Postoperative Delirium: A Review of Current Evidence

||||| Like It 0 Like |||||


Suraj Yalamuri, Srinivas Pyati and Charles Brudney*

* Department of Anaesthesiology, Duke University Medical Center and Durham Veterans Affairs Medical Center, Durham, North Carolina, US. (t: +1 919 286 6938 ; e:

Central nervous system (CNS) complications after surgery are a major source of morbidity and mortality. With an average reported incidence of 36.8%, postoperative delirium is a complicated process that increases patient mortality, hospital length of stay, healthcare cost, and the risk of long-term cognitive decline [1, 2]. A recent systematic review indicated that the incidence of delirium varies widely (10–89%), and specific at-risk populations include intensive care unit (ICU) patients (41%) and post-surgical elderly patients (89%) [3]. In the US, the cost of delirium has been estimated at between $38 billion and $152 billion [4]. Therefore, understanding the pathophysiology of postoperative delirium and identifying treatment strategies has the potential to improve patient outcomes significantly, whilst reducing resource utilisation and healthcare cost.

Clinical Presentation and Diagnosis
The Diagnostic and Statistical Manual of Psychiatric Disorders (DMS-IV) defines delirium as a clinical diagnosis characterised by an altered state of consciousness, change in cognition, and an acute, fluctuating time course. There are three main subtypes: hyperactive (25%), hypoactive (50%), and mixed (25%). The hypoactive subtype is routinely missed and is associated with a higher mortality rate [5]. The majority of studies define delirium as ‘postoperative’ if it occurs 24–72 hours after surgery [1]. It is important to distinguish postoperative delirium from similar clinical entities such as emergence delirium, postoperative cognitive dysfunction, and dementia. Emergence delirium is an acute reaction to emergence from anaesthesia, found primarily in children and young males, and can be seen in 5–21% of cases [4]. Postoperative cognitive dysfunction is defined by the International Society of Postoperative Cognitive Dysfunction as “an impairment in one of several cognitive domains such as attention, memory, sensorimotor speed, and cognitive flexibility” [6]. This condition typically develops over weeks to months.

While there is no standardised definition for the time course of postoperative delirium, the majority of studies define it as occurring between 24 and 72 hours after surgery, though some follow patients until hospital discharge. The Confusion Assessment Method (CAM) is the most commonly-used diagnostic method, and has a reported sensitivity of 94% and specificity of 89% [1, 7]. In the 37 identified risk-prediction models for postoperative delirium, the predictors most often used include age (20 models), preoperative Mini-Mental State Examination score (10 models), and preoperative increased alcohol use (7 models) [7]. There is no direct evidence that current pharmacological and non-pharmacological treatments improve patient outcomes for those at higher risk of delirium, as predicted by risk-prediction models [3]. Additionally, the lack of standardisation in the frequency of screening for delirium also accounts for the variability in outcomes [8]. Nevertheless, early diagnosis that includes the recognition and treatment of modifiable factors with current therapies helps patients recover from postoperative delirium [9].

The multiple risk factors for postoperative delirium can be categorised according to the timing of the perioperative period [9]. Preoperatively, three factors – age >70 years; malnutrition (low serum albumin and dehydration); and alcohol and drug abuse – all independently carried an odds ratio (OR) of 3.3 of developing postoperative delirium. Pre-existing cognitive impairment, such as dementia and psychiatric disorders, had an OR of 4.2. Other risk factors were decreased functional status (OR: 2.5), electrolyte abnormalities (OR: 3.4) [10], and type of surgery; orthopaedic, abdominal aortic aneurysm repair, and cardiothoracic procedures are associated with an increased OR as high as 8.3 [9]. Intraoperative risk factors include hypothermia, hypotension and hypoxia [9]. The risk of developing postoperative delirium is 40%, compared with 19% if excess depth of anaesthesia is avoided, as measured by bispectral index (BIS) scores [12-15]. Postoperatively, the greatest risk factors are persistent hypercarbia and hypoxia (17.2% increase in incidence of postoperative delirium) and an increased hospital length of stay (14% increase in incidence of postoperative delirium) [9]. Benzodiazepine use is pervasive in the perioperative period and its use in the elderly population significantly contributes to delirium. Each additional 1 mg of midazolam carries with it a 7% risk of developing delirium [16].

The molecular basis of postoperative delirium is an area of ongoing research. Most of the models have a similar theme: a susceptible patient who endures an acute neurological insult that leads to alterations in the levels of neurotransmitters or inflammatory mediators which, in turn, leads to delirium. Several markers are implicated, the most common being gamma aminobutyric acid (GABA), dopamine, acetylcholine, inflammatory makers such as c-reactive protein (CRP), cortisol, and many interleukins (IL-6, IL-8, IL-10) [1]. In the Maximizing Efficacy of Targeted Sedation and Reducing Neurological Dysfunction (MENDS) trial, McGrane et al. found a trend toward higher CRP in 87 post-surgical and ICU patients who developed postoperative delirium [17]. Patients who develop delirium after undergoing coronary artery bypass grafting have been found to have elevated cortisol levels and higher IL-6 levels [18]. The same patient population also showed a correlation between lower intraoperative BIS scores and higher cortisol levels. Cortisol has also been found to mediate cognitive function through the glucocorticoid receptors in the hippocampus and frontal cortex in Alzheimer disease patients [19].

Another area of interest is the association between pre-existing dementia and the risk of postoperative delirium. Patients with the epsilon 4 allele of apolipoprotein E (apoE4) are thought to be at higher risk of developing delirium [20], as it is also a risk factor for the development of Alzheimer’s disease [19]. However, the evidence for this is equivocal: whilst a study of 230 patients in the post-anaesthesia care unit showed no association between those with apoE4 and postoperative delirium [21], Leung and van Munster have suggested that apoE4 is nonetheless a risk factor [22]. Comprehensively, the studies on delirium show the process to be multifactorial. While the definitive pathways of delirium are still being explored, there are proven treatment strategies to both treat and minimise patient risk.

Treatment and Prevention
Delirium has a complex pathophysiology, so its treatment should also be multimodal (Figure 1). Moreover, postoperative delirium typically presents in the context of other acute medical problems that also require treatment. Therefore, the management of delirium demands thoughtful consideration and effective use of both pharmacological and non-pharmacological treatments. A recent systematic review and meta-analysis by Zhang et al. concluded there are three beneficial interventions, as follows: (1) In patients requiring sedation, dexmedetomidine produces less delirium, in comparison to propofol, midazolam or morphine; (2) both typical and atypical antipsychotics decrease the occurrence of delirium, compared with placebo; (3) multicomponent interventions (geriatrics service consultation, patient-specific non-pharmacological interventions, and targeted education for staff who manage patients with delirium) are effective in preventing delirium [2].

Figure 1: Summary of multimodal interventions for treating postoperative delirium

Brudney pic

Continue to the next page…

Download the full article here:

Return to the main table of contents by clicking here

Leave a Reply

Scroll to top