation management among patients with CS by noting the increased risk of delirium with benzodiazepines while also stating that propofol and dexmedetomidine are contraindicated in CS. 99 This serves to highlight the broader need to study the management of multi-organ dysfunction in CS, particularly respiratory failure requiring mechanical ventilation. Significant research is also necessary regarding the holistic management of patients with CS, with trials and observational datasets that examine strategies as well as devices and that assess outcomes beyond mortality. 100 In the same way that randomized trials established the benefit of strategies that minimize sedation and duration of mechanical ventilation such as routine SATs/SBTs, similar studies are necessary to identify optimal strategies for de-escalation of tMCS and patient mobilization while receiving tMCS. 91 This will be particularly important given the recent data from the DanGer Shock study showing improved mortality accompanied by increased complications in patients with AMI-CS treated with micro-axial tMCS. 8 As the positive results of this trial will likely lead to increased use of these devices, it will be essential to develop and test evidenced-based ICU practices for the care of patients treated with micro-axial tMCS while keeping patient-centered outcomes in mind. Ultimately, the greatest benefits in both short- and long-term CS outcomes may be found equally through advancing the science of protocols as much as pumps.
Conclusion
While cardiogenic shock carries a substantial risk of mortality, the majority of patients will survive to discharge, many after experiencing multi-organ dysfunction, invasive ICU interventions, and resulting complications. Subsequently, many of these CS survivors will face the ongoing sequelae of their critical illness long after ICU discharge. Survivors of other forms of critical illness frequently experience PICS, characterized by variable degrees of cognitive dysfunction, impaired physical ability, mental health challenges, and diminished quality of life. While patients with CS experience a unique form of critical illness with distinct pathophysiology and interventions such as tMCS, many CS survivors likely also experience long-term functional impairments. (Figure 3) Both through applying lessons of PICS from other forms of critical illness as well as further research to characterize and improve multi-domain functional outcomes after CS, our clinical and scientific community must embrace the challenge to help our patients not only survive the index hospitalization for CS but also come closer to returning to their prior lives.
