DOI: 10.1097/CCM.0b013e31827c02f0

,  

,  

PMID: 23528763

Issn Print: 0090-3493

Publication Date: 2013/04/01


Print

Dissecting Sedation-Induced Delirium*

Gilles L. Fraser; Craig P. Worby; Richard R. Riker

    loading  Checking for direct PDF access through Ovid

Excerpt

Understanding more about delirium will help focus research and practice efforts and potentially lead to improvements in associated short and long-term clinical consequences. Experts are careful not to ascribe causality to the increased morbidity, mortality, costs, and long-term cognitive impairment observed in patients with delirium because it is not known if it is a marker of acute illness or if it directly contributes to morbidity beyond what is independently caused by its precipitants (1). This distinction is crucial. If delirium has a direct link to serious illness, successful strategies for its prevention and management will likely affect important clinical outcomes. One such strategy, the avoidance of benzodiazepines in ICU patients at risk for delirium received a weak recommendation in recent pain, agitation, and delirium-management guidelines (2).
It is vital to establish whether benzodiazepine administration represents a modifiable risk factor for delirium and to determine what effect, if any, this has on important clinical outcomes. The question is whether the neurologic manifestations of the benzodiazepines, as measured by the Confusion Assessment Method for the ICU (CAM-ICU), truly represent delirium, and whether this portends the same dire consequences as delirium from other causes.
In this issue of Critical Care Medicine, Skrobik et al (3) present data that expand clinical equipoise on this subject and encourage debate and discussion of the assumptions that have framed the characterization of benzodiazepines as cognitively harmful to ICU patients (Table 1).
The CAM-ICU and the Intensive Care Delirium Screening Checklist are delirium screening tools validated using the American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, criteria, which were developed from experience in non-ICU patients (4). It is unclear how transferable these criteria are to ICU patients, an extremely heterogeneous group who commonly are nonverbal, suffer from multiple organ dysfunction, and receive sustained sedative and opioid therapy. These characteristics may explain differences noted between the delirium described for ICU vs. non-ICU patients, including different risk factors (age may not be a risk factor for ICU delirium), distribution of motoric subtypes (hypoactive delirium dominates in the ICU) (5), and its very high prevalence (up to 80%) (6). These unique features have prompted some to question whether a diagnostic gold standard really exists for ICU delirium (4, 7).
No clinical monitoring instrument is perfect. We accept errors and weaknesses when using pulmonary artery catheters, pulse oximeters, electrocardiograms, and biomarker measurements because most experienced users can distinguish artifact from true data. Our appreciation for potential “artifact” from ICU delirium diagnostic tools is just emerging (4). Many of the components of the CAM-ICU (altered consciousness, inattention, and disorganized thinking) are affected by benzodiazepine use, potentially leading to an overdiagnosis of delirium (8). Indeed, what is identified as hypoactive delirium may often be the consequence of expected pharmacology rather than a separate primary phenomenon (9).
Determining the importance and relevance of drug-induced altered consciousness is challenging at best. Consciousness was found to be the least accurate item during validity testing of the Intensive Care Delirium Screening Checklist (10). The impact of this feature has not been well studied within the CAM-ICU (11). The potential magnitude for artifact is illustrated by data demonstrating a 48% increase in delirium identification during sedation administration when compared with assessments made in the same patients after sedation was lightened to the point of wakefulness (12). An early trial hinted at a similar phenomenon (13).
Benzodiazepine therapy is often linked with the established poor outcomes associated with delirium (14). The assumption is that delirium has consistent features across wide-ranging patient populations and has similar dire consequences independent of its etiology.

Related Topics

    loading  Loading Related Articles