Outcomes and Management After COVID-19 Critical Illness

Abstract

Topic Importance

COVID-19 has caused > 7 million deaths worldwide since its onset in 2019. Although the severity of illness has varied throughout the pandemic, critical illness related to COVID-19 persists. Survivors of COVID-19 critical illness can be left with sequelae of both the SARS-CoV-2 virus and long-term effects of critical illness included within post-intensive care syndrome. Given the complexity and heterogeneity of COVID-19 critical illness, the biopsychosocial-ecological model can aid in evaluation and treatment of survivors, integrating interactions among physical, cognitive, and psychological domains, as well as social systems and environments.

Review Findings

Prolonged illness after COVID-19 critical illness generally can be classified into effects on physical, cognitive, and psychosocial function, with much interaction among the various effects, and includes a wide range of symptoms such as ICU-acquired weakness, prolonged respiratory symptoms, cognitive changes, post-traumatic stress disorder post-traumatic stress disorder, anxiety, and depression. Risk factors for COVID-19 critical illness developing are complex and include preexisting factors, disease course, and specifics of hospitalization in addition to psychological comorbidities and socioenvironmental factors. Recovery trajectories are not well defined, and management requires a comprehensive, interdisciplinary, and individualized approach to care.

Summary

The onset of vaccinations, new therapeutics, and new strains of SARS-CoV-2 virus have decreased COVID-19 mortality; however, the number of survivors of COVID-19 critical illness remains high. A biopsychosocial-ecological approach is recommended to guide care of COVID-19 critical illness survivors.

COVID-19 has strained health care systems worldwide, causing staggering volumes of severe illness and > 7 million deaths.¹ Evolving evidence sheds light on the dynamic patterns of hospitalizations and outcomes of this viral infection.^2,3 Although the number of patients requiring ICU care has decreased with vaccinations, new therapeutics, and newer strains of the virus, a significant burden of critical illness persists. In the United States alone, approximately 1,000 people are still hospitalized in an ICU for COVID-19-related illness in as recently as August 2023.⁴

Although mortality has improved,⁵ the persistent sequelae among the growing number of survivors of COVID-19 presents another challenge to health care systems navigating a global pandemic.⁵ Long-term consequences of critical illness are well described and included in post-intensive care syndrome (PICS). PICS is defined as new or worsening physical, cognitive, or psychological deficits, or a combination thereof, after an ICU admission.⁶ This clinical syndrome encompasses a heterogeneous group of patients that includes survivors of COVID-19 critical illness. Recently, the PICS framework has been used to organize deficits in survivors of an ICU stay with post-COVID conditions to help provide a patient-centered approach to recovery.⁷ Concurrently, SARS-CoV-2 may have unique and lasting consequences in those who do not experience severe disease. The sequelae of non-critically ill patients are beyond the scope of this review.

Together with the PICS framework, and given the multidimensional symptoms after COVID-19 critical illness, the biopsychosocial-ecological model can aid in evaluation and treatment of survivors, recognizing the key interactions among physical, cognitive, and psychological domains, as well as social systems and environments. The biopsychosocial model was introduced in 1977⁸ as a model for holistic medical care and since has been updated to include systems theory, viewing health conditions through a framework that considers interactions among individual characteristics, interpersonal relationships, community settings, and wider societal policies, with applications for prevention, public health, and clinical care.^{9, 10, 11, 12} Reference to these multidimensional frameworks is included throughout this review, which describes the current understanding and approach to survivors of COVID-19 critical illness with persistent symptoms and dysfunction.

Literature Review

Our team included experts in pulmonary and critical care medicine, PICS, rehabilitation, and neuropsychology. We developed areas of emphasis in this review through discussion and iterative revision. We completed a literature search in PubMed and PsycINFO to identify relevant literature using the following search terms: critical illness, intensive care unit, follow-up, and post-intensive care syndrome AND COVID-19, prioritizing literature that addressed physical, cognitive, psychosocial issues, or a combination thereof, in COVID-19 critical illness. However, because of the limited number of articles available that focused on COVID-19 critical illness, articles that included a mixed severity of COVID-19 illness also were reviewed.

Evidence Review

Epidemiology

The number of ICU admissions for COVID-19-related illness has varied throughout the pandemic, peaking in the United States in January 2021, with > 28,000 patients in an ICU. This number decreased to fewer than 1,000 patients in June and July 2023.⁴ Survival to hospital discharge of patients with COVID-19 critical illness also has varied throughout the pandemic. A systematic review completed in 2020 found that the overall ICU mortality rate was 25.7%, although 56% of patients were still in the ICU at the time of the study completion.¹³ A retrospective cohort review from February 2020 through March 2021 showed mortality rates of ICU admissions for COVID-19 at 30.7%.¹⁴ These results reveal large number of survivors of COVID-19 critical illness. Among these survivors, approximately 15% experience post-COVID-19 conditions, which refer to symptoms after COVID-19 that linger after recovery from acute infection.¹⁵ In addition to post-COVID-19 conditions, small cohort studies among survivors of COVID-19 critical illness found rates of PICS as high as 69%.^{16, 17, 18, 19, 20} Figure 1, adapted from Mikkelson et al,²¹ demonstrates COVID-19 and pandemic-specific factors superimposed on a model of PICS. This highlights acute illness and hospitalization factors that can contribute to poor outcomes as well as pandemic-specific factors that exacerbate PICS. These factors include²¹ a strained health care system, decreased access to care, increased health care provider burnout, increased isolation, and increased uncertainty on both the part of the patient and the provider related to the novel nature of the disease.^22,23

Figure 1.

Diagram showing COVID-19 pandemic and post-intensive care syndrome factors impacting post-ICU outcomes. (Adapted with permission from Mikkelson and Iwashyna.²¹ Persistent problems and recovery after critical illness. In: Warrell DA, Cox TM, Firth JD, eds. Textbook of Medicine. Oxford Univeristy Press; 2020:3925-3930.21. Copyright Oxford University Press, 2021.Reproduced with permission of the Licensor through PLSclear.)

Pathophysiology

The mechanisms underlying symptoms after COVID-19 critical illness remain uncertain and are likely multifactorial, possibly in part because of direct tissue damage in the setting of the viral infection and the immune response of the host.²⁴ A pronounced stress response and overactivation of the hypothalamic paraventricular nucleus may contribute to psychological factors.²⁵ Similar to the cause of PICS, which also is poorly defined, long-lasting effects likely are related to common ICU interventions such as sedation, prolonged immobility, and delirium, as well as social-ecological factors such as presence of interpersonal and systemic support.²⁶

General Risk Factors

Risk of prolonged illness after COVID-19 critical illness is complex, and risk factors include preexisting factors (age, comorbidities, frailty), disease course (organ system insults), COVID-19-specific factors (viral load, variant), as well as psychological comorbidities and social factors (social support). Vance et al²⁷ (Fig 2) proposes a model highlighting the biopsychosocial-ecological framework that encompasses these complex factors.

Figure 2.

Diagram showing possible factors influencing symptom reporting acutely, postacutely, and long after a COVID-19 infection. CDC = Centers for Disease Control and Prevention; PTSD = post-traumatic stress disorder.

Physical Function

Risk Factors

Increased age and frailty are risk factors for COVID-19 critical illness developing,^28,29 as well as risk factors for development of ICU-acquired weakness, defined as multidimensional functional disability including impairments such as compromised ambulation, impaired activities of daily living, swallowing difficulties, and prolonged mechanical ventilation.³⁰ Additional risk factors for physical impairment include longer duration of bed restriction, hyperglycemia, and use of neuromuscular blockade or corticosteroids, which became standard of care for COVID-19 critical illness.^{30, 31, 32, 33, 34} Additionally, multiple organ dysfunction or prolonged duration of illness also can contribute to development of ICU-acquired weakness.³⁵ Although pulmonary manifestations of COVID-19 are not limited to critically ill patients, one of the strongest predictors for post-COVID-19 pulmonary fibrosis developing is invasive ventilation during inpatient admission.³⁶

Prevalence and Clinical Manifestations

Physical impairments are reported frequently among survivors of COVID-19 critical illness and encompass a variety of conditions, often overlapping, with ICU-acquired weakness being the most common. Among a cohort of 92 survivors of COVID-19 critical illness in Belgium, 87.5% did not return to their previous level of activity at 3 months after discharge.³⁷ Among a cohort of 486 patients hospitalized with COVID-19, 52% of patients treated with mechanical ventilation had ICU-acquired weakness at ICU discharge.³⁸ In addition to ICU-acquired weakness, survivors of CCI are at risk for development of diaphragmatic dysfunction, laryngeal injury, dysphagia, and dysphonia primary due to prolonged invasive mechanical ventilation.^33,39

Pulmonary manifestations (eg, dyspnea, cough) after both severe and nonsevere COVID-19 infection are among the most commonly described sequelae. Multiple studies of survivors of COVID-19 critical illness have reported impairment of pulmonary function test results at 3 months after hospital discharge, including reduced diffusion capacity of lungs for carbon monoxide in 52% to 75% of patients and decreased total lung capacity (TLC) in 37% to 48% of patients.^{40, 41, 42, 43}

Radiographic evidence of interstitial lung disease has been observed after varying acuity of illness with COVID-19 pneumonia.⁴⁴ Residual lung abnormalities are common and can be seen in up to 11% of patients after hospitalizations for COVID-19-related illness; greater risk of these abnormalities can be seen with severe illness including invasive mechanical ventilation or extracorporeal membrane oxygenation. Organizing pneumonia is the most common persistent interstitial lung disease among survivors of COVID-19, accounting for 59% of interstitial changes in a cohort of 59 patients.⁴⁴ A meta-analysis found that 44.9% of survivors of COVID-19 demonstrate fibrotic changes on imaging.⁴⁵

Approach

The diagnosis and treatment of post-COVID-19 conditions can be difficult given the novelty, heterogeneity in presentation, and lack of standardized screening or diagnostic criteria.^46,47 The 2020 Society of Critical Care Medicine’s International Consensus Conference on Prediction and Identification of Long-Term Impairments after Critical Illness recommended that an assessment of patients for PICS should occur between 2 and 4 weeks after discharge.⁴⁸ Given the significant overlap of PICS and post-COVID-19 conditions in those who were critically ill and the evolving understanding of post-COVID-19 critical illness, currently a similar model can be used for evaluation of post-COVID-19 conditions. The 6-min walk test and EQ-5D-5L are recommended screenings for physical impairments seen in PICS.⁴⁹ The World Health Organization has developed a scale for assessing functional status across multiple domains (World Health Organization Disability Assessment Schedule 2.0), including physical functioning, and post-COVID-19-specific measures are beginning to emerge, with the Post-COVID-19 Functional Status Scale being particularly useful in assessing level of participation in usual daily activities with a physical function component; see Table 1^{50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63} for a comprehensive list of measures and references.

Table 1.

Suggested Assessment Methods After COVID-19

Domain	Assessment	Description	Considerations
Physical or general functional status	6-min walk test50	•Measures aerobic exercise capacity •Instructs patients to walk as far as possible for 6 min	•Can help provide information pertaining to pulmonary and cardiovascular systems, blood circulation, and peripheral circulation
	EQ-5D-5L51	•Assesses health-related quality of life across five dimensions: mobility, self-care, usual activities, pain or discomfort, and anxiety or depression	•Available in > 150 languages •Various methods of administration (eg, self-report, interviewer administered) •Learn more at https://euroqol.org/eq-5d-instruments/eq-5d-5l-about/
	World Health Organization Disability Assessment Schedule 2.052	•Assesses functioning across six domains: cognition, mobility, self-care, getting along with others, life activities, participation •Higher scores represent higher levels of disability (both total score and domain-specific scores reported)	•Multiple versions exist (eg, 36-item vs 12-item) •Various methods of administration (eg, self-report, interviewer administered) •Available in > 30 languages
	Post-COVID-19 Functional Status Scale53	•A structured interview that assesses degree of functional impairments in daily life after infection (ADLs and iADLs, participation in social roles, persisting symptoms, and care needs)	•Assessment of functional status before COVID-19 is recommended •Intended to be assessed at time of discharge, in the first few weeks after discharge, and up to 6 mo after COVID-19 to track functional status over time
Psychological functioning	Impact of Events Scale-Reviseda54	•Twenty-two items measuring symptoms of PTSD across the three categories of PTSD symptoms (avoidance, intrusion, hyperarousal symptoms)	•Available in several languages •Avoid diagnosis of PTSD based on screening alone
	Patient Health Questionnaire-955	•Nine items that assess the presence and intensity of depressive symptoms	•Widely used in health care settings, covers all symptoms for MDD based on DSM-5 criteria •Safety risk: includes suicidal ideation item that should be checked on completion •The first two items can be used for an ultrabrief screening (Patient Health Questionnaire-2) •Available in > 30 languages translations (see https://www.phqscreeners.com/)
	General Anxiety Disorder-756	•Seven items designed to assess the presence and severity of symptoms of generalized anxiety disorder	•Also adequately screens for panic disorder, social anxiety, and PTSD •The first two items can be used for an ultrabrief screening (General Anxiety Disorder-2) •Available in > 20 languages translations (see https://www.phqscreeners.com/)
	Health Anxiety and Depression Scale57a	•Fourteen items (seven anxiety and seven depression) assessing the presence of these symptoms in medical patients	•Designed to reduce the impact of physical symptoms on scoring (eg, depression items focus on anhedonia instead of sleep difficulties and fatigue) •Translated into > 70 languages
	Posttraumatic Stress Symptoms-1458	•Fourteen items assessing symptoms of PTSD according to DSM-4 criteria	•Commonly used in patient populations in the ICU
	PTSD Checklist for DSM-559	•Twenty-item measure that assesses presence and severity of PTSD symptoms based on DSM-5	•Commonly used within service members and veterans
Cognitive functioning	Montreal Cognitive Assessmenta60	•Sixteen items •Requires 10 min to administer	•Broad assessment of multiple cognitive domains, including executive functions •Designed to be sensitive to subtle forms of cognitive impairment •Requires certification (visit https://www.mocatest.org/training-certification/)
	Montreal Cognitive Assessment-Blind61	•Removes visuospatial executive and naming tasks of MoCA •Requires < 10 min to administer	•Can be used when in-person assessment is not feasible or with patients with vision or motor impairments
	5-min Montreal Cognitive Assessment62	•Includes four tasks based on MoCA (word list, verbal fluency, orientation, delayed recall) •Requires 5 min to administer	•Briefest version of the MoCA available •Does not require in-person assessment
	Repeatable Battery for Neuropsychological Status63	•Requires 30 min to administer •Stimulus book required	•Allows for continuously examining performance within the domains of immediate memory, attention, language, visuospatial, and delayed memory •Higher sensitivity to change than global cognitive screens •Alternative forms available for repeat assessment

ADLs = activities of daily living; iADLs = instrumental activities of daily living; DSM-4 = Diagnostic and Statistical Manual of Mental Disorders; DSM-5 = Diagnostic and Statistical Manual of Mental Disorders; MDD = major depressive disorder; MoCA = Montreal Cognitive Assessment; PICS = post-intensive care syndrome; PTSD = post-traumatic stress disorder.

^aRecommended by the Society of Critical Care Medicine.

Recommendations for primary care providers evaluating patients after COVID-19 illness focus on identifying those with more serious complications, including physical manifestations of pulmonary embolism, heart failure, stroke, myocardial infarction, and lung fibrosis,⁶⁴ for referral to specialists while managing mild to moderate symptoms in the general medicine setting in a collaborative fashion with the patient and family. Providing a comprehensive and user-friendly workup for primary care provider reference (eg, recommended evaluation and treatment by system and symptom, what can be treated by the primary care provider and when to refer to a specialist) may be useful for decreasing the need for multiple coordinated specialists.²⁷ For example, dyspnea and cough, common after varying acuity of COVID-19 illness, may require additional testing, including chest imaging or pulmonary function testing and consideration for pulmonary referral, particularly if the patient has progressive dyspnea, crackles on examination, symptoms of > 12 weeks’ duration, a sit-to-stand test with ≥ 4% oxygen desaturation, concerning findings on chest radiograph or spirometry, or a combination thereof.⁶⁵

Pulmonary rehabilitation and physical therapy also are important tools in COVID-19 critical illness recovery. Prospective studies have found that patients across the spectrum of COVID-19 illness who completed both inpatient and outpatient pulmonary rehabilitation programs showed improvement in objective measures of both functional and mental status on completion.^66,67 It is recommended to consider pulmonary rehabilitation for all patients who have been hospitalized for COVID-19.⁶⁸ As a result of the pandemic, use of telerehabilitation programs increased, allowing the patient to complete pulmonary rehabilitation at home; early studies of have shown that these virtual programs can provide a safe alternative to in-person sessions.⁶⁹

Cognitive Function

Risk Factors

Cognitive status before illness and during the ICU stay both are risk factors for development of prolonged post-COVID-19 cognitive symptoms. Preexisting dementia and other cognitive vulnerabilities, such as risk factors for vascular disease or mild cognitive impairment, increase the risk of delirium during acute COVID-19 infection, which subsequently may increase the risk of long-term cognitive impairment.⁷⁰ Similar to physical impairments, risk of cognitive impairment, particularly for those with the aforementioned vulnerabilities, is compounded by longer duration of mechanical ventilation, bed restriction, and use of neuromuscular blockade or corticosteroids, although these factors are likely important because they contribute directly to length of delirium.^{30, 31, 32, 33, 34}

Rates of ICU delirium for patients with severe COVID-19 range from 40% to 70%, making it one of the most common complications.^{71, 72} Limited data demonstrate variable rates of cognitive dysfunction after acute delirium among patients with COVID-19. Some studies show similar rates of cognitive dysfunction in those with and without delirium, although with much greater physical decline in those with delirium at the 4-week follow-up, whereas others show significant increase in cognitive dysfunction after COVID-19 with acute delirium.^71,73

Prevalence and Clinical Manifestations

Observed rates of cognitive difficulties have a wide degree of variability because of the heterogeneity in assessment methods (eg, self-report vs cognitive testing, different measurement tools) and timing of measurement. In an observational series of 45 of 51 consecutive patients with ARDS resulting from COVID-19 who were discharged from an ICU during the first epidemic wave, Daste et al⁷⁴ found that all patients experienced some level of cognitive dysfunction in the short term. The most common cognitive difficulties seen in the short term are in the domains of processing speed, working memory, and executive functions such as set shifting and divided attention, although significant improvement tends to occur by 3 months after discharge, which is consistent with other causes of PICS.⁷⁵ Longer-term cognitive manifestations after an ICU stay with COVID-19 can range from nonexistent to prolonged disorders of consciousness.^18,76 Many survivors of ICU stays for COVID-19 will not sustain clinically significant, long-term cognitive impairment when cognition is assessed objectively and appropriate comparison groups (norms) are used, although objective cognitive changes in the short term seem to be much more common.^75,77 Self-report of cognitive concerns tends to be higher, indicating that factors other than neurocognitive impairment, such as psychological distress, sleep disturbance, and medication effects, may contribute to cognitive difficulties after COVID-19.^75,78 Neurologic complications such as stroke increase the likelihood of long-term cognitive sequelae after COVID-19, again with similar presentation to strokes of other causes.⁷¹

Approach

Recommended screening tools for assessment of cognitive status include the Montreal Cognitive Assessment, or Montreal Cognitive Assessment-Blind if administering to a patient with visual difficulties or via telemedicine platforms. The 5-min Montreal Cognitive Assessment is an even briefer, although well validated, screening tool that may be useful for providers who have little time to assess general cognitive status. These tools now require that the test administrator be certified in the administration and interpretation of the instrument to maintain validity. Certification takes approximately 1 h.

The Repeatable Battery for Neuropsychological Status is a much more comprehensive cognitive screening measure that takes approximately 45 to 60 min to complete. The resultant data provide index scores for various cognitive domains, including attention, language, visuospatial and constructional abilities, and immediate and delayed memory. Indexes also exist that are fairly accurate in determining whether the patient’s presentation is most consistent with a cortical or subcortical process and may help to differentiate between something like Alzheimer disease and new-onset subcortical dysfunction related to hypoxia. This measure requires significant training before administration and interpretation, although typically is administered by a properly trained professional who is not a neuropsychologist. See Table 1 for a comprehensive list of measures and references.

Similar to the assessment and treatment of physical symptoms, nonspecialists may manage cognitive symptoms for many patients. Referral to a neuropsychologist is recommended if performance is impaired on a cognitive screener and cognition is preventing the patient from functioning as they would like.⁷⁹ Treatment of objectively confirmed cognitive impairment may include cognitive rehabilitation with a rehabilitation psychologist or speech language pathologist. Although randomized controlled trials are just being started to determine specific cognitive rehabilitation techniques for post-COVID-19 conditions, it is likely that findings will be similar to what is already known to be effective in other similar conditions.⁷⁹ Cognitive rehabilitation should include an emphasis on compensatory strategies and environmental supports to optimize independence because these strategies are more evidence-based than most remediation approaches.⁸⁰ Mass-marketed computer-based cognitive games have not been proven efficacious.⁸⁰ For subjective cognitive concerns that are not supported with objective measures, as well as objective cognitive impairment, assessment of and education on noncognitive factors that may be contributing to the experience of cognitive impairment, such as mood, anxiety, pain, and medication effects, should be provided, and these factors should be addressed with evidence-based treatment.

Psychosocial Function

Risk Factors

Psychosocial factors confer significant risk for development of COVID-19-related post-ICU conditions, and impairments in psychosocial function can result from COVID-19 critical illness. Risk factors can include psychological symptoms during acute illness, impaired adjustment or coping in the aftermath, and level of social support, both at the personal and the systems level.^16,17 Notably, psychological status impacts health behaviors and is a key factor in adherence with health care recommendations that will optimize outcomes.⁸¹ Improvements in depression and anxiety are associated with better adherence in recently hospitalized patients.⁸² Preexisting psychological conditions also may be a risk factor for development of severe disease itself. In a large systematic review and meta-analysis comprising 1,469,731 patients with COVID-19, of whom 43,938 showed preexisting psychological disorders, Vai et al⁸³ found that the presence of any psychological disorder was associated with an increased risk of COVID-19 mortality (OR, 2.00). This association also was observed for psychotic disorders (OR, 2.05), mood disorders (OR, 1.99), substance use disorders (OR, 1.76), and intellectual disabilities or developmental disorders (OR, 1.73), but not for anxiety disorders. Exposure to antipsychotics (OR, 3.71), anxiolytics (OR, 2.58), and antidepressants (OR, 2.23) also was associated with COVID-19 mortality.

Prevalence and Clinical Manifestations

Psychological manifestations after an ICU stay for COVID-19 span a wide range, from healthy adjustment to debilitating mental illness. Literature examining psychological manifestations specific to critically ill patients with COVID-19 is limited and has shown variable results. Some studies have found that patients meet post-traumatic stress disorder criteria less frequently and reported less severe symptoms of anxiety and depression than a historical critical illness cohort, whereas other studies indicate acute stress disorder occurs in about 40% of patients with COVID-19 in the ICU.^84,85 Taken together, the existing research indicates that psychological distress may be more pronounced early on in COVID-19 critical illness, but may dissipate more quickly than in other critical illnesses. In a qualitative study conducted early in the pandemic, researchers followed up patients through their hospitalizations, identifying COVID-19-specific attitudes and stressors, including high levels of fear, denial, and stigma associated with COVID-19 infection. The major sources of stress included the contagious nature of the disease, quarantine measures, and concerns regarding the health of family members. Manifestations of stress included excessive attention to symptoms, rumination, and changes in diet and sleep. Supportive factors included psychological flexibility in adjustment, perception of good medical care, and family and social support. For some, the disease resulted in psychological growth because patients viewed their experience with gratitude through the cherishing of life, family, bravery, and tenacity.⁸⁶

Approach

Recommended screening tools for assessment of psychosocial status include the Hospital Anxiety and Depression Scale for anxiety and depression and the Impact of Events Scale for post-traumatic stress disorder (see Table 1 for a more comprehensive list of measures).⁸⁷ Similar to the assessment and treatment of physical and cognitive symptoms, nonspecialists may manage psychosocial concerns for some patients, although a low bar should be in effect for referrals to social work for management of risk factors relating to social determinants of health and to a psychologist or other qualified mental care professional for psychological concerns that prevent the patient from functioning as they would like. Referral to a rehabilitation psychologist may be particularly beneficial because they have expertise in helping patients adjust to illness or disability. It is important to ensure that evidence-based techniques are used, such as cognitive-behavior therapy or acceptance and commitment therapy for anxiety and depression and cognitive processing therapy or prolonged exposure therapy for post-traumatic stress.⁸⁸ The referring provider can use motivational interviewing skills to encourage psychotherapy if a patient expresses ambivalence.⁸⁹

Encouraging and educating patients in self-management of chronic symptoms also is important for patient self-efficacy and optimal adjustment and to reduce unnecessary use of the medical system. PASCguide.com offers patient resources for education, self-management strategies by symptom, and methods for tracking symptom management that may aid with understanding what works or does not work for a particular patient.⁹⁰

Clinical Outcomes

Prospective research to inform understanding of recovery trajectories for this population is limited, and for the most part, long-term studies are limited to 6-month and 1-year follow-up periods. Marked variability in outcomes suggests that factors such as case mix (eg, age, disease burden, comorbidities) and other variables that likely impact findings (eg, selection or survival biases, assessment methods) should be considered, although this also highlights underlying variability in outcomes. One exploratory prospective multicenter cohort study of ICU survivors with COVID-19 found that at 1 year, 74.3% of patients reported physical symptoms (eg, overall weakness, joint stiffness and pain, muscle weakness, and myalgias), 26.2% reported psychological symptoms, and 16.2% reported cognitive symptoms.⁹¹ However, another study found that approximately 50% of patients with COVID-19 who were treated in the ICU experience full recovery of symptoms from 3 months to 1 year, although for some, symptoms of depression and anxiety remain largely unchanged despite improvements in physical functioning.¹⁸

Prior to the COVID-19 pandemic, studies evaluating pulmonary function after ARDS demonstrated normalization of spirometry values and lung volumes at 6 months for most patients with a more prolonged impairment in diffusion capacity of lungs for carbon monoxide.^43,92 A study comparing patients requiring mechanical ventilator for acute respiratory failure resulting from COVID-19 compared with those requiring mechanical ventilator for causes other than COVID-19 did not find a difference in rates of new disability at 6 months, suggesting that the overall trajectory and recovery of physical impairments are similar for ARDS resulting from both COVID-19 and all other causes.⁹³ A recent multicenter study of patients with severe to critical COVID-19 showed a high prevalence of radiologic and pulmonary function abnormalities at the 3-month follow-up, but continued improvement at the 6-month and 12-month follow-up.⁹⁴

Persistent cognitive dysfunction after COVID-19 critical illness is another commonly reported concern and likely is associated with acute factors discussed previously (eg, delirium and psychosocial complexity). Some degree of cognitive dysfunction based on screening measure results is present in 33% to 42% of survivors of ICU stays for COVID-19 6 months after discharge.^95,96 However, comprehensive studies assessing long-term cognitive outcomes in survivors of ICU stays for COVID-19 are limited. Another study examined patients with COVID-19 who were and were not treated in the ICU with comprehensive neuropsychological testing. Attention and executive functioning domains were impacted the most, and in contrast to studies suggesting no differences in cognitive functioning between hospitalized and not-hospitalized patients with COVID-19,⁹⁷ ICU-treated patients demonstrated more severe impairments. Studies with longer follow-up periods currently are lacking, although one survey suggests that clinically significant cognitive dysfunction is reported by 16% of patients 1 year after discharge.⁹¹

Clinically significant depression and anxiety symptoms are more prevalent than post-traumatic stress disorder both at the 6-month⁹⁵ and 1-year⁹¹ follow-up, with rates of depression, anxiety, and post-traumatic stress disorder of 42%, 38%, and 33% at 6 months and 18%, 18%, and 10% at 1 year, respectively. These rates are similar to those of survivors of ICU stays who do not have COVID-19.⁹⁸ Within patient populations with ARDS, psychological dysfunction has been documented up to 5 years after ICU discharge.⁹⁹ Taken together, these findings highlight the importance of understanding who is most at risk for persistent symptoms to provide early intervention.

Future Directions

A complex relationship exists between disease-specific and patient-specific factors in post-COVID-19 critical illness, with important implications for optimizing outcomes. Findings of the limited studies thus far often are in disagreement with one another. For example, some teams have reported unexpectedly similar cognitive performances among survivors of COVID-19 who required ICU care and those who did not, whereas others reported differing results commensurate with severity of acute illness.¹⁰⁰ In some studies, among those with impaired neuropsychological scores, no relationship was found between disease severity and cognitive functioning, suggesting that other key negative prognostic variables (eg, premorbid factors and other psychosocial-ecological complications, as well as secondary medical complications) should be examined carefully. These variable findings suggest that interpretation of any one test result should not be overgeneralized because many different causes underlying a patient’s experience of physical, cognitive, or psychological dysfunction are likely across the stages of severe acute and post-acute COVID-19. Additional studies examining these nuanced relationships, accounting for biopsychosocial-ecological factors, are required to tailor treatments better and to predict and optimize outcomes.

Understanding recovery trajectories with longer follow-up periods is another valuable area of future research, and more opportunities to accomplish this will be available as time elapses since the outset of pandemic. Given what has been observed with other patient populations, we know that post-traumatic stress symptoms can have a delayed onset of up to 1 year¹⁰¹ and that psychological symptoms can persist more generally,⁹⁹ indicating that this likely will be a patient population with ongoing follow-up and intervention needs. However, just as future research will be tasked with fully understanding negative prognostic variables, defining what constitutes and predicts a good recovery will further our understanding of post-COVID-19 trajectories and will inform possible intervention targets. For example, Sun et al⁸⁶ provide insight into protective factors associated with improved emotional adjustment, including psychological flexibility and perception of good medical care. Post-COVID-19 recovery clinics in addition to post-ICU recovery clinics have increased since the onset of the pandemic, with interdisciplinary care, but have variable approaches to structure, patient evaluation and assessment, and treatment.¹⁰² Continued investigation on the impact of these clinics on patients who have undergone an ICU stay and who have survived COVID-19 is warranted.

Limitations with study design also point to future areas of needed research. Some studies have strict exclusion criteria, including excluding those with preexisting psychiatric diagnoses or neurologic disease,^{97,103, 104, 105} limiting not only the generalizability of findings, but also our understanding of recovery trajectories in those who may be at highest risk of poor outcomes. Similarly, fully accounting for sample characteristics is necessary to avoid making misleading interpretations. For example, different conclusions may be drawn from studies with samples comprising mainly young, previously fit patients as opposed to studies including older patients with a high comorbidity burden. Additionally, many ICU-specific studies are limited by smaller sample sizes, which prohibits more nuanced analysis of data.^75,84,106

Summary

The advent of vaccinations and new therapeutics has lowered COVID-19 mortality, but the number of survivors of ICU stays for COVID-19 remains high. Research that is specific to caring for these patients is sparse, and additional methodologically sound research, such as prospective studies accounting for the nuanced interactions of disease-specific and patient-specific factors, as well as intervention randomized controlled trials, are needed. In the interim, early findings suggest that a model similar to critical illness recovery care may be helpful. A biopsychosocial-ecological approach is recommended to address the many factors that may contribute to prolonged symptom experience after COVID-19 critical illness.¹⁰⁷ Helping patients move toward self-management, when possible, will be an important part of the care model to decrease burden on the health care system and prevent provider burnout.

Funding/Support

This study was supported in part by the US Department of Veterans Affairs, Health Services Research and Development Service [Grant VA IIR 20-313]. T. E. is supported by the VA Office of Academic Affiliations through the VA/National Clinician Scholars Program (NCSP) and the University of Michigan Medicine at University of Michigan. T. V. is supported by the National Heart, Lung, and Blood Institute [Grants K23 140165 and R01 HL157361 and the Agency for Healthcare Research and Quality [Grant R01 HS028038].

Financial/Nonfinancial Disclosures

The authors have reported to CHEST the following: H. P. receives support from Agency for Healthcare Research and Quality (unrelated to this manuscript) and serves on the Surviving Sepsis Campaign Guidelines for Sepsis and Severe COVID. None declared (L. C., K. S., E. N., T. V., T. E., J. I. M.).