Covert Critical Illness Encephalopathy: Impairments That Escape Detection by Guideline Recommended, Protocolized Assessments

Abstract

Objectives: To determine whether cognitive impairments of important severity escape detection by guideline-recommended delirium and encephalopathy screening instruments in critically ill patients.

Design: Cross-sectional study with random patient sampling.

Setting: ICUs of a large referral hospital with protocols implementing the Society of Critical Care Medicine's ICU Liberation Bundle.

Patients: Patients with a heterogeneous mix of primary organ system conditions leading to critical illness and with no abnormal findings scored in Confusion Assessment Method for the ICU (CAM-ICU) screening, Richmond Agitation-Sedation Scale (RASS) 0, and Glasgow Coma Scale (GCS) 15, indicating they were alert, fully oriented, and following commands with no delirium or findings to indicate subsyndromal delirium.

Interventions: None.

Measurements and main results: We evaluated 50 patients, age 54 ± 16 years. Trained critical care nurses assessed patients at regular intervals using the CAM-ICU, RASS, and GCS per a protocol. We performed a battery of psychometric cognitive tests using the NIH Toolbox. Executive functions linked to attention and inhibitory control, and processing speed were 1.5 sd below population norm (both p < 0.01). Working memory and cognitive flexibility were also significantly, but less severely, impaired ( p < 0.01 and p = 0.026). Nearly two-thirds (64%) of the patients scored at least 1.5 sd worse than demographically adjusted means in two or more cognitive domains, a commonly used diagnostic criterion for cognitive impairment.

Conclusions: Substantial cognitive impairment is present among critically ill patients with no abnormalities detected by standard delirium and encephalopathy assessments.