Long-term outcomes of critical illness

Despite a continuous increase in the incidence of critical illness syndromes such as sepsis and acute respiratory distress syndrome (ARDS), improvements in supportive care have resulted in improved survival over the past few decades. For example, sepsis is estimated to affect between 30 and 50 million people each year worldwide. ^, Global sepsis mortality has declined from 30% in 1990 to 20% in 2017, resulting in an increasing number of sepsis survivors with variable prognosis. ^, Half of patients recover, one-third die during the following year, and one-sixth have severe persistent impairments. Impairments include development of an average of one to two new functional limitations; a threefold increase in prevalence of moderate to severe cognitive impairment; and a high prevalence of mental health problems, including anxiety, depression, or posttraumatic stress disorder, for which survivors will seek care from many types of clinicians ( Table 166.1 ). Current guidelines provide no recommendations on management for these complex patients after discharge, and there is a desperate need for improvement. According to an international survey of 1731 critical illness survivors, the majority reported dissatisfaction with or complete lack of support services after hospital discharge.

TABLE 166.1

Common Sequelae in Survivors of Critical Illness

Neurocognition	Behavioral	Neuromuscular	Immune System	Caregiver & Family
Cognitive dysfunction	Depression	Polymyopathy	Immunosuppression	Caregiver burden/burnout
Attention deficits	Anxiety	Polyneuropathy	Inflammation	PTSD
Impaired executive function	Posttraumatic stress disorder (PTSD)	Chronic pain	Recurrent infection	Anxiety
Chronic fatigue	Insomnia		Exacerbation of chronic disease	Depression

Patients who survive their initial acute illness but consequently experience persistent organ failures necessitating prolonged intensive care meet the definition of chronic critical illness (CCI). ^, This illness is characterized by high hospitalization costs, frequent post–acute care use, and poor long-term survival. For example, a third of sepsis survivors are readmitted within 90 days, and rehospitalization after sepsis accounts for 12.2% of all US hospital readmissions and 14.5% of readmission costs. The clinical and financial burden of CCI is expected to increase in the coming years because of both an aging population and improvement in short-term survival. In a population-based sample of five US states, we found that the prevalence of CCI based on a consensus definition has been increasing over time, with associated in-hospital costs exceeding $25 billion per year. These findings underscore the importance of CCI to the field of critical care, particularly for healthcare policy and planning. Spending on CCI is likely to rise further as the population ages, because the prevalence of CCI increases dramatically with age. The most common initial diagnoses that lead to CCI are acute respiratory failure requiring mechanical ventilation and sepsis. Hospital-acquired infections may contribute to mortality among hospitalized patients and to the burden of CCI.

With this shift in critical illness epidemiology, traditional outcomes such as short-term mortality need to be complemented by patient-centered outcomes that capture long-term health-related quality of life (HRQoL), physical and mental functioning, time spent at home, hospital readmissions, and morbidity and mortality ( Fig. 166.1 ).

In this chapter, we review recent advances in our understanding of long-term outcomes after critical illness. Most literature focuses on long-term outcomes after ARDS and sepsis. We describe the epidemiology, risk factors, clinical manifestations, management, and outcomes for each domain of post–intensive care syndrome (PICS).

Cognitive impairments after critical illness

Critical illness frequently results in new cognitive impairment, including deficits in memory, attention, and concentration. The reported incidence has been highly variable depending on study population and trial size, ranging from 4% to 62%. Cognitive impairment after intensive care unit (ICU) admission is also taxing to patients and their families and carries an enormous societal cost estimated at $18 billion per year.

Risk factors for long-term cognitive impairment, particularly preventable ones, are not well understood. A complicating factor is the uncertainty of whether critical illness and/or its treatment is the cause of the observed neurocognitive impairments or if it worsens preexisting subclinical disease. A prospective cohort study in elderly adults who had no known cognitive dysfunction at baseline showed that acute care or ICU hospitalization resulted in a greater cognitive decline and incident dementia compared with individuals who did not require hospitalization. Similarly, Pandharipande and colleagues reported new severe cognitive deficits affecting memory, attention, processing speed, and executive function in a broad population of 328 ICU survivors that were detectable at 3 and 12 months after critical illness. The etiology of cognitive impairments after critical illness is likely multifactorial and results from various factors that interact dynamically with preexisting and acute illness variables to produce adverse outcomes. Acute illness severity alone does not explain the cognitive impairments experienced by ICU survivors. Although current data do not suggest a clear association between age, acute illness severity scores, cumulative doses of analgesia and sedation, hypoxia, hypotension, hyperglycemia, and inflammation with adverse neurologic sequelae, the duration and severity of acute delirium have emerged as the strongest predictors of adverse cognitive outcomes in several recent studies.

In the study by Pandharipande and colleagues, longer duration of delirium was independently associated with worse cognitive performance at 3 and 12 months in tests of global cognition and executive function. Similar findings were reported in a small prospective study of 77 patients with ARDS who had in-person cognitive testing 3 and 12 months after ICU discharge. The duration of delirium was an independent risk factor for cognitive decline at both time points even after adjusting for preexisting cognitive impairment, age, severity of illness, level of education, severity of sepsis, and cumulative exposure to sedatives. Interventions to prevent or shorten delirium in critically ill patients are therefore much needed and center around promoting and improving the quality of sleep. Ziprasidone and haloperidol, two drugs commonly used to treat ICU delirium, neither increased the number of days alive without coma or delirium nor reduced 90-day mortality, time to freedom from mechanical ventilation, and time to ICU or hospital discharge in a large multicenter placebo-controlled trial. Nocturnal administration of low-dose dexmedetomidine (average dose 0.36 μg/kg/h) was recently reported to prevent delirium, reduce days spent with coma, and have an opiate-sparing effect in a randomized placebo-controlled trial in critically ill adults. Interestingly, dexmedetomidine did not improve patient-reported sleep quality. An earlier systematic review also suggested potential for dexmedetomidine and quetiapine to prevent ICU delirium, but neither drug influenced long-term outcomes.

Early identification of cognitive impairment should in theory expedite appropriate evaluations and treatment. Evaluation for cognitive dysfunction in the critical care setting must be brief, easy to administer, and widely applicable. Currently available tests such as the modified mini-mental state examination (MMSE) and Montreal Cognitive Assessment (MoCA), however, cannot be used to predict long-term cognitive dysfunction. ^, Evolving research suggests that patients who experience critical illness–induced delirium and persistent cognitive dysfunction may have a variety of abnormal findings on neuroimaging, including brain atrophy; leukoencephalopathy; and neuronal loss in the insula, frontal lobes, and thalamus. In addition, a multicenter prospective cohort study in 321 critically ill patients showed generalized slowing on routine clinical electroencephalography (EEG) correlated with delirium severity, length of hospitalization, functional outcomes, and mortality and may be a promising biomarker to identify patients at high risk for adverse outcomes early during hospitalization.

Integrated approaches specifically geared toward the management of mechanically ventilated patients have proven useful, with the most widely applied being the ABCDEF bundle. Initially focusing on daily sedation interruption and spontaneous breathing trials, this bundle also emphasizes the importance of delirium management, delirium prevention, early mobility, and family engagement. Despite never being formally tested in clinical trials, clinical practice suggests that the ABCDEF bundle is safe and may improve delirium outcomes in real-world settings. ^, These findings were confirmed in a recent prospective multicenter cohort of over 15,000 adults from the ICU Liberation Collaborative, in which use of the ABCDEF bundle was associated with a reduced incidence of delirium along with improvements in other patient-centered outcomes.

Psychiatric sequelae of critical illness

Psychiatric disorders, including depression, anxiety, and posttraumatic stress disorder (PTSD), are common among critical illness survivors. For instance, in a recent systematic review, Rabiee and colleagues reported that approximately 30% of ICU survivors had clinically significant depression. Early post-ICU depressive symptoms were a strong risk factor for subsequent depressive symptoms, and post-ICU depressive symptoms were associated with substantially lower HRQoL. Patients exposed to mechanical ventilation and longer ICU lengths of stay are at higher risk for developing a new diagnosis of “mental illness” compared with noncritically ill hospitalized patients. Further study is necessary to better understand patient predisposition, illness, and treatment-specific determinants of affective morbidity and appropriate tools for diagnosis and monitoring. Another important and understudied area is the frequency of suicide after critical illness, particularly given the increasing prevalence with age and burden of chronic illness.

Several studies have examined the relationship between critical illness and the development of PTSD. Schelling and colleagues were the first to introduce the concept of PTSD resulting from critical illness and traumatic experiences in the ICU. ^, Among their cohort of 80 long-term ARDS survivors, almost one-third reported impaired memory, nightmares, anxiety, and sleeping difficulties after ICU discharge, with a PTSD prevalence rate of 28%. A recent meta-analysis of prevalence, risk factors, and prevention/treatment strategies for PTSD symptoms in critical illness survivors reported clinically important PTSD symptoms in approximately one-fifth of critical illness survivors at 1-year follow-up. Patients with underlying psychiatric comorbidities, benzodiazepine use, or early memories of frightening ICU experiences had the highest prevalence of PTSD. A randomized controlled trial in 35 French ICUs studied the use of an ICU diary filled by caregivers and family members on PTSD symptoms, anxiety, and depression in patients and family members at 3 months after ICU discharge. In this study, the overall prevalence of significant PTSD symptoms was 29.9% among ICU survivors and was not reduced by an ICU diary. Equally disappointing, a nurse-led complex psychological intervention that was initiated in the ICU did not reduce the prevalence of self-reported PTSD symptoms at 6 months.

Critical illness–associated neuromuscular dysfunction

ICU-acquired weakness is common in patients with ARDS and other complex critical illness. Regardless of disease process, muscles and nerves are injured, resulting in prolonged mechanical ventilation and poor functional outcomes. Previous research has highlighted the concept of a continuum of weakness that begins with muscle injury documented within hours of mechanical ventilation that may persist with incomplete recovery for years after ICU discharge. Muscle weakness and impaired function constitute an important morbidity of severe critical illness. In a recent prospective cohort study that used daily point-of-care ultrasound to measure diaphragm thickness in mechanically ventilated patients, almost 50% of patients had a 10% reduction in diaphragm thickness by ICU day 4. Diaphragm atrophy was associated with prolonged mechanical ventilation and ICU length of stay. Interestingly, development of increased diaphragm thickness as a marker of excessive inspiratory effort was also associated with prolonged mechanical ventilation.

Critical illness polyneuropathy

Critical illness polyneuropathy (CIP) primarily manifests itself as a mixed sensorimotor neuropathy. CIP is quite common in patients with systemic inflammatory response syndrome and sepsis, with an occurrence of 70%–100% of patients with a longer ICU stay. It affects limb and respiratory muscles, whereas facial muscles are usually spared. Limb involvement is symmetric and most prominent in the proximal muscle groups of the lower extremities. Detection of the true incidence of CIP is complicated by lack of consensus on surveillance, timing, and nature and limitations of testing because of patient sedation or poor cooperation, formal definition, and diagnostic criteria. Weakness may initially be absent or difficult to detect clinically in these patients, but subsequent electromyography testing will demonstrate abnormalities showing an initial primary axonal degeneration of the motor neurons, followed by the sensory neural fibers, and this coincides with acute and chronic changes of denervation noted on muscle biopsies in affected patients.

In sepsis, the pathogenesis of CIP is linked to a perturbation in the microcirculation, with resultant axonal injury and degeneration. There is also evidence for a disruption of nerve action potential, which may be reversible over the course of the disease. Other risk factors associated with the development of CIP include hyperglycemia, and tight glycemic control has been shown to reduce the incidence of CIP in critically ill patients. The exact pathophysiologic link between glucose control and neuroprotection remains unclear but may involve preservation of mitochondrial functioning, calcium homeostasis, or modulation of nitric oxide production. In contrast to earlier associations between neuromuscular dysfunction and use of neuromuscular blockers, subsequent research has not been able to corroborate this previous association. ^, Data on the association between glucocorticoid use and weakness remain controversial; however, a national multicenter prospective trial in acute lung injury survivors reported a significant association between mean daily corticosteroid dose with impairments in physical outcomes after 1 year.

Critical illness myopathy

The reported incidence of critical illness myopathy (CIM) varies between 48% and 96% in prospective studies that have included muscle biopsy as part of their diagnostic evaluation. Pathologically, CIM is characterized by a diffuse, nonnecrotizing myopathy associated with fatty degeneration of muscle fibers, fiber atrophy, and fibrosis. This has been described in patients with sepsis and in those treated with corticosteroids and neuromuscular blockers. Patients clinically appear weak and paretic and are difficult to wean from the ventilator. They may be indistinguishable from patients with CIP. Muscle biopsy allows differentiation among these lesions. ^,

The pathophysiology of CIM entails catabolism, inflammation, and derangement of membrane excitability. Protein catabolism and an increase in urinary nitrogen loss are observed in CIM. Muscle biopsies in affected patients show low glutamine, protein, and DNA levels. There is evidence for the upregulation of the calpain, caspase-3, and ubiquitin proteolytic pathways in concert with an increase in apoptosis, but mitochondrial dysfunction and oxidative stress likely do not play a significant role.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here