People who survive critical illness often experience long-term cognitive impairment, even among those with normal or near-normal pre-hospital brain function. Cognitive impairment after critical illness is poorly understood; relatively few prospective clinical trials in critical care have followed patients after hospital discharge, and measuring cognitive impairment and determining its onset is difficult.
Cognitive impairment reduces quality of life, increases financial burdens and the risk for subsequent illnesses. With rising numbers of aging adults entering ICUs each year, “ICU brain” could well be a serious, under-recognized and growing public health problem. P.P. Pandharipande et al address the knowledge gap with the BRAIN-ICU study, a cohort study conducted at two Vanderbilit-affiliated hospitals in Nashville, TN.
What They Did
From 2007-2010, authors enrolled 826 truly critically ill patients (respiratory failure, cardiogenic shock, or septic shock) who had not recently been in ICUs who were not already moribund, anoxic-brain-injured, or severely demented. Median age was 61 and they were very sick (median SOFA=9). Only 6% were estimated to have mild-to-moderate cognitive impairment at baseline, as assessed by the IQCODE and CDR assessment tools (these are indirect / inferential assessments: patients’ cognitive function pre-critical illness was NOT directly measured).
Delirium and sedation were followed closely. 74% of patients were delirious according to the CAM-ICU tool; sedation doses were recorded and sedation depth followed with the RASS tool. Multiple other patient factors (organ failure severity, etc) were also followed as covariates.
Blinded psychologists assessed patients’ global cognition and executive function at 3 and 12 months after discharge, using the RBANS and Trail Making Test.
The study was funded by the NIH and the VA.
What They Found
At 3 months, 448 (79%) of the 569 surviving patients underwent cognitive testing:
- 40% had global cognition scores lower than moderate traumatic brain injured patients;
- 26% had scores similar to people with mild Alzheimer’s dementia (2 SD below population means).
At 12 months, 382 (75% of survivors) were retested:
- 34% had traumatic brain injury caliber scores;
- 24% had Alzheimer’s-type scores.
Younger patients (aged <49) were not spared these effects: at one year, 34% had traumatic brain injury-level scores (the same as the overall cohort) and 20% had mild Alzheimer’s-level scores.
Executive function (e.g., the organizational skills needed to perform most jobs) was impaired at 3 and 12 months, with values below population norms among all patient ages.
Longer periods of delirium predicted global cognitive impairment. However, neither duration of coma nor total dosage of sedatives (benzodiazepines, propofol, dexmedetomidine, or opioids) was associated with worse cognitive outcomes. (Higher doses of benzodiazepines predicted poorer executive function at 3 months, but not at 12 months.)
What It Means
Even a full year after recovering from critical illness, 1 in 4 survivors suffered persistent cognitive impairment on par with mild Alzheimer’s, and 1 in 3 had deficits comparable to those in moderate traumatic brain injury. Age was only minimally protective: patients under age 50 — even those with few comorbid conditions — became cognitively impaired as often and almost as severely as older patients.
Although pre-ICU cognitive function was not directly measured, indirect assessments suggested the cognitive impairment was new in a large majority of patients. A quarter of patients were not followed up, and this could have led to a significant bias in the results. However, those who completed follow up were better educated, less frail and more active — which suggests the missing patients might have been as cognitively impaired or more.
Duration of delirium was associated with the risk for cognitive dysfunction, whereas sedation exposure was not. Delirium is associated with neuronal cell death and brain atrophy, but it’s not clear whether delirium causes cognitive dysfunction or is simply a manifestation of another causative process. More to the point, it’s not clear whether interventions to reduce delirium can also reduce cognitive dysfunction. Early mobilization with physical therapy and sleep protocols are the only interventions that have been shown to reduce delirium in the ICU in randomized trials.
Because delirium is so common among ICU patients — 74% were delirious at some point in this trial — most intensivists are probably not attuned to the concept of delirium as an important clinical variable, unless the patients are agitated and climbing out of bed. That’s understandable, since delirium hasn’t yet been directly shown to be detrimental.
“For a long time, we thought delirium was just something that happened because people were in the ICU and that, as soon as they got out of the ICU, they would be okay,” Dr. Karin Neufeld, a psychiatrist at Johns Hopkins and unaffiliated with this trial, told Reuters Health.
If future inquiry can demonstrate interventions that prevent and reduce delirium can also improve ICU outcomes such as cognitive impairment, we could have an exciting new care pathway to reduce the long term burden of critical illness on patients, their families, and society.
For now, simply recognizing the potential severity and persistence of cognitive dysfunction after critical illness, and educating patients, families, and other physicians is an important first step.
Survival from critical illness is cause for celebration, but this study shows that often “you have a new problem to consider when they leave,” Ely said. “And if you didn’t consider that, they’re going to go back out into the world and have a lot of dysfunction in their life.”
“We as health care providers need to be aware of this when patients come to us with memory problems, with problems managing their finances, etc., and not blow it off as saying, ‘This is all going to get better,’ because we really don’t know.” Pandharipande said to USA Today.
P.P. Pandharipande, T.D. Girard, J.C. Jackson et al. Long-Term Cognitive Impairment after Critical Illness. N Engl J Med 2013; 369:1306-1316.