Antimicrobial exposure and the risk of delirium in critically ill patients

• Published in: Critical care (London, England) Vol. 22; no. 1; p. 337
• Main Authors: Grahl, Jessica J, Stollings, Joanna L, Rakhit, Shayan, Person, Anna K, Wang, Li, Thompson, Jennifer L, Pandharipande, Pratik P, Ely, E Wesley, Patel, Mayur B
• Format: Journal Article
• Language: English
• Published: England BioMed Central Ltd 12-12-2018; BioMed Central; BMC
• Subjects: Adult; Aged; Anti-infective agents; Anti-Infective Agents - adverse effects; Anti-Infective Agents - therapeutic use; Antibiotics; Cohort Studies; Critical care; Critical Illness - rehabilitation; Critical Illness - therapy; Critically ill persons; Cross-Sectional Studies; Delirium; Delirium - etiology; Encephalopathy; Female; Health aspects; Humans; Intensive Care Units - organization & administration; Intensive Care Units - statistics & numerical data; Length of Stay - statistics & numerical data; Male; Middle Aged; Multivariate Analysis; Neurotoxicity; Retrospective Studies; Risk factors; Delirium; Critical care; Neurotoxicity; Antibiotics; Encephalopathy
• ISSN: 1364-8535; 1466-609X; 1364-8535
• DOI: 10.1186/s13054-018-2262-z

Prior retrospective cross-sectional work has associated antimicrobials with a non-specific phrase: encephalopathy without seizures. The purpose of this study is to determine whether different classes of antimicrobials have differential associations with the daily risk of delirium after critical illness is adjusted for.
Our study was a nested cohort that enrolled non-neurological critically ill adults from a medical or surgical intensive care unit (ICU) with daily follow-up to 30 days. Our independent variable was exposure to previous-day antimicrobial class: beta-lactams (subclasses: penicillins, first- to third-generation cephalosporins, fourth-generation cephalosporins, and carbapenems), macrolides, fluoroquinolones, and other. We adjusted for baseline covariates (age, comorbidities, cognition scores, sepsis, and mechanical ventilation), previous-day covariates (delirium, doses of analgesics/sedatives, and antipsychotic use), and same-day covariates (illness severity). Our primary outcome of delirium was measured by using the Confusion Assessment Method for the ICU. A daily delirium logistic regression model was used with an ICU time-restricted sensitivity analysis including daily adjustment for sepsis and mechanical ventilation.
Of 418 ICU patients, delirium occurred in 308 (74%) with a median of 3 days (interquartile range 2-6) among those affected and 318 (76%) were exposed to antimicrobials. When covariates and ICU type were adjusted for, only first- to third-generation cephalosporins were associated with delirium (logistic regression model odds ratio (OR) = 2.2, 95% confidence interval (CI) 1.28-3.79, P = 0.004; sensitivity analysis OR = 2.13, 95% CI 1.10-4.10, P = 0.024).
First-, second-, and third-generation cephalosporins doubled the odds of delirium after baseline co-morbidities, ICU type, the course of critical care, and other competing antimicrobial and psychotropic medication risks were adjusted for. We did not find an association between delirium and cefepime, penicillins, carbapenems, fluoroquinolones, or macrolides.