Chapter 86: Post-Intensive Care Syndrome

Over the past decade, survival from critical illness has dramatically increased due to a better understanding of the pathophysiological mechanisms of disease, improved treatment strategies and advancements in medical technology. Several studies have shown improved survival and long-term outcomes in survivors of critically illness. However, surviving the intensive care unit (ICU) stay is just the start of a long road to recovery for a majority of these patients. The discharge from the ICU opens the path to a long journey of challenging physical rehabilitation, mood disorders, cognitive impairment, psychological distress, financial hardship, and caregiver burden and burnout.

In recent years there has been a growing recognition of impairments that affect the physical, psychological, social, and emotional aspects of the individual after ICU discharge that may adversely impact daily functioning and quality of life (QOL). Recently, the term “post-intensive care syndrome” (PICS) is used to describe any new or worsening impairments in physical, cognitive, or mental health status arising after critical illness and persisting beyond the acute care hospitalization.¹ PICS may persist for months to years after hospital discharge. Most impairments will diminish with time but some may linger on until the patient’s actual demise. This chapter will explore in detail the different domains affected in PICS, its impact on the individual and society, and offer insights into future developments.

Probably the most obvious and readily recognizable changes in patients immediately after discharge from the ICU are the physical impairments that are the result of the actual critical illness or the direct or indirect side-effects of interventions to treat the disease. An individual patient may lie at a particular point in the spectrum of physical disability according to the patient’s age, level of functioning prior to the onset of critical illness, and the burden of co-morbid conditions. At one end of this spectrum, clinicians encounter young, previously working and highly functional patients who develop a severe catastrophic illness. The middle range includes patients who are older and have a greater burden of comorbidities than the previous group. Finally, the opposite end of this spectrum includes those patients who have experienced chronic critical illness, or the very elderly, in whom ICU-level of care may not alter their ultimate outcome but may instead contribute to incremental disability and constitute part of a downward functional trajectory driven by progression of chronic illness.

Lung Function

It has been shown that in critical illness leading to respiratory failure, particularly in acute respiratory distress syndrome (ARDS), that lung function is decreased soon after recovery but improves to normal or near-normal over 6 months to 5 years. The diffusing capacity for carbon monoxide (DLCO) is the lung function parameter that seems to be mostly affected as DLCO values still remain mildly reduced or low-normal even after 5 years of follow-up.² This persistent impairment in gas transfer is probably due to injury at the capillary level, which promotes thickening in alveolar capillary interfaces, pulmonary fibrosis, and pulmonary vascular remodeling.

Anatomical changes in the lung are also observed in follow-up imaging studies of ARDS survivors. Localized changes in the nondependent lung zones including reticular changes, ground-glass opacities and minor pulmonary fibrosis are seen on high-resolution computer tomography (CT) scans.³ The association between the severity of lung injury and length of mechanical ventilation may reflect ventilator-associated lung injury. However, the relationship between the development of lung fibrosis after ARDS and any possible risk factor is not straightforward. Some studies have found significant correlation between CT scan impairment and duration of mechanical ventilation, level of positive end-expiratory pressure (PEEP), and oxygen fraction. These data may only reflect greater severity of ARDS, which can be responsible per se for lung fibrosis.

It is interesting to note that the severity of the patient’s dyspnea after recovery from ARDS does not seem to correlate with actual lung dysfunction but may reflect extrapulmonary muscle weakness and sometimes psychological impairments.

Chronic Respiratory Failure

The onset of respiratory failure requiring prolonged mechanical ventilation is associated with increased morbidity, mortality, and health care costs. ICU admission for pneumonia, ARDS, neuromuscular disease, head trauma, or postoperative intracerebral hemorrhage is one of the strongest predictors of prolonged mechanical ventilation.⁴ ICU-acquired weakness (ICU-AW) has also been shown to be a predictor of failure to be liberated from the ventilator. The mechanism that is responsible for a majority of ventilator dependence can be explained by an increase in respiratory load coupled with decreased respiratory muscle performance. Chronic ventilator dependence may result in complications similar to those receiving short-term mechanical ventilation. These include infections, tracheal bleeding or malformations, renal failure, pneumothorax, volume overload, ileus, and seizures.

The overall mortality in patients with ventilator-dependent chronic respiratory failure is high and up to 52% at one year from the initial hospitalization.⁵ Chronic irreversible neurologic diseases and presence of skin breakdown has been associated with increased risk for mortality. These patients may be discharged to home, long-term acute care (LTAC) facilities, skilled nursing facilities or hospice care centers. The QOL tends to be low but may improve over the years. In particular, ARDS survivors who require prolonged mechanical ventilation have poorer QOL than other ARDS survivors.⁶ Health care resource utilization in these patients have been shown to be exceeding high and much of this is spent on ongoing and recurrent medical care.

Weakness

Another important aspect of the physical impairment after recovery from critical illness is weakness. Potential contributors to weakness in survivors of critical illness are listed in Table 86–1. Risk factors for the development of weakness after critical illness are enumerated in Table 86–2. Several contrasting studies have shown variable association between the development of weakness and the severity of illness on ICU admission (eg, SAPS-2, APACHE II/III, SOFA). High blood glucose has been identified as a risk factor through an unknown mechanism while intensive insulin therapy has been shown to be a preventive factor against critical illness-associated weakness and to decrease the risk of critical illness polyneuropathy. The data on the association of corticosteroids with the development of weakness has been conflicting; however one study reported decreased neuromuscular dysfunction in patients on intensive insulin therapy while on corticosteroids, suggesting that when euglycemia is maintained, the anti-inflammatory effect of steroids may benefit the neuromuscular system.⁷ Low doses of neuromuscular blockers (eg, paralytics) do not seem to be associated with weakness but larger doses may be independently associated.

Formerly described by different entities such as critical-illness polyneuropathy (CIP), critical-illness myopathy (CIM), and critical-illness neuromyopathy, ICU-AW is the prototypical functional impairment that has been the subject of intense research even prior to the recognition of the other domains of PICS. The incidence of ICU-AW varies from 30% to 90%. It may be missed during the acute phase of critical illness when the patient is sedated, restrained and unable to communicate. Difficulty liberating from mechanical ventilation may be the first indication of an impairment.

CIP is an axonal polyneuropathy that affects both sensory and motor nerves. The causes of axonal degeneration include the systemic inflammatory response syndrome (SIRS), ischemia to nerves from hypotension, microthrombosis, changes in microcirculation, endoneural edema, inflammation, mitochondrial dysfunction causing bioenergetic failure, neurotoxic meds, and metabolic abnormalities. Physical examination may reveal distal sensory deficits, distal weakness, and preserved deep-tendon reflexes (DTRs). Electromyography-Nerve Conduction Velocity (EMG-NCV) studies demonstrate decreased amplitudes of sensory and motor nerve action potentials with normal motor unit potentials and excitability. Pathologic evaluation reveals fiber loss and primary axonal degeneration of both motor and sensory nerve fibers, most severe distally.

CIM includes flaccid tetraparesis with depressed or absent DTRs but sensory function is not affected. The causes include increased muscle breakdown, decreased protein production and abnormal muscle repair, derangement of energy delivery, changes in the muscle membrane with primary inexcitability of muscle fibers, inflammation, steroids, and immobility. NCV studies reveal decreased amplitudes of compound muscle action potentials with preserved sensory nerve action potentials. Direct muscle stimulation demonstrates reduced or absent muscle excitability. Pathology most commonly reveals loss of thick filaments with atrophy of the type II fibers.

In terms of assessing weakness, the six-minute walk test (6MWT) is most widely used as it is an integrated outcome parameter that is dependent on the motor, pulmonary, and circulatory function. This test has proven to be a simple but useful test of global physical recovery in former ICU patients.

Neuro-cognitive impairment spans a wide range of dysfunction in critically ill patients. From those who are comatose to patients with minor, sometimes subtle dysfunction that may only be revealed using specific tests of cognitive dysfunction, although close relatives may have already recognized behavioral changes in these patients. Among survivors from general, medical and surgical ICU, the incidence of cognitive impairment ranges from 4% to 71%, while the incidence of cognitive impairment in ARDS survivors ranges from 4% to 56%.³ Two studies in elderly critically ill patients have reported a prevalence of cognitive impairment varying from 17% to 56%.^8,9 Cognitive impairment is associated with a reduced QOL and is a major determinant of societal health care costs and care-giving needs. The elderly are prone to developing cognitive impairment; however, it appears that younger, relatively healthy patients are also at risk of cognitive impairment following critical illness. Risk factors for the development of cognitive dysfunction after critical illness are listed in Table 86–3. It is unclear whether a low performance on neuropsychological tests reflects impairment in cognitive functioning related to critical illness and ICU admission, or whether it is perhaps merely a marker of patients with poor health and an increased risk of ICU admission. However, studies that include premorbid cognitive data show that at least part of the measured cognitive impairment is related to the ICU admission and critical illness.

Memory, attention, verbal fluency and executive functioning are the domains most frequently impaired after critical illness. The pathogenesis is not fully understood but may represent an accelerated neurodegenerative process that develops in vulnerable patients. Cognitive impairments can also be associated with newly acquired brain damage due to insults from critical illness such as hypoxemia, hypotension, anemia, fever, hyper-/hypoglycemia, systemic inflammation, severe sepsis, pharmacologic agents, disrupted sleep, renal failure, and liver failure. Severe sepsis can lead to a neuroinflammatory response, resulting in increased levels of cytokines in the brain.¹⁰ Elevated cytokine levels are associated with impaired memory in healthy volunteers, and neuroinflammation is associated with the development of Alzheimer’s disease. Long-term cognitive impairment in patients may therefore represent a maladaptive version of cytokine-induced disease. Delirium in the ICU can cause long-term complications such as cognitive impairment and can also significantly decrease short-term and long-term survival and worsen QOL after critical illness. Increased duration of delirium likewise is associated with worse neurocognitive outcomes in a “dose-dependent” manner.¹¹

Radiologic and pathologic findings associated with cognitive dysfunction associated with critical illness include ischemic and hypoxemic hippocampal lesions, brain atrophy, ventricular enlargement, decreased superior frontal lobe and hippocampal volumes, and loss of white matter in the corpus callosum and internal capsule. At 1 year, these anatomical findings were associated with worse overall cognitive performance and worse executive functioning. Left hippocampal volumes on MRI were likewise markedly reduced in a cohort of patients with septic shock.

Individuals who are hospitalized for a critical illness have a greater likelihood of cognitive impairment, even after adjusting for premorbid cognitive screening scores and comorbidity. It has been suggested that critical illness may cause an abrupt loss of cognitive function rather than accelerate the decline in cognitive functioning. Cognitive dysfunction is probably very frequent in the immediate short period after intensive care but tends to normalize in most patients with time. Time to improvement varies from different studies. Improvement towards normal cognitive functioning has been reported to occur as early as 9 months after ICU discharge. Other studies have shown no improvement from 1 up to 5 years of follow-up.

Studies have so far been unable to identify patients at higher risk of cognitive impairment using brief cognitive screening tools at time of hospital discharge. One candidate predictor for cognitive impairment might be quantitative electroencephalogram (EEG). A recent study of sepsis survivors found EEG to be a potential predictor of cognitive impairment in these patients. Deficits in verbal learning and memory were associated with a significant reduction in left hippocampal volume and low-frequency activity on routine EEG. EEG may be able to provide prognostic information, possibly in combination with other modalities. Serum biomarkers may also be useful in predicting cognitive impairment. Elevated serum interleukin (IL)-6 and c-reactive protein (CRP) concentrations are associated with reduced cognition and contribute to accelerated functional decline in both the elderly and in post-operative cardiopulmonary bypass patients. Given the prevalence of delirium in disease states with a higher systemic inflammatory burden, inflammatory biomarkers may be useful for monitoring delirium disease activity and predicting risk of long-term cognitive impairment.

Lower percentages of patients with cognitive impairments were reported in studies using screening tests as compared to studies that utilized extensive neuropsychological testing. A major limitation of most studies is that a baseline assessment of cognitive status before the onset of critical illness is lacking and is oftentimes difficult to determine unless prior premorbid cognitive screening results were available. Ideally, pre-ICU admission cognition should be available because the real interest is not the absolute level of cognitive performance but rather the change in cognitive functioning.

Another limitation is that cognitive testing is seldom standardized from center to center and from cohort to cohort making the comparison between studies difficult. Furthermore, a large number of different tests and combinations have been used making the interpretation and application of the results to critically ill patients challenging. The most widely used tests include the mini-mental status examination (MMSE) for screening patients for cognitive dysfunction and the Cambridge Neuropsychological Test—Automated Battery (developed at the University of Cambridge) as a more in-depth test for cognitive dysfunction in research studies.

The increasing awareness of the PICS has led to the realization that new onset behavioral changes and psychiatric symptoms are the result of the ICU experience in survivors of critical illness. Akin to the psychological trauma experienced by survivors of natural or man-made catastrophes, the time spent in the ICU can be as harrowing as a near-death experience to many. The reported prevalence of anxiety and depression after ICU discharge varies between 6% and 28% among different studies depending on the “cut-off” scores and methodology used. Levels of anxiety and depression appear to decrease in the first year after ICU discharge; however, it is unclear whether this improvement remains or continues to full resolution of symptoms. On the contrary, in patients with ARDS, levels of depression increased from 16% at 1 year after discharge to 23% at 2 years. It has also been suggested that 14% to 27% of critically ill patients may develop a posttraumatic stress reaction or post-traumatic stress disorder (PTSD). Importantly, post-traumatic stress does not appear to decrease over time after ICU discharge and may endure for a number of years. Criteria for the diagnosis of PTSD are defined in the DSM-IV-TR. Factors associated with an increased likelihood of both anxiety, depression and PTSD after ICU discharge are listed in Table 86–4.

Patients who had neuropsychological impairment after ICU care had statistically significantly higher depression scores than those who did not require ICU-level of care.¹² Increased severity of illness does not seem to predict an increased risk of developing psychological dysfunction, but rather it is both the subjective and objective aspects of the ICU experience that seem to be associated with it. In addition, these subjective and objective indicators are associated with the severity of anxiety and depression experienced by patients after ICU discharge.

The subjective indicators include the reported unpleasant memories of being in the ICU. Patients often report disturbing recollections and these experiences are often persecutory in nature and are associated with feelings of being elsewhere reliving a previous life event or fighting for survival. The lack of memory for actual events may mean that this is the way in which patients process these delusions or unreal experiences that result in longer term psychological problems. In fact, it has been proposed that the content of the ICU memories was more important than their number, and that delusional memories were more likely to result in distress than factual memories, even if these were unpleasant.¹³ Patients frequently describe these memories as extremely vivid and real. The traumatic memories may also be linked to the recollection of certain persistent physical symptoms. Some patients may make strenuous efforts to block these traumatic memories but, paradoxically, this may only cause them to occur more frequently.

The objective indicators include variables such as length-of-stay (LOS), duration of sedation and/or neuromuscular blockade, mechanical ventilation, and other respiratory supportive therapies. Some studies show that decreased time under sedation or in mechanical ventilation result in decreased post-traumatic stress symptoms and anxiety. It is interesting to note that there has been some inconclusive evidence to support the role of corticosteroid supplementation during acute illness in preventing PTSD during recovery—as PTSD is associated with abnormalities of the hypothalamic–pituitary–adrenal axis and in stress response in general.

Psychological and behavioral outcome after ICU care has been mainly assessed using standardized questionnaires with demonstrated reliability and validity. Tools used to assess anxiety, depression, and PTSD are listed in Table 86–5.

Impact on the Patient

Health-related quality of life (HRQoL) sums up the effects of the ICU stay in survivors of critical illness and is probably the best documented of all non-mortality outcomes after hospital discharge. HRQoL encompasses both the physical and psychological aspects of the individual’s overall well-being. HRQoL after ICU discharge may range from normal to reduced in different patient populations and are usually compared to the “normal” population in different studies. Elderly survivors tend to have decreased HRQoL mostly in the physical domains. Survivors of severe trauma tend to have more permanent reductions in their HRQoL due to the fact that their HRQoL probably is already reduced at baseline from alcohol and substance abuse, even though these patients are typically younger and frequently do not suffer from prior chronic organ dysfunctions. Studies that compare pre-ICU HRQoL with post-ICU data indicate that ICU patients in general express a reduced HRQoL prior to ICU admission compared to the matched general population. The burden of preexisting medical conditions is one of the most significant factors associated with a reduction in HRQoL after ICU care. This strongly suggests that in most acutely ill ICU patients, their HRQoL after ICU discharge probably will never equal a matched general population. Working status or the ability to return to work is also a frequently reported outcome measure and is in general found to be low and reported to range from 23% to 35% in different studies.

Compounding the already burdened survivor, depression has been associated with poorer HRQoL. However, the direction of this relationship is not clear, and it may be that patients with a poorer HRQoL have a more prolonged recovery and thus are more likely to be depressed. This is an important issue because it suggests that if other aspects of HRQoL are better, then perhaps emotional outcome might be improved. Depression may interfere with the speed of recovery, and in some patients may lead to suicidal attempts.

Impact on Caregivers

There is no question that caregivers have a devastating, parallel, but different experience compared with their loved one. For the family and other caregivers, the daily and overwhelming stress experienced while the patient is in the ICU leaves its mark, and they similarly may experience compromised HRQoL and mental health, including PTSD, emotional distress, caregiver burden, depression and anxiety.¹⁴ There is a strong relationship between high levels of PTSD-related symptoms in family members and those in patients. In some cases, family therapy may be needed. As a result returning to the hospital or doctor’s office for outpatient appointments can be very stressful but, at the same time, it may be therapeutic and provide needed reassurance of their loved one’s recovery. For bereaved families, the incidence of PTSD may be even higher and often goes completely unrecognized. Even when home care is provided by family members, there is still great opportunity cost, as these individuals must sacrifice other endeavors to free up time at home. Caregivers themselves also may face significant cognitive morbidity simply as a consequence of caring for their chronically ill family members.

Impact on the Health Care System

Along with personal costs, there are significant financial costs associated with surviving an episode of critical illness. ICU survivors face long hospital LOS and post-discharge care that frequently involves expensive skilled nursing and long-term acute care (LTAC) facilities. Compared with the extensive resources devoted to helping the patient survive their ICU stay, far fewer resources are devoted to improving outcomes in the post-ICU period and systems are not in place to facilitate the patient’s transition from the acute to the chronic phase of their illness. Transferring patients to LTACs decreases hospitals costs, but increases total costs, with no consistent improvement in long-term outcomes. In the era of intensivist physician staffing and hospitalist medicine, it is becoming less common that a single physician provides both hospital care and post-discharge care. Instead, the responsibility for post-discharge care falls on busy general practitioners, who are frequently unaware of events in the ICU, and often lack the time and expertise to diagnose the myriad of sequelae of critical illness.

Physical and Neurocognitive Domains

Controlling the modifiable risk factors known to be associated with the different aspects of PICS is probably the first step in mitigating its effects on ICU survivors. These would include maintaining euglycemia, minimizing use of sedation and neuromuscular blockers, early liberation from mechanical ventilation, judicious use of corticosteroids, and early exercise and mobilization.¹⁵ Early identification of ICU-AW will be the key in initiating early physical and occupational therapy. Early exercise and mobilization has been shown to increase the number of patients being discharged from the hospital with an independent functional status. Studies have also demonstrated a link between exercise and cognition through improved cerebrovascular function. Identifying exercise as a mechanism that reverses or alters the trajectory of cognitive impairment would have an important impact on the HRQoL in survivors of critical illness.¹⁶ Furthermore, combining both cognitive and physical rehabilitation has been shown to result in significantly better executive functioning and fewer disabilities in instrumental activities of daily living. An ICU sleep-promotion initiative has likewise been shown to reduce incident delirium and cognitive impairment. Better sleep efficiency may contribute to improving HRQoL and reducing fatigue to allow more effective participation in physical rehabilitation.

Psychological Domain

The psychological stress resulting from the ICU experience has been shown to be partly contributed by inaccurate memories during the ICU stay. Therefore, it has been suggested that providing survivors with an ICU diary may improve their understanding of their ICU experience and act as a “debriefing” tool. ICU diaries may be compiled by any of the ICU staff members, particularly nurses, and also by family members. It can be a written account complemented by photographs or videos. It should be viewed as a form of exposure to accurate and potentially corrective information, thereby reducing the anxiety caused by inaccurate and oftentimes frightening memories of the ICU stay. Also, reading or viewing of the ICU diary is entirely voluntary at a time of the patient’s choosing and should be followed by a discussion of the contents with a member of the ICU staff. This should be viewed as a first step in learning to modify and control feelings and reduce physiological arousal as a step to recovery. Those patients who have received ICU diaries had lower levels of PTSD-related symptoms than those who did not.¹⁷

A stepped care approach with different degrees of therapeutic intervention depending on patient need may represent a more viable and flexible model for recovering critical illness survivors. This may involve designing or using off the shelf self-help programs to help individuals coping with lower levels of symptoms of anxiety, panic or depression. Those patients not responding to this or who exhibit greater symptom levels may then be referred to specialist psychological services for cognitive behavioral therapy or eye movement desensitization and reprocessing. These are therapies designed to reduce the emotional impact of distressing and traumatic memories and facilitate new ways of thinking. A stepped care approach may offer the appropriate and timely help to the maximum of patients, and may be more cost-effective compared with a more conventional approach of referring all patients to a counselor or clinical psychologist.

As the public is slowly becoming more aware of the PCIS because of increased reporting in mainstream media, efforts to further expand knowledge on this syndrome and to implement interventions to treat or mitigate its effects are more important than ever. With the percentage of the elderly population rising to much higher numbers in the next few decades, coupled with increasing survivorship from the ICU, we may be facing a public health emergency. Future research should be geared towards standardizing the definition of cognitive impairment, defining the spectrum of disability after critical illness, refining the associations of risk factors, detailing specific patient and family-centered outcomes, and improving study designs to address long-term patient-centered functional outcomes. In addition, much more research is needed to evaluate the effectiveness of educational strategies that raise awareness and promote treatment of post-ICU morbidity among general practitioners.

A potentially useful post-intensive care recovery intervention is the concept of the post-ICU clinic. From recent experience from the United Kingdom, it has been shown to increase understanding of the longer term recovery from critical illness.¹⁸ However, the provision of services in different clinics is varied, unspecialized and oftentimes inconsistent, and tend to have funding problems hence restricting the number of patients who can be seen. Currently, it is still unclear whether these clinics should be staffed by intensivists or general practitioners.

Nonetheless, a well-designed care model similar to disease management programs for diabetes, heart failure, and COPD, will facilitate the transition of patients and their families to the outpatient setting, allow for early recognition of post-ICU complications and sequelae, increase access to a variety of health care providers, and improve HRQoL. Likewise, principles of longitudinal care models developed from acute stroke care, cardiac rehabilitation, and post-traumatic brain injury may be applied to survivors of critical illness.

Recent data shows that PCIS is common and involves the physical, cognitive, and mental health status in survivors of critical illness. These impairments are frequently under recognized and adversely impact the daily functioning and overall QOL. It is critical that interventions to prevent the sequelae of acute critical illness begin during the ICU stay and involve not only the ICU staff but those involved in the care of the patient after the acute care hospitalization.