#30100: Understanding Postoperative Cognitive Dysfunction

Propofol is nearly universally used for both the induction and maintenance of general anesthesia [13,17]. Induction is the process whereby an intravenous bolus dose of propofol quickly renders the patient unconscious. A period of less than 30 seconds from injection to unresponsiveness is not uncommon. Propofol acts on a specific binding site on gamma-amino butyric acid proteins (i.e., type A or GABAA) on the surface of neurons in the brain [18]. These proteins act as channels that allow the passage of chloride ions from the plasma into the interior of the neuron. As chloride ions are negatively charged (anions), they lower the membrane potential inside the cell. This makes the cell less likely to generate an action potential, and less likely to send impulses to other brain cells. When enough neuronal transmissions are inhibited, classical brain theory states that unconsciousness occurs [16]. Note that this inhibition occurs along a continuum, where smaller doses of the agent lead not to unconsciousness, but sedation (response to a voice while feeling unconcerned). This continuous spectrum of effects impacts executive function. In small doses, executive function is diminished, but still in effect. For example, a patient may receive a local anesthetic technique resulting in the numbing of an arm or leg. In order to diminish the anxiety associated with such a technique, patients may receive a continuous intravenous infusion of a low dose of propofol [17]. They will still respond to questions, take deep breaths when requested, and move other unblocked extremities in response to a verbal command. Propofol wears off in a fashion that inversely mirrors its onset; patients awaken extremely rapidly when the agent is discontinued [13]. Further, they have no hangover effect, such as that seen with barbiturate agents; indeed, they seem to have a feeling of well-being upon awakening, and some have reported feelings of euphoria [19].

Ketamine is a dissociative anesthetic with a different mechanism of action than propofol. While binding with numerous receptors in the brain, its primary effect derives from binding with and blocking the N-methyl d-aspartate (NMDA) receptor [17]. When activated, these receptors excite the brain by allowing the entry of positive ions (or cations) into neurons, thus moving the resting membrane potential of the neurons closer to threshold, making it easier for the brain to become excited and aroused. Ketamine affects neurons in both the brain and the spinal cord, making both an anesthetic as well as a potent analgesic. Sufficient doses of the agent completely abolish executive function, and even small doses result in altering inhibition and cognitive ability. Indeed, patients receiving ketamine demonstrate markedly reduced executive function, which may continue after emergence from anesthesia. Calvey and Williams have described ketamine's emergence phenomena as such [13]:

Almost counterintuitively, however, is the use of small doses of ketamine being administered to patients with neuropsychiatric disorders, particularly treatment-resistant depression and post-traumatic stress disorder [20,21]. The degree to which this agent changes executive function places it in an area of interest for those studying POCD. At this time, there is no research that has been done to determine the extent to which ketamine helps or hurts the development of POCD.

At (or close to) the end of the surgical procedure, the anesthesia provider turns off the device administering the anesthetic (a pump for intravenous infusions, or the vaporizer for the inhaled agents). It is important to ensure the patient has adequate analgesia, as the inhibition of painful stimuli by the inhaled or IV agents will soon stop. As patient begins to recover, depth and regularity of respirations increases, and the patient may begin to move his or her extremities. In each case, patients pass through what is referred to as stage 2 (or the "delirium stage") of anesthesia, first defined in a paper by Arthur Guedel in 1936 [30]. During stage 2, patients may experience a transient delirium during which they will not obey instructions or respond to verbal stimuli and may thrash about the operating bed. Inadequate analgesia makes this much worse, while the use of benzodiazepines and analgesics may ameliorate it. Usually, this is a very transient period, lasting two to four minutes, after which the patient has recovered sufficiently to be transported safely to the postanesthesia care unit. Theoretically, this is attributed to the late offset of anesthetics to inhibitory neurons, as is reflected by irregular respiratory status, dilated pupils, and divergent gaze. This is normal if the patient spontaneously and quickly passes through this stage. Some patients, particularly those in whom surgery is less complex and/or who have received opioids, seem to effortlessly pass through this stage and awaken with minimal disruption. In other cases, the term stormy emergence seems inadequate to describe the several minutes of complete disinhibition that can occur as the patient passes through this stage. Although risk factors for stormy emergence have not been clearly identified, there have been studies linking hypochondriasis to POCD [28]. A long-standing truism in anesthesia is: "Go to sleep scared, wake up scared." This is not a new discovery, having been first reported in the literature in the 1960s [31]. Indeed, it is incumbent upon the anesthetist to reassure the patient during the preoperative evaluation and to administer antianxiety agents when indicated. This is done to decrease the likelihood (or at least the severity) of going to sleep "scared." In either case, smooth emergence or stormy emergence, the normal course shows an increasing degree of executive function from the time the anesthetic is discontinued until the patient is discharged from the facility.

Delirium is a state in which the patient's mental functions are profoundly degraded and executive function nearly completely absent. One of the earliest studies of postoperative delirium characterized the condition as, "an acute brain syndrome characterized by impairment of orientation, memory, intellectual function, and judgment with lability of affect" [31]. This study matched 57 patients not experiencing delirium with 60 who had. The researchers found the principle pathologic states associated with delirium included metabolic imbalance, excessive surgical stress, cardiac failure, infection, intoxication, pre-existing brain disease, and anemia [31]. In the more than 50 years that followed, knowledge about delirium has grown, though not as exponentially as might be imagined.

The types of hyperactive behavior the average clinician would associate with delirium actually account for only 20% of cases [32]. Most delirium is hypoactive in nature, which is far more difficult to observe in the patient recovering from anesthesia; the signs include disorientation, memory problems, and disorganized thought, which can be mimicked by the sedated postsurgical patients [32]. Delirium may occur in 5% to 10% of postoperative patients [33]. Delirium may develop from the preoperative self-administration of drugs or alcohol and thus may be seen after emergency surgery in patients who have used these substances. Other causes of delirium include tumor, psychiatric disturbance, or other pre-existing neuropathology [33,34]. Delirium is thus differentiated from POCD by the presence of either an inherent disease process or extrinsic substance abuse. Ordinarily, in the case of those with substance use disorder, as the drug effects wear off, so does the delirium. Some have linked delirium to hypoxia or perioperative anemia, though this remains under investigation. There is also an age component involved in patients developing delirium, as will be discussed in more detail later in this course [32].

It is important not to confuse dementia, a diagnosable and named disorder, with POCD. In extremes, patients with POCD may experience dementia, but for the purposes of this course, dementia refers to an organic condition of cognitive impairment the patient brings to the operating room before any planned or emergent procedure is performed. Age-related cognitive decline is an expected aspect of the aging process. The likelihood of surgery increases with age, and one study reported that 55% of all surgeries in the United States are performed on patients 65 years of age or older [35]. Delirium rates also increase in a stepwise fashion as patients age [35]. Patients presenting with pre-existing diagnoses, such as Alzheimer disease, organic brain syndromes, or senility, are at significantly increased risk for poor postoperative cognitive outcomes [36]. In one study of older adult patients being screened for elective surgery, 44% showed a preoperative baseline measurement consistent with impaired cognition [36]. Patients with cognitive deficits also developed delirium at higher rates, and their hazard for death was 2.77 times that for patients who had impaired cognition but did not develop delirium [36]. Numerous studies have reported that patients presenting with some form of cognitive impairment preoperatively are far more likely to experience postoperative cognitive effects that are both more severe and longer lasting than those without pre-existing impairment [29,34,35,36,37]. The clinician should, therefore, be observant of patients' existing cognitive status during preoperative evaluation and seek input from family members who also may be present to determine the degree of preoperative dysfunction.

Neurologic insults may occur during anesthesia and surgery. While their incidence grows smaller each year as advances in both anesthetic agent safety and monitoring modalities occur, patients may rarely experience some form of pathologic condition during the conduct of anesthesia. Such insults include stroke, central nervous system hemorrhage, and/or embolism to the vasculature of the brain. Stroke during surgery is quite rare, with rate of occurrence ranging from 0.08% to 0.7% [38,39]. Stroke incidence also varies by the type of surgery involved. The rates vary quite widely, and operations concerning any aspect of the vascular tree, as may be expected, have far higher rates. It is important to note that the incidence of perioperative stroke increases, on average and across types of surgery, approximately 0.15% for each decade of life over the age of 65 [40]. It is extremely difficult to determine the presence of brain ischemia during general anesthesia, as the paralyzed, anesthetized and intubated patient has nearly all the signs and symptoms of stroke effectively hidden from view. In some cases, types of electroencephalography are used to monitor brain function, but this is quite rare. Ordinarily diagnosis of ischemic stroke is made in the post anesthesia recovery room, further adding to the long list of postoperative central nervous system dysfunctions which may occur.

As noted, cardiovascular procedures have been extensively studied for their relationship to POCD. A significant percentage of patients undergoing cardiovascular or extensive vascular procedures manifest POCD, as has been observed almost since the genesis of open-heart surgery. As early as 1972, altered states of postoperative consciousness ranging from mild impairment to major delusional and hallucinatory activity after open heart surgery were being reported [42]. In a 1972 study, 43 of 86 patients (52%) experienced some form of cognitive impairment in the intensive care unit following open-heart surgery [42]. Studies such as this provided the impetus for further research. The road to finding out the nature of POCD after cardiac surgery has not, however, been an easy one. In one systematic review of 420 studies related solely to POCD after cardiac surgery, the researchers found little consensus on the incidence, severity, and time course of symptoms [43]. This is primarily because no one neuropsychologic test covers all areas of executive function, and the sheer volume and time of testing that would be required is quite daunting.

In a systematic review of 62 studies, researchers analyzed POCD prevalence among a broad spectrum of surgical patients [44]. The presence of POCD was measured at various time periods postoperatively. Researchers found a median incidence rate of 48% in the first 21 postoperative days, 24% from day 22 to 5 months, 24% from 6 months to 1 year, and 42% after 1 year [44]. They cautioned that this failed to account for atherosclerotic or other organic postoperative changes, but the rates of POCD remain significant. Another systematic review supported these findings. Patel and associates found the percentage of patients experiencing some degree of cognitive dysfunction after cardiac surgery to be high, but with a wide degree of variability [43]. Significantly, 10% to 20% of patients studied were still experiencing dysfunction after one year, and after five years, the incidence increased to approximately 30% [43]. There should be some care taken with the longer-term data, as the evidence for diminished executive function due solely to age is quite robust, and the changes may unrelated to surgery or anesthesia.

While systematic reviews are helpful to our understanding, prospective studies are also of value. One such study of 101 patients 45 to 75 years of age assessed the extent to which preoperative comorbidities and perioperative management might decrease the rate of POCD [45]. The research team administered the Wechsler Memory Test (WMT) preoperatively, three to five days postoperatively, and then once more five months postoperatively. While not the broadside of tests required for complete executive function analysis, the researchers nonetheless used their postoperative WMT score as a proxy measure for POCD. The findings provided some information about why people develop POCD. First, the researchers found a significant correlation between the preoperative WMT score and the postoperative score; in other words, 59% of the memory test score variation could be attributed to the memory ability patients demonstrated prior to surgery [45]. Patients who had pre-existing cognitive issues were likely to continue to display impairment after surgery. Another noncardiac contributing factor identified was the presence of chronic obstructive pulmonary disease (COPD), which correlated negatively with WMT scores (i.e., those who had COPD before surgery experienced diminished cognitive ability after surgery) [45]. There are numerous theories about the mechanisms of action leading to POCD, and this study supports the concept that existing hypoxia (or perhaps hypercarbia) may result in decreased cognitive functioning, as will be discussed in greater detail later in this course. In terms of actual relationship of cardiovascular surgery and POCD, the only factor this research team found to be statistically significant was a correlation with aortic cross-clamp time and postoperative WMT score. As cross-clamp time increased, memory ability after surgery decreased; the correlation was statistically significant, but weak [45].

Another team of researchers tested the hypothesis that cerebral regional oxygen saturation (rSO₂) deficits (i.e., hypoxia in the brain) resulted in POCD [46]. They tested 57 patients by performing complete neurocognitive test batteries on the day before surgery, and four to seven days after (dysfunctions in this group were described as early POCD), as well as one month after surgery (dysfunctions in this group were described as late POCD). Of importance in this study is that the researchers set a high bar for POCD; patients had to experience a score decline of one standard deviation below their preoperative test score in two or more of the cognitive tests administered postoperatively [46]. Finally, in an effort to determine the extent to which placing the patient on a cardiac bypass pump created dysfunction, the patients were analyzed as two groups: one whose surgery was conducted with the patient "on-pump" and the other whose surgery was conducted with the heart still beating "off pump." In this study, researchers found that 80.7% of the patients overall experienced early POCD, while 38% experienced late POCD [46]. When the researchers looked at rSO₂ values, they found that a value of 50% or lower was predictive for the development of early POCD. In patients who had this level, approximately 40% experienced early POCD, while 8% did not. Also, they found patients were more likely to have POCD if placed on bypass pump [46].

Unfortunately, the preponderance of evidence related to POCD after cardiac surgery is still evolving. Two studies show the divergence of opinion on both the cause and the extent of POCD. In the first study. researchers sought to determine the extent to which cerebral hypoxia resulted in POCD [47]. These authors found that cerebral oxygen levels were not predictive of POCD, and their reported rate of POCD was 23%. In a second study, researchers looked at 1,156 patients who had open heart surgery and compared the rates of POCD between the two groups [48]. This was a large, multisite study, in which participants received an extensive neurocognitive evaluation after open heart surgery. In this sample, 581 of the participants had surgery using cardiopulmonary bypass pump, while 575 had off-pump surgery. The neurocognitive scores were measured approximately two days before surgery (baseline) and one year after surgery. In this large sample, the only predictors of POCD were lower baseline cognitive score, older age, lower level of education, and non-White ethnicity [48]. Use of the bypass pump had no effect on score differences. Further, the researchers reported that at one year, the rate of POCD was less than 14% in both sub-samples [48]. This is markedly different from some studies showing levels higher than 30%.

In one final study, researchers followed 326 patients for 7.5 years after their cardiac surgery to determine the extent to which POCD progressed into dementia [49]. In contradistinction to the previously described study, these researchers found a POCD rate of 32.8% and dementia rate of 30.8% in the patients they followed [49]. In the patients they followed, pre-existing cognitive impairment was present in 32.2% of those with POCD or dementia. They also found that pre-existing peripheral vascular disease was a significant predictor for POCD. As will be discussed later in this course, atherosclerotic vascular disease is associated with increased levels of systemic inflammation, which may be significant. Finally, the presence of POCD at 3 months and 12 months was associated with increased mortality [49].

While cardiovascular surgery may have initiated the interest in POCD, the predictive indicators are far from settled. Considering only cardiovascular procedures, it is safe for the clinician to infer that:

Orthopedic procedures have been linked to systemic postoperative problems, ordinarily linked to the development of thrombi, often described as bone marrow or fatty embolisms. These are usually linked to long-bone procedures (e.g., repairs of the femur or humerus, especially those involving the insertion of hardware into the damaged area). Therefore, it is not surprising that the interest in POCD after orthopedic procedures has developed. Further, orthopedic implantations (e.g., total hip placements, total knee replacements) may be held in place by an exogenous bone cement (e.g., polymethyl methacrylate), which is associated with transient hypotension in patients.

Total joint arthroplasty (replacement) of hips and knees has become commonplace, particularly in older adult patients. These procedures are associated with low mortality rates (around 0.2%) and provide patients with an improved quality of life [50]. However, they are associated with significant rates of POCD. As early as 1998, research indicated that 26% of patients were experiencing POCD one week after orthopedic surgery, and despite decades of progress in surgery and anesthesia techniques, the rate remains 18.9% [51].

Clinicians have sought approaches to decrease these rates or risk factors that would be helpful in predicting the occurrence of POCD. Orthopedic surgery may entail significant loss of blood, necessitating transfusion. This is especially pronounced in older adult patients, who tolerate hemorrhagic anemia poorly. One group of researchers looked for an association between the administration of blood (and its concomitant release of inflammatory mediators) and the presence of POCD [52]. In this study of 313 patients older than 65 years of age, all of whom received total hip replacements, the research team measured neurocognitive function preoperatively, and seven days postoperatively, using the MMSE [53]. The researchers found a statistically significant decline in MMSE scores, which they used to identify POCD. This POCD group was 1.5 times more likely to have received three or more units of transfused blood than the non-POCD group [52].

Another study sought to determine the relationship, if any, between the stress hormone cortisol and the presence of POCD in patients after hip fracture surgery [54]. In addition to viewing POCD as being linked to inflammation, this group hypothesized that cognitive dysfunction would occur after a stressful event, and that cortisol, as a stress biomarker, might be important in finding a link between orthopedic surgery and POCD. They looked at 175 patients who presented for hip fracture surgery, all of whom were older than 65 years of age. All patients received spinal anesthesia rather than general anesthesia. The research team reported that POCD occurred in 29% of their patients when measured at postoperative day seven [54]. Plasma cortisol was inversely correlated with MMSE score—as the stress biomarker increased, the neurocognitive scores went down [54].

Other researchers sought to determine the extent to which the administration of a general anesthetic would lead to the development of POCD when compared to regional anesthesia [55]. A problem with studies like this is that patients receiving regional anesthesia frequently are moderately to heavily sedated; this study ameliorated that confounding variable by not administering any sedatives before or after the application of the regional anesthetic or during hip replacement surgery. Researchers found that patients receiving general anesthesia had significantly lower MMSE scores on postoperative day one than both the scores on their own baseline, as well as when compared to the scores of the regional anesthesia group [55]. By the fifth postoperative day, neither of the groups had neurocognitive scores that differed significantly from their preoperative scores. However, the researchers also ran a univariate regression on amyloid beta proteins (Ab), which have been found to be biomarkers for Alzheimer disease [55,56]. They found a strong inverse correlation between the presence of Ab and MMSE score. In short, 74% of the variance in neurocognitive scores was predicted by Ab levels in the blood [55]. This is interesting evidence that POCD may have the same pathogenic model as Alzheimer disease.

Another orthopedic study sought to further use these data to predict POCD by measuring Ab and levels of the Tau protein on patients undergoing total knee replacement and total hip replacement [56]. As background, Tau protein helps to construct the microtubules for neurons and helps neurons maintain their structural integrity. This group of researchers enrolled 80 patients having knee replacements or hip replacements and a matched sample of 80 patients not having surgery into their study. They tested neurocognitive function before surgery and then seven days, one month, and three months after surgery. In an effort to decrease the confounding variables associated with general anesthesia, all of the patients were anesthetized using a regional technique (combined spinal and epidural) [56]. Their measurements showed that 40% of patients experienced POCD at day 7, 25% at the end of one month, and 15% at the end of three months. When they performed a ratio of the concentration of Ab protein to tau protein, patients with POCD had significantly lower ratios than those not experiencing POCD. The authors found this ratio to serve as a potential biomarker for the development of POCD [56].

Other cases with high postoperative rates of POCD have also been reported. It is important that such cases are studied, as other procedures may be more frequently performed than cardiovascular surgeries. One consideration is that the complexity and invasiveness of the surgery may not be correlated to the development of POCD. One of the first attempts to describe the incidence of POCD in patients after minor surgery enrolled 372 patients older than 60 years of age, of whom 199 remained in the hospital for one night and 173 were discharged home after surgery [57]. Neurocognitive tests were performed preoperatively and then repeated seven days and three months after surgery. Of this group, researchers noticed a significant difference between those recovering at home compared with those recovering in the hospital. The one-week POCD rate for those recovering in the hospital was 9.8%, while those recovering at home had a rate of 3.5%, a significant difference [57]. When the cognitive tests were repeated at the three-month mark, there was no statistically significant difference between inpatients (8.8%) and out-patients (4.5%) [57]. The authors did believe that the decreased stress of home recovery probably accounted for the one-week difference in scores, which mirrors the previously discussed studies using cortisol and other stress biomarkers. The lack of difference after three months was not explained, though the incidence was lower than that found in other studies of more complex surgeries.

Another study followed 1,064 adult patients (older than 18 years of age), testing their neurocognitive status preoperatively, one week after surgery (or at discharge if hospitalized less than one week), and once again three months postoperatively [58]. One of the major strengths of this study was a comparison of POCD effects across ages of participants as well as across time. Major surgery was defined as procedures requiring postoperative hospitalization for a minimum of two days. A complete battery of neuropsychiatric tests was administered to all the patients in the study, evaluating all areas of executive function. The participants were divided into three age groups: younger (18 to 39 years of age), middle-aged (40 to 59 years of age), and elderly (60 years of age and older) [58]. At discharge (or one week postoperative), the POCD rates were 36.6% among younger patients, 30.4% for middle-aged patients, and 41.4% for elderly patients. At three months post-discharge, POCD rates fell to 5.7% for young patients, 5.6% for middle-aged patients, and 12.7% for elderly patients [58]. The authors also followed up on mortality rates one year after discharge. Of concern was the finding that for those who had no POCD, the one-year mortality rate was 4%, compared with a mortality rate of 10.6% for those who did have POCD [58]. In this case, POCD was clearly prognostic for increased mortality.

Laparoscopic cholecystectomy (excision of the gall bladder via a small incision and the use of an endoscopic device) is a common surgical procedure. Its sheer volume reflects the importance of a study concerning the development of POCD in this group of 120 elderly patients. The study also compared the utility of the alpha-2 agonist dexmedetomidine anesthetic regimen compared to a control group receiving a placebo [59]. The utility of dexmedetomidine will be discussed later in this paper, of concern at this point is the rate of POCD for cholecystectomy. The incidence of patients who experienced "mild" POCD impairment was 38% in the control group, and 18% in the dexmedetomidine group, while those experiencing "moderate" impairment was 4% in the control group and 2% in the dexmedetomidine group [59]. The presence of POCD remains consistent, despite the intensity of surgery performed.

Acetylcholine (ACh) is one of the key neurotransmitters in the brain. The primary receptor responsible for this neurotransmission is the nicotinic acetylcholine receptor (nAChR), which is expressed ubiquitously in numerous subtypes throughout the body [60]. Anticholinergic drug use has been related to both POC and post-operative delirium [61]. The normal role of ACh in the brain is primarily one of exciting neurons and causing their discharge. The neurons most affected by ACh are ones that send projections throughout the cortex and striatum, areas responsible for wakefulness, learning, attention, and motivation [62]. Degradation of these neurons, or interruptions in their discharge, would lead to signs and symptoms associated with POCD [63]. Inhaled and intravenous anesthetic agents inhibit the nAChR [64]. This in turn blocks the entry of sodium ions into the cell as well as the entry of calcium ions. The former moves the electric potential within the cell closer to threshold, while the latter functions as a second messenger for numerous intracellular effects. One of the other effects of anesthetics is the desensitization of these receptors, such that even if ACh can bind, the receptor will not respond [64]. This desensitization may last after the anesthetic is no longer in the vicinity of the receptor and may account for prolonged POCD effects.

The one area in which there was common agreement was that of inflammation after surgery. In one study, the authors found POCD in patients three-months post-procedure in such disparate surgical interventions as open heart surgery (16%), total hip replacement (16%) and coronary angiography (21%), despite the fact that the last procedure was minimally invasive [65]. After nearly any procedure or injury, the natural inflammatory response of the body is activated, and numerous mediators are released into the bloodstream. When inflammation occurs, the blood brain barrier undergoes a breakdown, allowing these circulating mediators to enter the brain [66]. Numerous mediators may result in neuronal damage (Table 1).

The presence of brain inflammation appears to satisfy all the requirements needed to explain POCD. Most patients who have surgery share the commonalities of a heightened stress condition preoperatively (whether from a physical condition causing pain, or the psychic stress from the knowledge of impending surgery) and a normal inflammatory response to surgical intervention postoperatively. The release of inflammatory cytokines during this normal response has been associated with numerous neural and behavioral outcomes. One review, which was primarily concerned with psychiatric illness, discussed in depth the effect of cytokines on what was described as sickness behaviors [5]. Using an animal model, physiologic changes in the brain were described, which resulted in diminished social interactions, coupled with cognitive and affective dysfunction, the latter characterized by disturbed operant performance and anhedonia [5]. If this is compared to descriptions of patients experiencing POCD, the similarities are striking. In animals, these behaviors are viewed as adaptive; they create an environment in which the animal can engage in rest and recuperative behaviors. This latter finding has not been assigned to humans, in whom POCD is viewed as maladaptive in nature. The author also reported that the presence of cytokines in the brain may result in motor impairment, a finding in neurocognitive tests of patients manifesting POCD. The elevated presence of mediators such as Il-6, Il-8, and Il-10 was found in postoperative hip replacement patients four weeks after surgery [5]. In the animal model, the presence of the inflammatory mediators TNF-α and Il-1β was found to negatively impact the ability of the hippocampus to create long-term potentiation in the brain, the mechanism by which memories are stored.

Other studies have shown that activation of glucocorticoid and mineralocorticoid receptors during stress and inflammation, while enhancing memory in younger patients, degrades memory in older patients [9]. This would help account for the difference in the incidences of POCD across the lifespan. These steroid receptors are found prominently in the hippocampus, and there is good evidence showing that anesthetics also influence other receptors here, thus resulting in the inability to form memories [68]. A 2013 study found that older patients having hip surgery had an inverse correlation between MMSE scores and plasma cortisol levels [54]. In another study of 94 patients, all of whom were older than 60 years of age and undergoing major noncardiac surgery, salivary cortisol ratios were tested to assess for a link between endogenous steroid activation and POCD [67]. All patients received neurocognitive testing preoperatively, and then had their saliva sampled for cortisol in the morning (between 6:00 a.m. and 7:30 a.m.) and in the evening (between 9:30 p.m. and 11:00 p.m.) on the day before surgery. It is well-established that glucocorticoid levels are diurnal in nature, thus the sampling enabled researchers to obtain the widest range of values. The research team then set up a ratio of morning cortisol value to evening cortisol value [67]. Neurocognitive tests were administered one week after surgery and the results evaluated against the cortisol ratios obtained. The preoperative morning:evening cortisol ratio was nearly twice as high in patients with POCD as in the those without POCD [67]. The researchers suggested that this finding could be used as a simple screen to preoperatively identify those at risk for the development of POCD. Additionally, it lends credence to the theory that the development of POCD is related to stress and inflammation.

Since the 1950s, researchers have been looking for the pathologic nature of postanesthesia cognitive dysfunction. Most research followed the pathway that leads to neuronal damage, meaning that something in the anesthetic or surgical intervention was responsible for causing damage to neurons, resulting in visible behavioral changes in the patient's ability to deal with day-to-day activities. Slowly, studies emerged that provided a list of things that probably were not responsible. This is not to say that any of these, in excess, could not result in neurologic damage to a given patient, but rather that minor changes in these variables were probably not the primary cause of POCD.

Lower oxygen concentrations did not prove to be predictive of POCD, nor did hypercarbia or hypotension. This makes sense based on the brain's ability to maintain constant cerebral blood flow over a wide range of mean arterial pressures (60–160 mm Hg) [66]. This is not to say that such conditions could not damage the brain, but unless found in extreme conditions, these situations were not predictive for POCD. Studies in all of these areas produced varied and inconclusive results.

Next, it was assumed that the action of anesthetics themselves could be linked to POCD. In one study that extensively discussed the problems of neurotoxicity associated with inhaled anesthetics, the researchers concluded that there is limited clinical data directly linking anesthetic agent exposure to cognitive dysfunction in the aging population [23]. POCD exists in varying rates, regardless of the anesthetic used. Further, as evidenced in some of the studies discussed, a number of patients who received regional anesthesia without any form of sedation still developed POCD [54].

Dexmedetomidine, an alpha-2 agonist, is used as an adjunct agent for anesthesia as well as a sedative in the intensive care unit [70]. Alpha-2 agonists bind on presynaptic neurons and act to diminish neurotransmitter release, with an overall effect of diminishing the output of the sympathetic nervous system. As the sympathetic nervous system is integrated into the inflammatory response, the administration of dexmedetomidine suppresses inflammatory cytokines after surgery [59]. In one study, researchers sought to determine if the use of dexmedetomidine as part of an anesthetic regimen resulted in a lower incidence of POCD when compared with a placebo group. They found significantly less mild cognitive impairment 24 hours after surgery (18% in the dexmedetomidine group vs. 38% in the placebo group); however, there was no difference in MMSE scores two days after surgery [59]. The authors did not follow the patients beyond 72 hours, so they were unable to assess if there was a prolonged effect from the dexmedetomidine. The study does add to the preponderance of evidence that POCD is inflammation related.

The American Society of Anesthesiologists (ASA) Perioperative Brain Health Initiative recommends that prevention of perioperative neurocognitive disorders should involve optimized postoperative pain control, preferably with minimally sedating multimodal pain management.

(https://www.bjanaesthesia.org/article/S0007-0912(20)30925-9/fulltext Last Accessed: February 28, 2026)

Level of Evidence: Expert Opinion/Consensus Statement

Another study compared the administration of the nonsteroidal anti-inflammatory drugs (NSAIDs) to that of traditional opioid administration in a large sample of 1,062 patients presenting for noncardiac surgery [71]. These researchers investigated numerous postoperative sequalae that they defined as clinically meaningful events. Some of these clinically meaningful events included outcomes commonly linked to POCD, including fatigue, inability to concentrate, and confusion [71]. These were self-reported by the patients and not based on any rigorous neurocognitive tests, but the results still bear review. When compared with the patients controlling pain solely with opioids and placebo (rather than NSAIDs), patients reported fewer symptoms, and the NSAID group showed better patient-described neurocognitive function [71]. As NSAIDs work primarily by interrupting inflammatory pathways, the increasing use of perioperative NSAIDs may hold out some hope in decreasing POCD.

Another study tested the degree to which the perioperative infusion of the local amide anesthetic lidocaine would decrease POCD [72]. The authors theorized that lidocaine would act in a neuroprotective fashion and sought to determine if the mechanism of action of this neuroprotection had to do with its effects on levels of circulating inflammatory cytokines and mediators, including IL-6, TNF-α, S100β, neuron-specific enolase, and malondialdehyde. A 2019 study found malondialdehyde levels to be elevated in patients with POCD [73]. Returning to the lidocaine research, Chen and colleagues studied two groups of 40 patients presenting for spinal surgery, one group receiving lidocaine 1 mg/kg over five minutes after the induction of general anesthesia followed by a continuous infusion of 1.5 mg/kg/hour until the completion of surgery, while the second group received an equal fluid volume of normal saline [72]. The team measured blood levels before surgery, at the end of surgery, and three days after surgery. They administered the cognitive tests preoperatively and three days postoperatively. Blood cytokine tests showed significantly higher levels of IL-6, S100β, and neuron-specific enolase in the placebo group at both postoperative time points, indicating that the lidocaine infusion did effectively decrease circulating cytokines [72]. Serum malondialdehyde was significantly lower in the lidocaine group only on the third postoperative day. In terms of MMSE scores, the lidocaine group had a significantly higher mean score than the control group on the third postoperative day, and the postop scores of the lidocaine group did not differ from their preoperative scores. In contrast, the placebo group postoperative scores were significantly lower than their preoperative scores. Further, their postoperative scores were lower than the postoperative scores of the lidocaine group [72]. Lidocaine is a well-known and inexpensive agent and supplements general anesthesia, thus its use in patients who have known risk factors for POCD may be warranted.

One idea for decreasing the incidence of POCD and treating possible postoperative effects is the administration of statin drugs prior to surgery [78]. In addition to their cholesterol-lowering effects, statins reduce inflammation throughout the body, including in the brain. Note that the inflammatory cytokines produced by surgery are nearly all impacted by the administration of statins, and there is research supporting their possible usefulness in treating delirium (rather than POCD). A 2009 study showed that the administration of statin drugs to a group of more than 1,000 patients reduced their incidence of postoperative delirium by 46% [79]. Further, studies are showing some cautious optimism in the use of statins as adjuncts to treat neuropsychiatric disorders, including depression [80]. Because these symptoms are similar to those experienced by patients with POCD, and because these agents are useful for conditions that are based on inflammatory processes, it seems reasonable to assume the same positive effects would occur in POCD.

The dissociative anesthetic ketamine has recently been introduced as a potential option for treatment-resistant major depressive disorder and post-traumatic stress disorder [20,21]. Further, at least one study has shown that the use of perioperative ketamine decreased the likelihood of POCD [81]. Some researchers have advocated its use to decrease the incidence of POCD, but so far, its use in treating existing cases of POCD in postsurgical patients has not been reported [23].

A 2017 animal study showed that rats responded well to electroacupuncture [82]. This research team took four groups of 20 rats and operated on three of the groups (partial hepatectomy), retaining one group as a control. In the three surgical groups, the rats received either surgery only, surgery and the anti-inflammatory drug minocycline, or surgery and acupuncture in the post-surgery period. They found the rats receiving electroacupuncture had their memory improved compared with rats in the other two surgery groups and that all rats had memory declines after surgery compared with the control group [82]. No human studies have used this technique to date.

Cognitive-behavioral therapy (CBT) has been touted as the criterion standard for the treatment of a number of neuropsychiatric and neurocognitive disorders [83,84]. Its basis is on engaging patients to reflect on how their thoughts and emotions are interlinked and to use the cognitive skills they have to influence how they think about given circumstances [84]. One might be excused for thinking that in a patient with POCD with impaired cognitive abilities this technique might be of limited utility. In fact, this type of therapy has been used in others with cognitive dysfunction, including those with major depressive disorder, attention deficit hyperactivity disorder, obsessive compulsive disorder, and others [83]. As many of the symptoms of these disorders result in degradation of executive function, it may be possible in the future that such a technique might be attempted to ameliorate POCD degradation. For now, there is no evidence supporting use of CBT for the treatment of POCD.

One of the more interesting studies concerned the use of photobiomodulation to mitigate POCD symptoms. This technique is not well-known, but in its most basic form, photobiomodulation is the effect of red light in very specific wavelengths that is shined on the head and thought to affect neurotransmitters and initiators of protein construction in DNA [85]. It is reported that one of the effects is to decrease NFK-b, a cytokine that diminishes neurocognitive function [85]. The use of photobiomodulation has been suggested both in preconditioning patients to resist the effects of anesthesia and surgery on neurons as well as postoperatively to treat the adverse effects of POCD [86]. To date, no studies have been published regarding the use of this technique in humans with POCD.