Recently, there have been more articles published on post operative delirium (POD) in the anesthesia literature. The ASA has added additional training related to this topic as we become more aware of the significant consequences of post operative delirium in our aging population. We have learned that those who experience POD will likely see an increased duration of hospitalization of two to three days. In addition, it has become apparent that POD is associated with 30 day mortality of 7 to 10% vs. only 1% in those without POD. Other associated consequences include increased healthcare costs, increased likelihood of requiring longer term nursing care and significant functional decline.
Furthermore, much research into this topic has identified several factors that increase the risk for POD. Among the largest risks include patients with preexisting neurocognitive disorders, elderly patients, those undergoing complex and prolonged surgery or emergency surgery especially if requiring post operative ICU admission. A simple scoring tool has been developed and validated named the delphi score to help identify patients at 'high' risk for POD. I have listed the items considered in this scoring tool to place patients at higher risk to provide further guidance on factors increase risk in our patients.
- age > 70, 1 pt age > 80 2 pts
- Needs assistance for physical activity - 2 pts
- heavy alcoholism - 1pt
- hearing impairment - 1pt
- history of delirium - 2 pts
- Emergency surgery - 1 pt
- Open Surgery - 2 pts
- ICU admission - 3 pts
- C reactive Protein - >/= to 10 mg/dL
Our understanding of POD has also increased over the last several years. For example multiple studies have found that POD is associated with increased levels of inflammatory markers such as CRP and IL-6. It has been shown that with peripheral trauma, there is damage to the BBB resulting in leakage of inflammatory modulators into the CSF. Alteration of neurotransmitter balance has also been associated with POD. For example Ach is thought to be involved in neuroplasticity and is involved in several pathways responsible for attention and memory. Studies have also found that patients experiencing POD had lower levels of acetylcholinesterase both preoperatively and up to two days post operatively. Also, it has been found that low levels of acetylcholinesterase was an independent risk factor for developing POD. In addition, central acting anticholinergic medications have also been associated with increased incidence of POD. In addition, there is some evidence that non optimized perioperative perfusion pressure of the brain may be associated with increased frequency of POD. It has been suggested that subclinical cerebral ischemia may manifest in the post operative period as delirium. This thesis is still being explored. What is clear, is that perioperative delirium can be mitigated by perioperative interventions over which we as anesthesiologist control.
A recent review of the literature on POD provided a list of interventions with evidence supporting benefit [1]. From this article is a list of preoperative interventions we can consider in at risk patients.
1. Avoid perioperative polypharmacy
2. Avoid prolonged fluid fasting (i.e. > 6 hrs)
3. Adequate and aggressive perioperative pain management
There are also intraoperative interventions we can take that have evidence supporting its ability to reduce the incidence of POD. From the same article these include:
1. Monitoring depth of anesthesia. In a meta-analysis it was reported that depth of anesthesia monitoring is associated with a significantly lower risk of delirium [3]. In Anesthesiology, a meta analysis of 13 RCT was published demonstrating that the OR for POD was 0.62 when processed EEG was used to guide anesthesia [2]. In cardiac surgery three studies demonstrated that patients who had intraoperative burst suppression had an increased incidence of POD [4,5,6]. In a study specific to low BIS values in aortic surgery, it was shown that patients with a lower BIS were more likely to suffer from post op stroke, TIA, increased ICU length of stay, and increased incidence of POD [7]. In an RCT looking at non cardiac surgery patients experiencing cognitive decline both acutely (1 week) and prolonged (up to 1 year), found that significant cognitive impairment occurs after major non cardiac surgery in patients greater than 60 years old which can still be evident up to one year after surgery compared to aged match controls [8]. Importantly, patients receiving the monitoring intervention were within the optimal BIS target range (40–60±5) for a significantly higher proportion of the intra-operative period and there was a significant relationship between the proportion of the intra-operative period spent in this optimal window and both cognitive outcome and S100B, a marker of brain injury [8]. This study prompted another study one year later (2013) which was an RCT of BIS v. no BIS in major non cardiac surgery. The study fount that in multivariate analysis, the percentage of episodes of deep anaesthesia (BIS values <20) were independently predictive for postoperative delirium (P=0.006; odds ratio 1.027). BIS monitoring did not alter the incidence of postoperative cognitive dysfunction, however [9]. In yet another study looking at elderly frail patients undergoing spine surgery, comparing those patients who had burst suppression on EEG it was found that the group that suffered burst suppression also had increased POD compared to the no burst suppression group. This study also demonstrated that elderly frail patients were more likely to have burst suppression during surgery [10]. In another observational study, Fritz et al in 2016 [11] found that after adjusting for confounders using statistical methods, patients who who had greater degrees of intraoperative burst supression suffered increasing POD in a dose dependent fashion (greater time in burst suppression). Importantly, the authors noted that increased volatile agents concentration AND lower opioid dosages both were predictive of greater burst suppression. As part of this theme, Fritz et al. in the BJA were able to show that intraoperative processed EEG monitoring can allow the astute anesthesiologist to predict which patients are very likely to suffer from POD. In a paper published in 2018 [26], they found that patients who suffered suppression at lower volatile anesthetic concentrations than there age calculated value would indicate are 2.6 times more likely to develop POD in ICU.
In yet another study, Chen at al. were able to demonstrated that implementation of BIS type monitoring dramatically reduced the incidence of POD in elderly patients. This benefit only became statistically significant in the cohort aged 75 years or older. For patients aged 60 to 74, there was only a trend showing improvement with BIS type monitoring [12]. In a particularly interesting RCT study patients aged 50 or older undergoing major surgery were targeted to a BIS of 50 or 35. The group targeted to a BIS of 35 had an increased incidence of delirium (57% v. 34%; p=0.009 OR=0.53). Importantly, in this same population, those who experienced POD, experienced other negative outcomes including: increased unplanned ICU admissions, increased hospital stay, and higher post op MI rate. They also did a sub analysis of time in burst suppression and found that time in burst suppression was significantly longer in those who went on to experience delirium and those who experienced delirium had an higher mortality rate at one year (12% v. 6%). It should be noted that this trial differed from the results of the ENGAGES trial that randomized patients 60 years and older to BIS guided v. non BIS guided anesthesia for major surgery and did not find a statistically significant difference in POD. This is likely due to the fact that the difference in MAC of the volatile anesthetics between groups was similar (difference of 0.11 MAC). Other studies have shed light on depth of sedation and its correlation to incidence of POD. Sieber et al. [14] found that in hip fracture patients (high risk for POD), using a BIS targeted light sedation during spinal anesthesia did not decrease incidence of POD compared to deeper sedation. The BIS was on average 82 in the light sedation group vs. 57 in the deep sedation group. Two takeaways from this study that inform our practice today are the following: 1) both groups had a BIS greater than 50 which may represent a threshold above which no further improvement can be achieved. 2) the subset of patients in this study with lowest co morbidity had a 2.3 fold increase in POD in the deep sedation group. This suggests that in very frail patients that also have significant co morbidities have such a high baseline risk for POD, that reducing BIS may provide little protection.
To utilize processed EEG to inform us in the OR on our patients risk for POD, it is important to spend a little time understanding how the EEG is affected by different brain states. In particular, as the brain ages the component of the EEG that makes up what are called alpha oscillations is decreased. It has been well established that decreasing alpha oscillations is typical of older brains and as the degree of alpha oscillations are depressed, you see increased risk of POD. Additionally, younger patients with a vulnerable brain (by trauma, disease, etc) can be identified by a decrease in alpha power on spectral array and mimicks the aged brain in many way. It is well known that as alpha power falls, the chances of burst suppression increases. Furthermore, in older brains, the component oscillations known as beta may be elevated. This is a normal pattern in older brains, but on average can cause the processed EEG to read out a higher number by essentially fooling the algorithm. In the BIS, the average increase in BIS reading for patients older than 70 is about 3.5 units on the BIS. In other words, if you have a 72 year old patient with a BIS reading of 40, on average, you can expect that the true BIS value is closer to 36.5. Below is from a study where it was demonstrated that beta power is increased in elderly patients (>70 yrs) compared to younger patients which confuses the BIS algorithm.
 |
In standard commercially available BIS monitors, the suppression ratio is also provided. This number is not typically followed as closely as the BIS in standard clinically practice. The suppression ratio represents the percentage of time in the previous 63 seconds in which the EEG signal is considered suppressed. Suppression is recognized as those periods longer than 0.5 seconds, during which the EEG voltage does not exceed approximately ± 5 μV. Typically a period of suppression is followed by a burst of activity which varies depending on the anesthetic agent used. For example Propofol produces bursts that are shorter in duration and of lower amplitude than volatile anesthetics. This "burst" of activity is the thalamus activating as it tries to kick start the cortex back into action.
This figure demonstrates a quirk in the BIS algorithm. Essentially, if the suppression ratio is greater than 40%, then the BIS value displayed will drop 1 index value for every 2% increase in the suppression ratio. In other words, there is a linear inverse relationship between suppression ratio and BIS value where as suppression ratio increases, BIS drops as a result. However, if the suppression ratio is less than 40%, the displayed BIS value will not consider the suppression ratio in its algorithm and may be artificially high. An example is provided below.
This displays a BIS of 40, however, the raw EEG data on the display shows a rather flat line that looks depressed and the suppression ratio is 20%. This would represent a patient who is most likely deeply anesthetized, far more than the BIS value of 40 might indicate. In other words, when the BIS value reads between 30 and 40, it can represent a wide range of brain states. This is where looking at the suppression ratio becomes critical to better determine the state of your patient's brain. The take home is maintaining the BIS above 40 helps maintain the integrity of the data. Additionally, shooting for a suppression ratio of 0 is also important. If the the suppression ratio is less than 40% but greater than 0, the displayed BIS value is likely artificially high.
Also important to note is that burst suppression can be augmented by different agents. For example, the MAC burst suppression (MACbs) for ISO is 1.3 x MAC, while for Sevo it is 1.4 x MAC (i.e. patients would be more vulnerable to burst suppression when using iso v. sevo). Propofol causes burst suppression at 4.9 mcg/mL, in 50% of patients, but requires 15.2 mcg/mL to inhibit movement in 50% of patients. This indicates that propofol is excellent at cerebral suppression (i.e. burst suppression), but very ineffective at suppression of spinal function.
We should also recognize what clinical factors might put a patient at higher risk for suffering burst suppression. The following are risks for increased likelihood of going into burst suppression: Increasing age, particularly above about 60 years, coronary artery disease, and male sex. Of course significant co morbidities also place patients at risk as well as critically ill patients. In most patients it has been shown that below a BIS of 40, you start to run into problems with burst suppression. However, given the above information, the cut off may change depending on clinical risk factors. In relation to age, Besch et al. were able to demonstrate that when the BIS is maintained between 40 and 60 in patients greater than 80 years old, they are 10 times more likely to be in burst suppression than patients younger than 60 years old. Furthermore, patients 60 to 80 were nearly five times as likely to be in burst suppression as younger patients at the same BIS levels (i.e. BIS targeted to 40 to 60) [27].
It should be noted that many events during an anesthetic can result in burst suppression. These include significant hypothermia, hypotension, significant anemia, hypoglycemia, hypoxia and vascular brain injury.
The correlation of lower BIS values at lower concentrations of inhaled anesthetics was also recently shown to correlate with increased rates of POD [28]. This study utilized the ‘The Duke Anesthesia Resistance Scale” to determine how resistant each patient was to the inhaled anesthetic during surgery. The scale was calculated with the following equation:
DARS= (1/(2.5 - average Age adjusted MAC))*BIS
They found that a lower DARS predicted patients likelihood of developing POD in multi variant regression analysis. In other words, a lower DARS indicates that the patient has a “less active brain than would be expected than would be expected for a given inhaled anesthetic dose”. One caveat is important to keep in mind. The results of the study were non linear. In other words, the likelihood of POD did not increase with each incremental decrease in score on the DARS. Instead, the chances were increased only when a certain critical value was hit (the value was less than 28). Also, this study was done in elderly at risk patients. Importantly, the DARS score was a better predictor than age alone in predicting the likelihood of POD. Unfortunately, the methodology of the study does not allow for easy routine clinical implementation, however, experienced anesthesia providers use heuristics everyday and this study I believe is amenable to the same heuristic process we use everyday.
To summarize the various results from multiple studies it becomes clear that in at risk patients, particularly as the age approaches 70 years old, targeting a BIS of closer to 50 is likely to make a difference. In otherwise healthy and younger patients (60 year and younger), targeted BIS is not likely to make a clinically significant difference in the incidence of POD. In the highest risk patients, frail, elderly patients with hip fracture (or other major surgery) with significant co morbidities, targeting BIS to a higher level may have little impact on decreasing incidence of POD, however, it is unlikely to provide harm and thus, it would still seem reasonable to spend the money and effort on BIS monitoring.
Processed EEG is also useful in our ability to recognize the vulnerable brain. A vulnerable brain is one that may exhibit increased sensitivity to anesthesia or be more likely to develop problems after anesthesia. A vulnerable brain may be present in other (atypical) patient populations such as those with a TBI, a significant brain tumor or even significant metabolic process or infection. There is data to suggest that neurons in vulnerable brains have a lower mitochondrial production of energy substrate leading to lower electrical activity from decreased synaptic transmission. Since it is known that the alpha band (alpha power on spectral array analysis) is related to thalamocortical electrical activity and thus reflective of overall brain health, we can look at this specific band to diagnose the probability of the vulnerable brain. During a propofol or volatile anesthetic, we should expect to see a clear "alpha band" that is stable. When this alpha power is diminished from the expected value given the age of the patient, we should consider this an indicator of the vulnerable brain state. In this same vein, patients with low alpha power during the intraoperative power are known to have reduced cognitive function in the PRE operative period, and thus, again, lower alpha power is a significant indicator for higher POD risk in the post operative period.
2. Use of a multimodal opioid sparing analgesia. Multiple observational studies have demonstrated that increased pain scores after surgery are associated with increasing POD incidence and severity. Conversely, the use of post operative opioids particularly long acting opioids are also associated with increasing POD. Therefore, controlling pain with non opioid methods is ideal in at risk patients. Furthermore, two larger observational studies have been able to demonstrate a 20 to 40% lower incidence of delirium when regional anesthesia is used. Another important aspect of multimodal analgesia is the use of NSAIDs and tylenol. In animal studies, NSAIDs have been shown to reduce neuroinflammation secondary to cerebral ischemia-reperfusion, and neuroinflammation secondary to remote insults. In another large observational study of more than 1 million surgical patients, parecoxib was associated with a significant reduction in POD. Paracetamol was shown to reduce the risk of POD from 28% to 10% (NNT=5.6) in cardiac surgical patients.
Also part of multimodal analgesia is dexmedetomidine; a highly selective alpha 2 adrenoceptor agonist. In animal models dexmedetomidine administration reduced the expression of inflammatory mediators, microglial activation and neuroapoptosis. Wang et al. [15] conducted a meta-analysis of 67 studies, and found that intraoperative dexmedetomidine was associated with decreased concentrations of stress hormones (cortisol, epinephrine), CRP, and TNF-Alpha after surgery. Duan and colleagues [16] conducted a meta-analysis of 18 clinical trials and found that intraoperative and post operative dexmedetomidine administration significantly reduces the risk of POD (OR 0.35). Several recent clinical studies have followed and also demonstrated an improvement in POD with intraoperative dexemedetomidine [17, 18, 19]. Another meta analysis was published by Zeng et al. [20] showing that dexmedetomidine could reduce the incidence of POD in elderly patients after non cardiac surgery with a NNT=10.
3. Avoidance of certain medications in perioperative period
Today, most institutions have instituted a pre formulated pathway to enhance recovery from surgery. Known as ERAS, institutions and many surgeons have jumped all in to this paradigm. ERAS calls for a multitude of things including perioperative multimodal analgesia. Many ERAS multimodal regimens at most institutions are pre determined and fixed as a one size fits all formula. This has unintended consequences in some cases. As a case in point, gabapentin is often given pre op by the pre operative nurse prior to any one evaluating the patient risks and benefits of this medication. This is largely because the orders are presriptive and come from the surgeons office. As a result many elderly and frail patients will receive a cocktail of medications orally in the pre op holding bay usually including between 600 mg to 900 mg of gabapentin. However, a large observational study has reported that perioperative gabapentinoids are associated with a slightly increased risk of of POD [21].
Scopolamine is an anticholinergic anti emetic. There have been several case reports of scopolamine-induced delirium in the perioperative setting and several expert guidelines have recommended against its use in older patients.
4. Choice of general anesthetics (volatile agents vs propofol)
Miller and collegaues conducted a meta analysis and identified five relevant studies and found no difference between propofol and volatiles anesthetics in the incidence of POD [22]. Furthermore, xenon is considered to have neuroprotective properties due to its ability to inhibit neuroinflammation and apoptosis. However, when compared to xenon, sevoflurance is non inferior in terms of the incidence of POD. This provides support to the idea that volatile anesthetics are likely not inferior to propofol. Currently, at my institution, there is a widely held belief that in any patient who is elderly and therefore, at percieved risk for POD, propofol TIVA is the anesthetic of choice. However, in one RCT, Mei and colleagues compared propofol to Sevoflurane [23], and found that the incidence of delirium was less in the Sevoflurane group compared to propofol, but the difference did not reach statistical significance. In another RCT comparing propofol to Sevoflurane when combined with an epidural for anesthesia for intraabdominal laparoscopy, patients in the propofol group had worse scores on the delirium rating scale used to evaluate the patients compared to Sevoflurane [24]. Nevertheless, there have been several retrospective and prospective observational trials showing that patients receiving propofol had a lower incidence of POD. Recently (2021) in Anesthesiology, a paper was published to address the problem with previous poorly done studies comparing propofol to volatile anesthetics to reduce POD. In this regard a large multicenter RCT was performed where all aspects of POD were carefully considered to reduce confounding influences. In this large RCT, there was no difference in the incidence of post operative cognitive dysfunction between propofol and sevoflurane in at risk patients having abdominal surgery [25]. However, the authors did note that elevated IL-6 at one hour after incision was predictive of patients who would go on to develop post operative cognitive dysfunction confirming previous evidence that neuroinflammation plays a large role in POD. Given the above literature, in summary, one can conclude that lower quality studies with small numbers of patients have shown conflicting results. However, in larger and higher quality studies, there is no difference or sevoflurane is preferred to propofol to reduce POD.
In summary, POD occurs frequently in certain patient populations. Anesthesiologists are becoming more aware of this problem. While much of the risk for POD cannot be mitigated by modifying our clinical approach, we have gotten much better at identifying which risk factors are modifiable. It appears that one of the larger risk factors relates to over dosing of anesthetics leading to significant time under burst suppression. Better education on the use of intraoperative processed EEG can help us avoid burst suppression or at least reduce the time under burst suppression. Avoiding polypharmacy, aggressive use of regional anesthesia in appropriate cases, and the addition of multimodal analgesia (especially with precedex) are other important factors to be considered when attempting to reduce POD in at risk patients.
1. Jin, Z et al. 2020, BJA
2. Sumner M, Deng C, Evered L, Frampton C, Leslie K, Short T, Campbell D. Processed electroencephalography-guided general anaesthesia to reduce postoperative delirium: a systematic review and meta-analysis. Br J Anaesth. 2023 <https://pubmed.ncbi.nlm.nih.gov/35183345/>
3. MacKenzie KK, Britt-Spells AM, Sands LP, Leung JM. Processed Electroencephalogram Monitoring and Postoperative Delirium: A Systematic Review and Meta-analysis. Anesthesiology. 2018 Sep;129(3):417-427. doi: 10.1097/ALN.0000000000002323. PMID: 29912008; PMCID: PMC6092196.
4. Soehle M., Dittmann A., Ellerkmann R. K., Baumgarten G., Putensen C., Guenther U. (2015). Intraoperative burst suppression is associated with postoperative delirium following cardiac surgery: a prospective, observational study. BMC Anesthesiol. 15:61
5. Momeni M., Meyer S., Docquier M. A., Lemaire G., Kahn D., Khalifa C., et al. (2019). Predicting postoperative delirium and postoperative cognitive decline with combined intraoperative electroencephalogram monitoring and cerebral near-infrared spectroscopy in patients undergoing cardiac. J. Clin. Monit. Comput. 33 999–1009.
6. Pedemonte J. C., Plummer G. S., Chamadia S., Locascio J. J., Hahm E., Ethridge B., et al. (2020). Electroencephalogram Burst-suppression during cardiopulmonary bypass in elderly patients mediates postoperative delirium. Anesthesiology133 280–292.
7. Santarpino G., Fasol R., Sirch J., Ackermann B., Pfeiffer S., Fischlein T. (2011). Impact of bispectral index monitoring on postoperative delirium in patients undergoing aortic surgery.HSR Proc Intensive Care Cardiovasc Anesth. 347–58.23.
8. Ballard C., Jones E., Gauge N., Aarsland D., Nilsen O. B., Saxby B. K., et al. (2012). Optimised anaesthesia to reduce post operative cognitive decline (POCD) in older patients undergoing elective surgery, a randomised controlled trial [published correction appears in PLoS One. 2012;7(9). Amaoko, Derek [corrected to Amoako, Derek]] [published correction appears in PLoS One. 2013;8(9).] PLoS One 7:e37410.
9. Radtke F. M., Franck M., Lendner J., Krüger S., Wernecke K. D., Spies C. D. (2013). Monitoring depth of anaesthesia in a randomized trial decreases the rate of postoperative delirium but not postoperative cognitive dysfunction. Br. J. Anaesth. 110(Suppl. 1) i98–i105. 10.1093/bja/aet055
10. Ren, S., Zang, C., Yuan, F. et al. Correlation between burst suppression and postoperative delirium in elderly patients: a prospective study. Aging Clin Exp Res 35, 1873–1879 (2023). https://doi.org/10.1007/s40520-023-02460-5
11. Fritz BA, Kalarickal PL, Maybrier HR, Muench MR, Dearth D, Chen Y, Escallier KE, Ben Abdallah A, Lin N, Avidan MS. Intraoperative Electroencephalogram Suppression Predicts Postoperative Delirium. Anesth Analg. 2016 Jan;122(1):234-42. doi:
12. Chen, YC., Hung, IY., Hung, KC. et al. Incidence change of postoperative delirium after implementation of processed electroencephalography monitoring during surgery: a retrospective evaluation study. BMC Anesthesiol 23, 330 (2023). https://doi.org/10.1186/s12871-023-02293-9
13. Evered LA, Chan MTV, Han R, Chu MHM, Cheng BP, Scott DA, Pryor KO, Sessler DI, Veselis R, Frampton C, Sumner M, Ayeni A, Myles PS, Campbell D, Leslie K, Short TG. Anaesthetic depth and delirium after major surgery: a randomised clinical trial. Br J Anaesth. 2021 Nov;127(5):704-712. doi: 10.1016/j.bja.2021.07.021. Epub 2021 Aug 28. PMID: 34465469; PMCID: PMC8579421.
14. Sieber F., Neufeld K.J., Gottschalk A., et al. Depth of sedation as an interventional target to reduce postoperative delirium: mortality and functional outcomes of the Strategy to Reduce the Incidence of Postoperative Delirium in Elderly Patients randomised clinical trial. Br J Anaesth. 2019;122:480–489.
15. Wang K, Wu M, Xu J, Wu C, Zhang B, Wang G, Ma D. Effects of dexmedetomidine on perioperative stress, inflammation, and immune function: systematic review and meta-analysis. Br J Anaesth. 2019 Dec;123(6):777-794. doi: 10.1016/j.bja.2019.07.027. Epub 2019 Oct 24. PMID: 31668347.
16. Efficacy of perioperative dexmedetomidine on postoperative delirium: systematic review and meta-analysis with trial sequential analysis of randomised controlled trials.
- Lin H.M.
- et al.
- Li C.J.
- Wang B.J.
- Mu D.L.
- et al.
- Su X.
- Meng Z.T.
- Wu X.H.
- et al.
- Zeng H.
- Li Z.
- He J.
- Fu W.
- Memtsoudis S.
- Cozowicz C.
- Zubizarreta N.
- et al.
- Miller D.
- Lewis S.R.
- Pritchard M.W.
- et al.
No comments:
Post a Comment