Better but Not Well: Disability, Frailty, and Cognitive Impairment One Year after COVID-19 Critical Illness

The first wave of the coronavirus disease (COVID-19) pandemic, from December 2019 to August 2020, was responsible for over 24 million confirmed cases worldwide, with approximately 5% of infected adults developing rapidly progressive respiratory failure (1, 2). Mortality rates for critically ill patients varied significantly during 2020 across regions, ranging from 15% to 60%, with a global assessment of the in-hospital mortality rate estimated at 28% (3). The only figure perhaps more staggering than the death estimates from 2020 may be the unprecedented number of COVID-19 critical illness survivors. Persistent physical, mental, and cognitive impairments after critical illness, termed the post-intensive care unit (ICU) syndrome (4), are common, but how they might present and evolve in critical COVID-19 is as yet poorly understood.

In this issue of AnnalsATS, Taniguchi and colleagues (pp. 289–295) present a prospective cohort study of 428 critically ill COVID-19 survivors in Brazil and report on longitudinal recovery trajectories of a wide range of clinical outcomes (5). Expanding on their prior report describing increased rates of frailty and disability at 90 days after hospital discharge (6), this new study extends serial assessments of frailty and disability as well as cognitive impairment over 12 months of follow-up. The authors found that new disabilities, frailty, and cognitive symptoms peaked 3 months after discharge and then slowly decreased, but many patients never returned to the prehospitalization assessments. At 12 months of follow-up, over one-third of patients had more disabilities than in the pre–COVID-19 period. Nearly half of the 12% of patients with frailty at 12 months had not been frail before COVID-19, and cognitive impairment appeared to be new for most (60%) of the 12% of participants experiencing it at the end of follow-up.

Key strengths of the study include the rigorous methods used to determine baseline and longitudinal assessments of disability and frailty status with relatively little missingness. Many COVID-19 cohorts from this timeframe lacked baseline assessments, limiting our understanding of whether physical and cognitive symptoms were indeed new and were limited by higher rates of missing data (7). The serial assessments at multiple time points also generate new knowledge regarding the recovery trajectories of patients over the course of a year. The incorporation of geriatrics-trained medical investigators to assess frailty outcomes is another important strength of this study, as geriatricians have been shown to have superior interrater reliability on the Clinical Frailty Scale (8).

A key limitation of the study is the baseline measure of cognitive status. Preexisting cognitive impairment was defined as a previous diagnosis of dementia, significant memory complaints, or disability in any activity of daily living. However, this definition was not validated; it is unclear how significant memory complaints were assessed, and disability in any activity of daily living does not equate to baseline cognitive impairment. Moreover, the baseline measure of cognitive status is unlikely to capture baseline mild cognitive impairment, which is a risk factor for post-ICU cognitive impairment (9).

Persistent symptoms and chronic disability are not limited to patients experiencing critical illness from COVID-19 (10, 11). The current study does not include a comparison group of other critically ill patients without COVID-19, making it challenging to understand whether the recovery trajectories and persistent deficits are unique to COVID-19 itself. However, we can compare their results to those described in other critically ill patient cohorts from before the COVID-19 pandemic. A United States-based multicenter cohort study previously demonstrated similar patterns of post-ICU frailty, disability, and cognitive impairment trajectories but had overall higher rates of transitions to a worse frailty state between enrollment and 12-month follow-up (12). Other prior studies assessing ICU outcomes at more frequent intervals have consistently shown that disability peaks shortly after ICU discharge, with functional recovery generally occurring over the 6 months thereafter (13, 14). The peak observed by Taniguchi and colleagues at 3 months is likely a function of the longer intervals between follow-up assessments; had outcomes been assessed at 1 and 2 months after ICU discharge, transitions in disability and frailty status may have been even more striking. Future studies, including comparisons across patients with different degrees of severity of acute COVID-19, may help distinguish between deficits likely attributable to more general post-ICU syndrome and postcritical–COVID-19–specific disability.

The finding that even after 1 year, nearly one-third of patients were in a worse physical state than before the ICU admission has major implications for the lasting burden of illness. However, it remains unclear how much of this is driven by COVID-19–specific pathology versus a shift in practice patterns (15) during the peak of capacity strain caused by the first wave of the pandemic. The patients enrolled in this cohort were all hospitalized from March to July 2020, well before the advent of vaccination and improved treatments for COVID-19. It is uncertain how directly we can apply the findings from Taniguchi and colleagues to current clinical practice. Ongoing longitudinal assessments of hospitalized and critically ill patients with COVID-19 among more contemporary cohorts are needed. Moreover, detailed patient-level information about treatment characteristics, such as mode of mechanical ventilation, use of paralytics and sedatives, and receipt of physical therapy, and important potential mediators, such as the onset of delirium, would enhance our understanding of how frailty, disability, and cognitive impairment are related to both underlying disease and ICU management. Future studies could help disentangle how much of a patient’s recovery trajectory is predicted by a patient’s baseline functional status, pre-ICU trajectory, and characteristics of the ICU stay.

The study by Taniguchi and colleagues makes an important contribution to the literature on COVID-19 critical illness and identifies areas in need of further research and intervention. Although the recovery trajectories of COVID-19 survivors and rates of persistent disability and frailty may be similar to those seen with other critical illnesses, the sheer volume of patients and resultant downstream implications will be profound. Understanding these impacts is crucial to help families, health systems, and policymakers prepare for the ongoing needs of survivors and their communities.