Clarifying the Confusion Surrounding Drug-Associated Delirium in the ICU

Delirium occurs frequently in the critically ill and may adversely affect both short and long-term outcomes.[1, 2] Treatment options for delirium in the ICU remain limited. Therefore, clinicians should focus on delirium prevention and risk reduction strategies.[3] While predisposing (e.g., older age) and many precipitating (e.g. severity of illness) delirium risk factors are not reversible, other precipitating factors, such as patient immobility, aspects of the ICU environment (e.g., noise) and administered medications, may be modifiable.[3] The list of medications that have been reported to cause delirium in the critically ill is long. [4] This is not surprising given the large number of medications administered in the ICU setting, the frequent presence of end-organ dysfunction that may influence the pharmacodynamic response observed, the presence of conditions such as sepsis or stroke that may impair blood-brain barrier integrity, and the fact that use of medications having psychoactive properties that may lead to delirium-like symptoms is frequent. [5] Many of the proposed pathways for drug-induced delirium overlap with those for delirium itself (e.g., anticholinergic activity, gabaminergic activity) although few are confirmed.[7] Ascribing delirium at the ICU bedside solely to the use of a particular medication is potentially fraught with error.[5] The causes for delirium in the ICU are usually multifactorial and seldom obvious, the temporal relationship between medication initiation and delirium onset remains poorly characterized and medications (e.g. benzodiazepines) that may be used to treat delirium-associated agitation may also influence both delirium recognition and duration. The drug-associated delirium literature consists primarily of case series and uncontrolled cohort studies whose design makes it nearly impossible to determine whether a particular medication is an independent risk factor for delirium. One recent systematic review evaluating 80 different potential ICU delirium risk factors was forced to exclude 1,593 (98%) of 1,626 published ICU delirium risk factor studies given that neither a randomized controlled design nor a multivariable approach was used.[6] Moreover, given the fluctuating nature of coma, disease severity, and delirium over the course of the ICU, coupled with a number of important baseline and daily risk factors for its occurrence; it is important to use a competing risk Markov model with time-dependent multinomial methods when evaluating the daily risk from a non-delirium to delirium state. . Data from investigations that incorporate these advanced modelling techniques to evaluate the odds of transitioning on any particular ICU day from an “awake and no delirium state” to “delirium” now help inform delirium-associated risks for benzodiazepines, corticosteroids and medication(s) with anticholinergic effects. [7-11] A landmark 2006 study 198 mechanically ventilated adults found that lorazepam administration was an independent risk factor for a daily transition to delirium [OR = 1.2 (95% CI, 1.1 to 1.4, p = 0.003)] [7]. A more recent analysis of 1,112 critically ill adults found that midazolam administration was an independent risk factor for a daily transition to delirium [OR = 1.04 (95% CI, 1.02 to 1.05), p < 0.001 per every 5mg per day of midazolam administered] [8]. This latter study therefore suggests that for every 5mg of midazolam that is administered to a patient who is awake and without delirium, there is a 4% chance that this patient will have delirium the next day. It is important to note that this risk is additive and thus the administration of 20mg of midazolam in a 24 hour period would be associated with a 16% chance of delirium the next day. The lower odds for transitioning to delirium in this latter study is a reflection of the trend over the past decade to reduce benzodiazepine dosing in an effort to promote patient wakefulness. Given an increased risk for delirium with continuous benzodiazepine use, along with the fact that the risk for delirium is dose-dependent, clinicians should employ strategies known to reduce the daily amount of benzodiazepine administered and convert patients when possible to an intermittent administration regimen. In a patient deemed to require continuous sedation, clinicians should consider non-benzodiazepine sedatives not strongly associated with delirium (e.g., dexmedetomidine, propofol). [3] Despite the uncertain efficacy of corticosteroids for many critical illnesses and their relatively extensive risk profile, corticosteroids are frequently administered in the ICU. Although corticosteroids decrease inflammation, and therefore theoretically reduce neuro-inflammation and the incidence of delirium, results of the (Dexamethasone for Cardiac Surgery) DECS trial suggest that the administration of a corticosteroid prior to cardiac surgery does not influence the prevalence of post-operative delirium.[12] While delirium has long been assumed to be a potential consequence of corticosteroid use, particularly when high doses are administered, its association with delirium has only recently been rigorously evaluated. [9, 10] Using multivariable Markov modeling techniques, one cohort analysis of 520 mechanically ventilated adults with acute lung injury (ALI) found that systemic corticosteroid use was significantly associated with transitioning to delirium from a non-delirious, non-comatose state.[10] However, in a larger analysis of 1,112 patients, who received a corticosteroid on 35% of their ICU days at a median dose of 50 (25-75) mg, corticosteroid administration was not associated with a daily transition to delirium [OR = 1.08, 95% CI, 0.89 -1.32] per each 10mg increase in prednisone equivalent administered. A secondary analysis of the 45% of patients who had severe hypoxemia (and thus likely had ARDS) revealed a risk for delirium that remained unchanged. Differences between these 2 studies in the way in which additional delirium risk factors were identified and incorporated, the daily frequency by which delirium was evaluated, and the differing patient populations likely account for these discordant results. Regardless of the exact risk for delirium with corticosteroid exposure, ICU clinicians should continue to evaluate their patients daily to ensure that they are receiving the lowest effective corticosteroid dose. A number of reports have proposed cholinergic deficiency as being an important mechanistic cause for delirium occurrence. This is not surprising given attention is regulated in part by the cholinergic neurotransmitter system. Prior ICU studies attempting to evaluate the association between anticholinergic medication exposure and delirium occurrence have suffered from important methodological limitations and have yielded inconclusive results. For example, time-dependent multinomial models were not developed and the effect of increased age and an acute inflammatory state, both factors that can lead greater cholinergic neurotransmitter system dysfunction and thus potential over estimate delirium risk, were not considered. In one prospective study of 1, 112 critically ill adults, anticholinergic burden was calculated on a daily basis using the sum of the Anticholinergic Drug Scale score for each medication administered [median (IQR) = 2 (1-3)]. [11] The transition from being in an "awake without delirium" state to "delirium" occurred on 562 of ICU days (6%). Using a first-order Markov model that adjusted for eight covariables, a one-unit increase in the Anticholinergic Drug Scale score resulted in a nonsignificant increase in the probability of delirium occurring the next day (odds ratio, 1.05; 95% CI, 0.99-1.10). Neither age nor the presence of acute systemic inflammation modified this relationship. While medications without strong anticholinergic properties are preferred in the critically ill, the results of this investigation suggest that the association between anticholinergic medication use and delirium in the critically ill may not be as significant as previously thought. The association between medication use and delirium occurrence increasingly has become better investigated. [7-11] Clinicians should develop a basic understanding of the Markov multinomial modelling techniques that are being used in many of these new studies given they have become the gold standard method by which to estimate medication -associated delirium risk in the critically ill. The medication profile in ICU patients should be reviewed daily, and the lowest effective dose for each medication should be administered [5]. In patients with delirium, the medication list should be carefully reviewed to identify medications that could be increasing delirium burden. References 1. Zaal IJ, Slooter AJ: Delirium in critically ill patients: epidemiology, pathophysiology, diagnosis and management. Drugs 2012; 72(11):1457-71 2. Wolters AE, Slooter AJ, van der Kooi AW, van Dijk D: Cognitive impairment after intensive care unit admission: a systematic review. Intensive Care Med 2013; 39(3):376-86 3. Barr J, Fraser GL, Puntillo K, et al. Clinical practice guidelines for the management of pain, agitation and delirium in adult ICU patients. Crit Care Med 2013; 41(1):263-306 4. Accessed August 29 2014. 5. Devlin J.W., Fraser GL, Riker RR. (2011). Drug-induced coma and delirium. In: J. Papadopoulos, B. Cooper, S. Kane-Gill, S. Mallow-Corbett, J. Barletta (Eds). Drug-Induced Complications in the Critically Ill Patient: A Guide for Recognition and Treatment. Society of Critical Care Medicine. (1st ed.). Chicago 6. Zaal I, Devlin JW, Slooter A: A systematic review of the risk factors for delirium in the intensive care unit. Crit Care Med 2013; 41 (Suppl):A558 7. Pandharipande P, Shintani A, Peterson J, et al: Lorazepam is an independent risk factor for transitioning to delirium in intensive care unit patients. Anesthesiology 2006; 104(1):21-6 8. Zaal I, Devlin JW, Hazelbag M, et al: Benzodiazepine associated delirium in critically ill adults. Intensive Care Med 2015 (in press) 9. Schreiber MP, Colantuoni E, Bienvenu OJ, et al.: Corticosteroids and transition to delirium in patients with acute lung injury. Crit Care Med 2014; 42(6): 1480-6. 10. Wolters AE, Veldhuizjzen DS, Zaal IJ, et al. Systemic corticosteroids and transition to delirium in critically ill patients. Crit Care Med 2015 (in press) 11. 11. Wolters AE1, Zaal IJ, Veldhuijzen DS, et al. Anticholinergic medication use and transition to delirium in critically ill patients: a prospective cohort study. Crit Care Med. 2015 Sep;43(9):1846-52. 12. Sauër AM, Slooter AJC, Veldhuijzen DS, et al: Intraoperative dexamethasone and delirium after cardiac surgery: a randomized clinical trial. Anesth Analgesia 2014; 119 (5): 1046-52.



Drug-Associated Delirium in the ICU
The association between drugs and delirium is very complex and difficult to articulate. In addition to all the points highlighted by Dr. Devlin, delirium due to abrupt discontinuation of medications that needs appropriate weaning is another concern. We often focus on opiates and benzodiazepine when we think about appropriate weaning to prevent withdrawal; however there are many other medications with significant withdrawal symptoms if not appropriately weaned. Enhanced sensitivity to medications due changes in blood brain barrier (BBB) during severe illness is another possibility. The impact of each of these phenomena to a given delirious case is very difficult to determine or quantify. For the bedside clinician, the challenge is identifying the most likely culprit/s at the particular point in time abreast of the rapidly changing clinical condition of the patient.
Elizabeth Udeh, PharmD, BCPS
Hartford Hospital
Hartford, Connecticut


The assumption of course is that current delirium screening tools can distinguish between simple benzo pharmacology (altered mental status, inattention, change in level of consciousness) and delirium that carries associated outcomes risks. It is indeed possible that the relationship between benzo's and delirium is largely due to artifact within the assessment tool.


In terms of delirium detection, what do you think about using the EEG as described in an der Kooi, A. W., et al. (2015). "Delirium detection using eeg: What and how to measure." Chest 147(1): 94-101.
BACKGROUND: Despite its frequency and impact, delirium is poorly recognized in postoperative and critically ill patients. EEG is highly sensitive to delirium but, as currently used, it is not diagnostic. To develop an EEG-based tool for delirium detection with a limited number of electrodes, we determined the optimal electrode derivation and EEG characteristic to discriminate delirium from nondelirium.METHODS: Standard EEGs were recorded in 28 patients with delirium and 28 age- and sex-matched patients who had undergone cardiothoracic surgery and were not delirious, as classified by experts using Diagnostic and Statistical Manual of Mental Disorders, 4th edition, criteria. The first minute of artifact-free EEG data with eyes closed as well as with eyes open was selected. For each derivation, six EEG parameters were evaluated. Using Mann-Whitney U tests, all combinations of derivations and parameters were compared between patients with delirium and those without. Corresponding P values, corrected for multiple testing, were ranked.RESULTS: The largest difference between patients with and without delirium and highest area under the receiver operating curve (0.99; 95% CI, 0.97-1.00) was found during the eyes-closed periods of the EEG, using electrode derivation F8-Pz (frontal-parietal) and relative δ power (median [interquartile range (IQR)] for delirium, 0.59 [IQR, 0.47-0.71] and for nondelirium, 0.20 [IQR, 0.17-0.26]; P = .0000000000018). With a cutoff value of 0.37, it resulted in a sensitivity of 100% (95% CI, 100%-100%) and specificity of 96% (95% CI, 88%-100%).CONCLUSIONS: In a homogenous population of nonsedated patients who had undergone cardiothoracic surgery, we observed that relative δ power from an eyes-closed EEG recording with only two electrodes in a frontal-parietal derivation can distinguish among patients who have delirium and those who do not.

My colleague, Dr. Gen Shinozaki and Dr. John Cromwell here at UIHC are embarking on their own study, see link