Perioperative complications involving the central nervous system (CNS) are common. However, these issues have generally received little attention in the presurgical assessment process. New data allow us to better guide physicians and patients regarding perioperative CNS risk and ways to mitigate that risk ( Fig. 7.1 ).
Postoperative delirium (POD) is an acute confusional syndrome that is frequently encountered in the postoperative setting. In fact, it is the most common surgical complication among older adults, with an incidence ranging from 15% to 50% in this population, depending on the procedure. Delirium can be defined as an acute, fluctuating, multifactorial disorder characterized by a decline of attention, cognition, and awareness, and it is a distinct disorder that is separate from dementia or chronic cognitive decline. Three subtypes of delirium exist: hyperactive, hypoactive, and mixed. Despite hypoactive being the most frequent subtype, patients who have hypoactive delirium are less frequently diagnosed and have poorer prognoses than those with hyperactive delirium. Unfortunately, a diagnosis of delirium carries many implications for a patient because those diagnosed with delirium have increased morbidity, mortality, loss of independence, likelihood of subsequent institutionalization, and health-care costs.
Establishing a diagnosis of delirium requires that the patient’s baseline mental status be defined with a preoperative screen or by a knowledgeable informant. Once this is established, a cognitive screening test may be used to determine what, if any, change from baseline the patient is experiencing. There are many different diagnostic tools and severity scales for delirium, with two of the most common being the Confusion Assessment Method (CAM) and CAM-ICU, which is a derivative of the CAM that is intended for mechanically ventilated patients in the intensive care unit (ICU) ( Fig. 7.2 ).
The pathophysiologic mechanisms underlying delirium are poorly understood and may include neurotransmitter imbalance and/or neuroinflammation. Despite the lack of a definite cause of delirium, multiple predisposing and precipitating factors for delirium have been elucidated ( Table 7.1 ). The summation of predisposing and precipitating factors will determine a patient’s likelihood of developing delirium; a higher burden of predisposing factors will necessitate a lower number of precipitating factors. Multiple drugs have been shown to increase the probability of delirium in the geriatric population and should be avoided in the perioperative period if possible. These include benzodiazepines, anticholinergics, diphenhydramine, hydroxyzine, meperidine, histamine-2 antagonists, and sedative hypnotics.
|Factors known to predispose to delirium||Factors known to precipitate delirium|
|Older age |
High burden of coexisting disease
Mild cognitive impairment
|Drugs (benzodiazepines, anticholinergics, meperidine, opioids, diphenhydramine, hydroxyzine, histamine-2 antagonists, sedative hypnotics) |
Acute exacerbation of chronic illness
Currently, optimal intraoperative management of patients at risk for delirium is not yet well defined. The only intraoperative delirium-reducing interventions for which strong recommendations exist are avoidance of predisposing drugs and adequate pain control, with nonopioid medications and regional anesthesia being a consideration when possible or appropriate. However, several interventions are currently being investigated. Some studies have shown that deeper levels of anesthesia are associated with increased incidence of POD and that bispectral index (BIS)-guided anesthetics are associated with a lower incidence of POD. It has been hypothesized that BIS-guided anesthetics may have fewer periods of electroencephalogram (EEG) burst suppression. However, the increased burden of burst suppression and other EEG patterns identified in patients with POD may simply be a marker of fragility and not a causative factor. The largest randomized trial to date, the ENGAGES [Electroencephalography Guidance of Anesthesia] trial, randomized over 1200 patients to EEG-guided anesthesia or control and found no difference in the incidence of POD. Other studies are ongoing.
Another intervention that has been investigated in recent years is the use of dexmedetomidine. Recent studies have shown that dexmedetomidine administered perioperatively as a low-dose prophylactic infusion may decrease not only POD but also the rates of stroke and both operative and in-hospital mortality. These findings may be explained in part by evidence that dexmedetomidine acts via the JAK2/STAT3 pathway to attenuate isoflurane-induced neurocognitive deficits in senile mice. However, other studies have been unable to replicate these findings and have found that dexmedetomidine does not decrease POD. A recent report found that low-dose nocturnal dexmedetomidine reduced POD in the ICU, but this awaits validation in larger trials. As the data are conflicting at this time, it would be reasonable for the anesthesiologist to consider the perioperative use of dexmedetomidine in the at-risk patient if it is considered clinically feasible/appropriate. It remains unclear whether intraoperative, postoperative, or combined dosing is most effective.
Much like dexmedetomidine, ketamine showed initial promise as an intraoperative intervention to prevent POD. However, a recent study was unable to replicate this result, and patients receiving ketamine experienced higher incidences of hallucinations and nightmares (but not delirium). Although ketamine does show benefit in decreasing delirium in children, there are still relatively few studies examining ketamine in adult delirium prevention. Ketamine may allow lower perioperative opioid doses to be used, and the anesthesiologist may still consider ketamine use in at-risk patients as part of a multimodal anesthetic.
Finally, both cholinesterase inhibitors such as rivastigmine and antipsychotics such as haloperidol administered perioperatively in a preventative capacity have been investigated. Cholinesterase inhibitors have not been found to decrease delirium incidence and furthermore may be associated with increased adverse effects, including mortality. It is not recommended that patients not taking preoperative cholinesterase inhibitors be started on one in the perioperative setting. Data on prophylactic perioperative use of antipsychotics are less conclusive but generally do not favor their use, owing to mixed results on delirium prevention, no decrease in mortality, and a high rate of adverse effects.
Emergence delirium is an acute, short-lived phenomenon that occurs in the early/immediate postoperative period after the patient has awakened. These patients experience a short-term impairment in consciousness that may manifest as disorientation, hallucinations, confusion, restlessness, and hyperactive behavior that may be violent or harmful to the patient or hospital staff. The incidence is reported to range from 3% to 21%, and these patients experience longer recovery periods, use more resources, and pose a greater risk for harm to themselves and hospital staff.
Emergence delirium is not as well investigated in the adult population as it is in the pediatric population. However, some risk factors that have been identified for adults include preoperative benzodiazepine administration, post-traumatic stress disorder, breast surgery, abdominal surgery, long duration of surgery, young age, presence of an endotracheal tube and/or urinary catheter, sevoflurane anesthesia, recent smoking, and postoperative pain ≥ 5 on the numerical rating scale.
Treatment of emergence delirium is also not well studied, but there are some case reports and case series describing the effective use of dexmedetomidine, both as an intraoperative infusion in patients with a history of emergence delirium and via intravenous bolus dosing as rescue therapy for patients after extubation. It is therefore reasonable to consider use of dexmedetomidine in patients who have a history of emergence delirium, who have risk factors for emergence delirium, or who experience significant delirium upon awakening from their anesthetic.
Treatment of Postoperative Delirium After the Immediate Recovery Period
As up to 40% of delirium has been shown to be preventable, the best treatment for delirium is the prevention of its occurrence altogether. Currently recommended preventative strategies include cognitive reorientation, sleep enhancement with nonpharmacologic sleep protocols and hygiene, adaptations for visual and hearing impairment, early mobility, nutrition and fluid repletion, avoidance of precipitating medications, adequate oxygenation, prevention of constipation, pain management, education of health-care professionals, and use of an interdisciplinary team.
If a patient does develop delirium, the aforementioned nonpharmacologic therapies should be implemented as first-line treatment if not already in place. In addition, the patient should be evaluated for the presence of any precipitating factors (such as a urinary tract infection or poorly controlled pain) that warrant intervention. In the event that the above interventions are ineffective and the patient is experiencing hyperactive delirium that is posing a threat to him-/herself and/or hospital staff, antipsychotics may be used. It is recommended that the lowest possible dose be used for the shortest possible time. Benzodiazepines are not recommended unless medically indicated, such as in the instance of alcohol or benzodiazepine withdrawal.
Identify patients with predisposing and/or precipitating risk factors for POD.
In patients at risk for POD, consider regional anesthesia, multimodal analgesia, and the use of dexmedetomidine.
EEG-guided anesthesia does not prevent POD but may be desirable in allowing optimization of anesthetic dosing.
Avoid precipitating medications, such as preoperative benzodiazepines, when possible in high-risk patients.
Implement multidisciplinary, nonpharmacologic measures, such as geriatrics consultation, to both prevent and treat POD.
Use low-dose, short-course antipsychotics or dexmedetomidine for delirious patients at risk of harm to themselves or others. Reserve benzodiazepines for patients withdrawing from alcohol or benzodiazepines.
Postoperative Cognitive Dysfunction
Postoperative cognitive dysfunction (POCD) describes a syndrome of prolonged impairment or deterioration of cognitive function with onset usually occurring weeks to months after surgery. Assessment of decline is made with preoperative and postoperative testing. Features include limitations in memory, intellectual ability, and executive function. A standardized approach to the identification and study of POCD has been limited by the lack of consensus diagnostic criteria. In general, the diagnosis involves identifying a worsening in postoperative formal neurocognitive testing relative to preoperative levels and is hence difficult outside of a research protocol. The International Study of Postoperative Cognitive Dysfunction (ISPOCD) made the diagnosis of POCD based on a combined Z-score > 2 (i.e., 2 standard deviations from the mean) across a multitude of individual cognitive tests or at least two Z-scores > 2 for single test parameters. Other investigators have looked at changes in cognitive domains such as memory and executive function, based on similar individual neurocognitive tests, relative to the patient’s baseline or to a control group.
Although a consensus definition remains elusive, most investigators agree the morbidity of cognitive decline is significant. POCD is associated with longer hospital stays and cost, premature withdrawal from the workforce, and greater 1-year mortality, and it may possibly cause a change in the trajectory toward dementia.
Attention to preoperative recognition of patients at risk for POCD has increased in recent years in the hope that identification, preoperative counseling, risk factor modification, and possible interventional strategies can help mitigate the untoward effects of POCD on individual patients and health-care systems as a whole.
Since 1955, with the publication of “Adverse Cerebral Effects of Anaesthesia on Old People” in The Lancet , clinicians have observed cognitive decline after surgery that can persist for months or even years. These early observations inspired research into the phenomenon of cognitive decline and the frequency with which it occurs. Current studies report an incidence ranging between 10% and 80% due to different patient populations and approaches to measuring POCD.
Initial investigations focused on cognitive impairment after cardiac surgery. A higher prevalence of POCD was originally described following coronary artery bypass grafting surgery than after noncardiac surgery, with rates as high as 24% at 6 months and 42% at 5 years. More recent work suggests underlying coronary artery disease and associated risk factors may contribute to cognitive decline more than cardiac surgery and anesthesia; ongoing research is necessary to dissect out the contribution of each of these risk factors.
It soon became clear that POCD was present after noncardiac surgery as well as after cardiac surgery. The ISPOCD was a series of multicenter, prospective, case control studies on POCD in noncardiac surgery. The ISPOCD1 study (mean age, 68 years) found POCD in 25% of patients at 1 week, in 10% at 3 months, and in 1% at 1 year. Subsequent work looking at young, middle-aged, and elderly groups found an increased rate of POCD (12.7%) in the older age group (mean age, 70 years) at 3 months, compared with 5.7% in the young and 5.6% in the middle-aged group. In fact, those with POCD at both discharge and 3 months were five times more likely than those without POCD to die in the first year after surgery.
Risk factors for POCD are comprised of both modifiable (extrinsic) and nonmodifiable (intrinsic) factors ( Table 7.2 ). It is important to screen for intrinsic risk factors, even though these variables are difficult or impossible to modify, because this may impact decisions on purely elective surgery. Intrinsic risk factors include increasing age, preexisting cognitive impairment, lower education level, and comorbid conditions. Extrinsic factors, such as hemodynamic instability, metabolic derangements, or pain control, can more easily be addressed and optimized in the perioperative period in an effort to reduce the incidence of POCD.
|Advanced age |
Fewer years of education
Preexisting cognitive impairment
Declined functional status
|Major surgery |
Increased duration of anesthesia
Uncontrolled postoperative pain
Postoperative respiratory complications
More definitively linked to POD, the contribution of intraoperative hypotension and hypoxia to POCD is less clear. Early studies such as the ISPOCD1 study found that mild, brief hypoxemia and hypotension were independent risk factors for the development of POCD. Newer evidence indicates low blood pressure may accelerate cognitive decline in elderly patients. More research is needed in the perioperative arena.
Although there is consensus that testing for preoperative cognitive dysfunction should be performed, there is not consensus on how we should be testing or what tool we should use. A standard, validated tool that is both thorough and practical has not yet been developed. However, POCD is hard to diagnose and quantify without documentation of baseline status.
The 2012 American College of Surgeons National Surgical Quality Improvement Program/American Geriatrics Society Best Practice Guidelines for Optimal Preoperative Assessment of the Geriatric Surgical Patient recommend early cognitive testing of any patient older than age 65, which should include an interview with spouse or family members to supplement evaluation. Also recommended is referral to a geriatrician for perioperative management, should cognitive dysfunction be present. These guidelines recommend use of the Mini-Cog tool (which consists of a three-item recall test and clock drawing exercise); however, other tools exist throughout the POCD literature ( Table 7.3 ).
(Montreal Cognitive Assessment)
(Mini Mental State Examination)
(Saint Louis University Mental Status Exam)