Effects of early postoperative neurocognitive disorders on clinically relevant outcomes: a meta-analysis

Article information

Korean J Anesthesiol. 2023;76(5):490-500
Publication date (electronic) : 2023 May 26
doi : https://doi.org/10.4097/kja.23126
1Department of Clinical Trials, V. Negovsky Reanimatology Research Institute, Moscow, Russia
2Department of Anesthesiology and Resuscitation, First Moscow State Medical University, Moscow, Russia
3Department of Anesthesia and Intensive Care, IRCCS San Raffaele Scientific Institute, Milan, Italy
4School of Medicine, Vita-Salute San Raffaele University, Milan, Italy
5Department of Anesthesiology and Resuscitation, Vladimirsky Moscow Regional Research and Clinical Institute, Moscow, Russia
6Department of Intensive Care Medicine, Kameda Medical Center, Kamogawa, Japan
7Department of Anesthesiology and Resuscitation, A. Loginov Moscow Clinical Scientific Center, Moscow, Russia
8Department of Anesthesiology and Resuscitation, V. Demikhov Municipal Hospital №68, Moscow, Russia
9Department of Intensive Care Medicine, Federal Research and Clinical Center of Reanimatology and Rehabilitology, Moscow, Russia
Corresponding author: Giovanni Landoni, M.D. Vita-Salute San Raffaele University, Department of Anesthesia and Intensive Care, IRCCS San Raffaele Scientific Institute, Via Olgettina, 60, 20132 Milan, Italy Tel: +39-02 2643 6158 Fax: +39-02 2643 6152 Email: landoni.giovanni@hsr.it
Received 2023 February 19; Revised 2023 May 18; Accepted 2023 May 25.

Abstract

Background

Early postoperative neurocognitive disorders (ePND), include both emergence delirium, which is defined as very early onset postoperative delirium, and emergence agitation, defined as motor arousal. Although research on anesthesia emergence is limited, ePND are likely associated with unfavorable outcomes. This meta-analysis assessed the effect of ePND on clinically relevant outcomes.

Methods

A systematic search of studies published between 2002 and 2022 on MEDLINE, PubMed, Google Scholar, and the Cochrane Library was performed. Studies that included adults with emergence agitation and/or delirium and reported at least one of the following outcomes: mortality, postoperative delirium, length of post-anesthesia care unit stay, or length of hospital stay were included. The internal validity, risk of bias, and certainty of the evidence were assessed.

Results

A total of 16,028 patients from 21 prospective observational studies and one case-control retrospective study were included in this meta-analysis. The occurrence rate of ePND was 13% (data excluding the case-control study). The mortality rate was 2.4% in patients with ePND vs. 1.2% in the normal emergence group (risk ratio [RR]: 2.6, P = 0.01, very low quality of evidence). Postoperative delirium occurred in 29% of patients with ePND and 4.5% of patients with normal emergence (RR: 9.5, P < 0.001, I2 = 93%). Patients with ePND had a prolonged length of post-anesthesia care unit stay (P = 0.004) and length of hospital stay (P < 0.001).

Conclusions

This meta-analysis suggests that ePND are associated with twice the risk of mortality and a 9-fold increased risk of postoperative delirium.

Introduction

Every year, more than 300 million patients undergo surgeries worldwide [1]. Postoperative complications are a major public health issue that have a negative impact on survival, quality of life, and economic costs [24].

Neurological disorders (e.g., delirium and cognitive decline) are common postoperative complications and are associated with increased mortality and morbidity in various populations [5]. Although no consensus on the classification of postoperative neurological disorders currently exist [6,7], experts recommend they be divided into short-term (postoperative delirium) and long-term (postoperative cognitive disorders) disorders [7].

Early postoperative neurocognitive disorders (ePND) occur immediately after surgery and consist of two entities: emergence agitation [6] and emergence delirium [6,7]. Agitation is a well-known condition [8] defined as ‘motor arousal’ that is manifested by purposeless behaviors such as fidgeting, shifting, fiddling, inability to sit or stand still, and wringing of the hands [9]. Emergence delirium, on the other hand, has more recently been assessed in the literature and is defined as very early onset postoperative neurological disorders (e.g., immediately after anesthesia, before or on arrival in the recovery room) [6]. Although the negative effects of short- and long-term postoperative neurological disorders are well established, no comprehensive data on the effect of ePND on mortality and other clinically relevant outcomes are currently available.

This study aimed to evaluate the hypothesis that patients with ePND have an increased risk of mortality and morbidity compared to those with normal emergence from anesthesia.

Materials and Methods

This study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [10]. The study protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO, CRD 42022382008).

Eligibility criteria

The inclusion criteria for this study were as follows: prospective or retrospective studies on adults with emergence agitation and/or emergence delirium that reported at least one of the following outcomes: all-cause mortality, postoperative delirium, post anesthesia care unit (PACU) length of stay, and overall hospital length of stay.

We excluded studies that fulfilled at least one of the following criteria: 1) letters without patient data, 2) case reports, 3) reviews, 4) meta-analyses, 5) pediatric patients, 6) interventional studies, and 7) non-English language articles.

Search strategy

A systematic literature search of MEDLINE, PubMed, Google Scholar, and the Cochrane Library for studies published between 2002 and 2022 (total of 20 years) was performed by three independent researchers (LB, NE, and MY). Medical Subject Heading (MeSH) terms were used. Details regarding the queries are available in the supplemental material (Supplementary Material 1). Additionally, the backward snowballing method (analysis of the references of the included articles and retrieved reviews) was used to retrieve further studies.

Study selection

After removing duplicates, the titles and abstracts of the remaining studies were independently screened by two researchers. Prospective and retrospective observational studies comparing postoperative outcomes of patients with ePND to those with normal emergence were considered.

During emergence after anesthesia, patients with ePND exhibit signs of either agitation or delirium. Agitation was defined as motor arousal according to the Richmond Agitation-Sedation Scale or a similar condition according to other motor scales matching the definition ‘agitation’ in the International Classification of Diseases (ICD)-11 [9] or Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) [11]. Delirium was defined using the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) [12] or the Nursing-Delirium Screening Scale (Nu-DESC) [13] and/or DSM-5 criteria [11].

The inclusion and exclusion criteria were applied to the full texts of potentially eligible studies by LB, NE, and MY. Disagreements were resolved by consensus with the involvement of the supervisor.

Outcome measures and data extraction

Data were extracted by two independent authors (LB and MY). These data included the first author, year of publication, study design, ePND criteria, sample size, mean age, sex, American Society of Anesthesiologists (ASA) status, method of anesthesia administration, operation and anesthesia time, presence of postoperative delirium, PACU and hospital length of stay, and all-cause mortality. The primary outcome of this study was the rate of all-cause mortality at the longest follow-up available for patients with or without ePND. The secondary outcomes were postoperative delirium and PACU and hospital length of stay. Cohen’s kappa coefficient was calculated to assess agreement between the two researchers regarding literature selection and data extraction.

Internal validity and risk of bias assessment

The internal validity and risk of bias of the studies included in the meta-analysis were assessed by four reviewers (LB, MY, KK, and NE) according to the latest version of the new Cochrane tool ROBINS-E (Risk Of Bias In Non-randomized Studies - of Exposures) [14]. Risk of bias was assessed by four researchers (LB, MY, KK, and NE) using the Delphi method in two rounds, and disagreements were resolved by consensus. Differences in estimates were also resolved by consensus. Publication bias and small-study effects were assessed using Egger’s test (MedCalc Statistical Software, version 19.5.6, Belgium) [15] and funnel plot analyses. A systematic Grading of Recommendation, Assessment, Development and Evaluation (GRADE) approach was used to rate the certainty of the evidence. The baseline evidence level was high in the studies on prognostic factors [16].

Data analysis and synthesis

We used STATA 17 (StataCorp LLC, USA) and Cochrane Review Manager (RevMan version 5.3, Denmark) to perform the meta-analysis. Inter-study heterogeneity was assessed using the I2 and Cochrane Q tests. If the value was ≥ 50% and/or the P value was < 0.05, the effect estimate was considered significant for heterogeneity and a random-effects model was used. If necessary, data presented as medians were converted to the mean ± standard deviation (SD) [17]. The effect size for continuous data is expressed as the standardized mean difference (SMD; Hedges’ g) with a 95% CI. We used the Cochrane handbook recommendations to re-express SMDs using the rules of thumb for effect sizes (< 0.40 = small effect, 0.40–0.70 = moderate effect, > 0.70 = large effect) [18]. Binary research results were used to calculate the odds ratio (OR) with 95% CI using the inverse variance method (Mantel-Haenszel method). To understand the meaning of the OR in terms of changes in the number of events, we converted the OR into a risk ratio (RR) [19]. Statistical significance was set at P < 0.05 for hypothesis testing. Meta-regression analysis using the restricted maximum likelihood random-effects model was performed to assess whether the association between exposure and outcomes varied by scale type, age, sex, ASA status, postoperative delirium prevalence, type of anesthetics, operation time, and anesthesia time [20]. Covariates were first tested using a univariate model and all covariates that were available for most of the studies were assessed using a multivariate model. Due to the nature of the study design, case-control studies were excluded from the calculations on overall occurrence of ePND and mortality but were included in the meta-analysis to calculate the OR [21].

Sensitivity analysis

Sensitivity analyses were performed by analyzing only the studies with a moderate risk of bias.

Results

Baseline characteristics of the included studies

During the initial search, 920 articles were identified, and the full texts of 481 were reviewed for eligibility. A total of 16,028 patients from 22 studies were ultimately included in this systematic review and meta-analysis (Fig. 1) [2243]. The major exclusions are reported in Supplementary Table 1 (Cohen’s kappa for study selection, 0.88; for data extraction, 0.93). The overall occurrence rate of ePND was 13% (1,979 of the 15,008 total patients from 21 studies, excluding the case-control study [23]), and the overall reported mortality rate was 1.4% (33 of the 2,359 patients from two studies [28,38]).

Fig. 1.

PRISMA flow diagram for study selection. POD: postoperative delirium. A detailed list of major exclusions is presented in the Supplement.

Among the 22 included studies, 20 were prospective observational [22,24,25,2743] and eight used motor scales to assess awakening (9,277 patients, 58%) [22,23,2628,34,40,41] (Table 1). The mean ages of the patients ranged from 39 to 79 years, and the proportion of patients with an ASA status ≥ 3 varied from 0% to 55%.

Characteristics and Description of the 22 Studies included in the Meta-Analysis

Quantitative data synthesis

All-cause mortality

Among the 22 included studies, three [23,28,38] (3,379 patients) reported all-cause mortality. The all-cause mortality was 16/677 (2.4%) in the ePND group and 32/2,702 (1.2%) in the normal emergence group (OR: 2.69, 95% CI [1.25, 5.76], RR: 2.6 [1.3, 5.5], P = 0.01; I2 = 0%) (Fig. 2, Table 2).

Fig. 2.

Forest plot for mortality representing the odds ratio for the effect of normal emergence versus early postoperative neurocognitive disorders (ePND) on all-cause mortality for the included studies. The plot displays the study, sample size, odds ratio, CI, and P value. The size of the squares indicates the weight of the studies (considering sample size and standard deviations); the diamond represents the pooled OR with CI.

Outcomes and Sensitivity Analyses

Postoperative delirium

Postoperative delirium was reported in nine studies with 4,310 patients [23,24,26,30,32,34,37,41,43]. Patients with ePND, compared with those with normal emergence, had an increased incidence of postoperative delirium: 377/1,301 (29%) vs. 136/3,009 (4.5%) (OR: 15.72 [5.68, 43.53], RR: 9.5 [4.7, 14.9], P < 0.001; I2 = 93%) (Fig. 3, Table 2).

Fig. 3.

Forest plot for postoperative delirium representing the odds ratio for the effect of normal emergence versus early postoperative neurocognitive disorders (ePND) on postoperative delirium incidence for the included studies. The plot displays the study, sample size, odds ratio, CI, and P value. The size of the squares indicates the weight of the studies (considering sample size and standard deviations); the diamond represents the pooled OR with CI.

PACU length of stay

Sixteen studies (13,919 patients) [22,23,2629,31,33,34,36,3843] reported a PACU stay that was longer in patients with ePND (SMD: 0.59 [0.19, 0.98] - moderate effect size, P = 0.004; I2: 98%) (Supplementary Fig. 3, Table 2).

Hospital length of stay

Data from 12 studies and 7,452 patients [23,2527,31,33,34,36,38,40,41,43] showed that patients with ePND had a prolonged hospital stay (SMD: 0.62 [0.36, 0.88] - moderate effect size, P < 0.001; I2: 93%) (Supplementary Fig. 4, Table 2).

Publication bias and GRADE assessment

Egger’s test and the funnel plot analyses did not reveal small-study effects for the majority of the study outcomes (Table 2, Supplementary Figs. 58).

Owing to their observational nature, > 50% of studies had a high or very-high risk of bias for all studied outcomes (8/9 for postoperative delirium, 9/16 for length of PACU stay, 10/12 for length of hospital stay, and 2/3 for all-cause mortality) (Supplementary Figs. 1 and 2, Supplementary Table 2).

According to the GRADE approach, high quality evidence was only found for an increased risk of postoperative delirium in patients with ePND (Table 3).

Certainty of Evidence for Studied Outcomes (GRADE Approach)

Meta-regression

Multivariate meta-regression revealed no predictor of postoperative delirium or length of PACU stay (Supplementary Table 3); however, a meta-regression for all-cause mortality could not be performed due to the insufficient number of studies.

Discussion

Key findings

Our systematic review and meta-analysis of the available literature allowed us to document, for the first time, an increased mortality risk (RR = 2.6) in adult patients who experience ePND (very low quality of evidence). We also confirmed that ePND is associated with a 9-fold increase in postoperative delirium (high heterogeneity with I2 = 93%) and a prolonged PACU and hospital stay. However, the meta-regression analysis did not show an association between ePND and postoperative delirium.

Relationship with previous studies

Short- and long-term PNDs are associated with mortality [44]. ePND are a very early type of PND that can cause serious complications (e.g., aspiration pneumonia and hemorrhage) [28]. In addition, these complications often require physical restraints or pharmacological interventions that prolong hospitalization. Therefore, it is reasonable to assume that ePND are associated with increased mortality risk. The association between ePND and mortality has previously been suggested by Lepouse et al., who found a mortality of 4% vs. 1% in patients with and without ePND, respectively; however, the difference was not statistically significant due to the limited number of included patients [28].

Postoperative delirium is already a well-known predictor of mortality [45-47]. Our findings suggest that the presence of ePND is not a benign sign, but a harbinger of detrimental outcomes in terms of an increased risk of mortality and postoperative delirium, which warrants further prospective investigations.

Significance of study findings

Future studies should focus on prevention and treatment of ePND. Previous studies have suggested that non-pharmacological and pharmacological prevention and treatment may reduce the risk of emergence delirium in children [48]. However, evidence of this phenomenon in the adult population is lacking. Surgeons and anesthesiologists may consider these complications life-threatening and manage patients accordingly (e.g., prolonged or advanced postoperative monitoring, nonpharmacological preventive strategies for delirium, and early rehabilitation).

Strengths and limitations

Our study has some limitations. First, several studies included in the meta-analysis were classified as having a high risk of bias; however, the quality of evidence for the impact of ePND on the incidence of postoperative delirium was ‘high.’ Second, the included studies were heterogeneous from a clinical point of view (patient cohorts varied in age, sex, ASA status, type and duration of surgery, type of anesthetics, and incidence of postoperative delirium); however, this can also be considered a strength as this increases the external validity of our findings. Third, the findings of the meta-regression analyses should be interpreted with caution because of the small number of studies included in our meta-analysis [49]. Fourth, the inclusion of a case-control study [23] might have affected the strength of the association between ePND and clinical outcomes; however, baseline characteristics were matched between patients with and without emergence agitation [23], which could have enhanced the homogeneity between the two groups. Another strength of this study is the high number of included patients (16,028 patients over a 20-years period), 3,379 of which had vital status data.

Future studies and prospects

Consensus among physicians and researchers regarding the definition and classification of emergence delirium and postoperative delirium is needed. Currently, it is not clear whether they are different entities or refer to the same postoperative neurocognitive disorder occurring at different time points. Evered et al. [7] suggested combining them under the term ‘postoperative delirium’ regardless of the time of appearance, and the findings of our meta-analysis support this recommendation.

Because considerable confusion exists regarding the terms emergence agitation and emergence delirium, these terms should be defined more clearly. Most studies used emergence agitation and delirium interchangeably, while some researchers argued that emergence delirium does not exist, as it simply reflects continuation of the anesthetized state [50]. A consensus definition is necessary for the further investigation of ePND.

Our systematic review and meta-analysis showed that ePND increase the risk of all-cause mortality by 2.6 times and the risk of postoperative delirium by 9 times, suggesting that further study and clarification of the similarities and differences between ePND and postoperative delirium is needed.

Notes

Funding

None.

Conflicts of Interest

No potential conflict of interest relevant to this article was reported.

Data Availability

Availability of data: template data collection forms, data extracted from included studies, data used for all analyses on other data are available from the corresponding author on request.

Author Contributions

Valery V. Likhvantsev (Conceptualization; Writing – original draft; Writing – review & editing)

Giovanni Landoni (Conceptualization; Writing – original draft; Writing – review & editing)

Levan B. Berikashvili (Formal analysis; Writing – original draft)

Nadezhda V. Ermokhina (Formal analysis)

Mikhail Y. Yadgarov (Formal analysis; Writing – original draft)

Yuki Kotani (Formal analysis; Writing – original draft)

Kristina K. Kadantseva (Formal analysis)

Dmitry M. Makarevich (Formal analysis; Writing – original draft)

Andrey V. Grechko (Formal analysis; Writing – original draft)

Supplementary Materials

Supplementary Material 1.

Search strategy for each database.

kja-23126-Supplementary-Marterial-1.pdf
Supplementary Table 1.

Excluded studies with a cause of exclusion.

kja-23126-Supplementary-Table-1.pdf
Supplementary Table 2.

Judgement of each domain in ROBINS-E tool.

kja-23126-Supplementary-Table-2.pdf
Supplementary Table 3.

Meta-regression.

kja-23126-Supplementary-Table-3.pdf
Supplementary Fig. 1.

Risk of bias evaluation of the included trials using ROBINS-E (The Risk Of Bias In Non-randomized Studies - of Exposure) tool.

kja-23126-Supplementary-Fig-1.pdf
Supplementary Fig. 2.

Risk of bias assessment graph: proportion of studies with moderate, high or very high risk of bias for studied outcomes.

kja-23126-Supplementary-Fig-2.pdf
Supplementary Fig. 3.

Forest plot for PACU length of stay. The plot displays the study, sample size, standardized mean difference (SMD), confidence interval (CI), and P value.

kja-23126-Supplementary-Fig-3.pdf
Supplementary Fig. 4.

Forest plot for hospital length of stay. The plot displays the study, sample size, standardized mean difference (SMD), confidence interval (CI), and P value.

kja-23126-Supplementary-Fig-4.pdf
Supplementary Fig. 5.

Funnel plot for mortality.

kja-23126-Supplementary-Fig-5.pdf
Supplementary Fig. 6.

Funnel plot for postoperative delirium.

kja-23126-Supplementary-Fig-6.pdf
Supplementary Fig. 7.

Funnel plot for PACU length of stay.

kja-23126-Supplementary-Fig-7.pdf
Supplementary Fig. 8.

Funnel plot for hospital length of stay.

kja-23126-Supplementary-Fig-8.pdf
Supplementary Fig. 9.

Bubble plot depicting the univariate meta regression of SMD (Hedges’g) with anesthesia time (for association between ePND and PACU LoS).

kja-23126-Supplementary-Fig-9.pdf

References

1. Weiser TG, Haynes AB, Molina G, Lipsitz SR, Esquivel MM, Uribe-Leitz T, et al. Estimate of the global volume of surgery in 2012: an assessment supporting improved health outcomes. Lancet 2015;385 Suppl 2:S11.
2. Moonesinghe SR, Harris S, Mythen MG, Rowan KM, Haddad FS, Emberton M, et al. Survival after postoperative morbidity: a longitudinal observational cohort study. Br J Anaesth 2014;113:977–84.
3. Archer S, Pinto A, Vuik S, Bicknell C, Faiz O, Byrne B, et al. Surgery, complications, and quality of life: a longitudinal cohort study exploring the role of psychosocial factors. Ann Surg 2019;270:95–101.
4. Patel AS, Bergman A, Moore BW, Haglund U. The economic burden of complications occurring in major surgical procedures: a systematic review. Appl Health Econ Health Policy 2013;11:577–92.
5. Mashour GA, Woodrum DT, Avidan MS. Neurological complications of surgery and anaesthesia. Br J Anaesth 2015;114:194–203.
6. Aldecoa C, Bettelli G, Bilotta F, Sanders RD, Audisio R, Borozdina A, et al. European Society of Anaesthesiology evidence-based and consensus-based guideline on postoperative delirium. Eur J Anaesthesiol 2017;34:192–214.
7. Evered L, Silbert B, Knopman DS, Scott DA, DeKosky ST, Rasmussen LS, et al. Recommendations for the nomenclature of cognitive change associated with anaesthesia and surgery-2018. Anesthesiology 2018;129:872–9.
8. World Health Organization. The ICD-10 classification of mental and behavioural disorders: clinical descriptions and diagnostic guidelines [Internet]. Geneva: World Health Organization; 1992 [cited 2023 Feb 19]. Available from https://apps.who.int/iris/handle/10665/37958.
9. The Lancet. ICD-11. Lancet 2019;393:2275.
10. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med 2009;6e1000100.
11. Diagnostic and statistical manual of mental disorders: DSM-5 5th edth ed. Washington DC: American Psychiatric Publishing, Inc; 2013. p. 596–852.
12. Ely EW, Inouye SK, Bernard GR, Gordon S, Francis J, May L, et al. Delirium in mechanically ventilated patients: validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA 2001;286:2703–10.
13. Gaudreau JD, Gagnon P, Harel F, Tremblay A, Roy MA. Fast, systematic, and continuous delirium assessment in hospitalized patients: the nursing delirium screening scale. J Pain Symptom Manage 2005;29:368–75.
14. ROBINS-E Development Group. Risk Of Bias In Non-randomized Studies - of Exposure (ROBINS-E) [Internet]. Launch version, 2022 Jun [cited 2023 Feb 19]. Available from https://www.riskofbias.info/welcome/robins-e-tool.
15. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34.
16. Foroutan F, Guyatt G, Zuk V, Vandvik PO, Alba AC, Mustafa R, et al. GRADE guidelines 28: use of GRADE for the assessment of evidence about prognostic factors: rating certainty in identification of groups of patients with different absolute risks. J Clin Epidemiol 2020;121:62–70.
17. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol 2005;5:13.
18. Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al. Cochrane handbook for systematic reviews of interventions [Internet]. London: Cochrane; 2022 Feb [cited 2023 Feb 19]. Available from https://training.cochrane.org/handbook/current.
19. Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al. Cochrane handbook for systematic reviews of interventions version 6.3 [Internet]. London: Cochrane; 2022 Feb [cited 2023 Feb 19]. Available from www.training.cochrane.org/handbook.
20. Harbord RM, Higgins JP. Meta-regression in Stata. Stata J 2008;8:493–519.
21. Irony TZ. Case-control studies: using “real-world” evidence to assess association. JAMA 2018;320:1027–8.
22. Assefa S, Sahile WA. Assessment of magnitude and associated factors of emergence delirium in the post anesthesia care unit at Tikur Anbesa Specialized Hospital, Ethiopia. Ethiop J Health Sci 2019;29:597–604.
23. Fields A, Huang J, Schroeder D, Sprung J, Weingarten T. Agitation in adults in the post-anaesthesia care unit after general anaesthesia. Br J Anaesth 2018;121:1052–8.
24. Gutiérrez RG, Reyes FI, Penna A. Delirium in the post-anesthesia care unit may be associated with the development of postoperative delirium in a cohort of elderly patients. Minerva Anestesiol 2019;85:326–7.
25. Hesse S, Kreuzer M, Hight D, Gaskell A, Devari P, Singh D, et al. Association of electroencephalogram trajectories during emergence from anaesthesia with delirium in the postanaesthesia care unit: an early sign of postoperative complications. Br J Anaesth 2019;122:622–34.
26. Huang J, Qi H, Lv K, Chen X, Zhuang Y, Yang L. Emergence delirium in elderly patients as a potential predictor of subsequent postoperative delirium: a descriptive correlational study. J Perianesth Nurs 2020;35:478–83.
27. Kim HC, Kim E, Jeon YT, Hwang JW, Lim YJ, Seo JH, et al. Postanaesthetic emergence agitation in adult patients after general anaesthesia for urological surgery. J Int Med Res 2015;43:226–35.
28. Lepousé C, Lautner CA, Liu L, Gomis P, Leon A. Emergence delirium in adults in the post-anaesthesia care unit. Br J Anaesth 2006;96:747–53.
29. Munk L, Andersen G, Møller AM. Post-anaesthetic emergence delirium in adults: incidence, predictors and consequences. Acta Anaesthesiol Scand 2016;60:1059–66.
30. Neufeld KJ, Leoutsakos JM, Sieber FE, Wanamaker BL, Gibson Chambers JJ, Rao V, et al. Outcomes of early delirium diagnosis after general anesthesia in the elderly. Anesth Analg 2013;117:471–8.
31. Olbert M, Eckert S, Mörgeli R, Kruppa J, Spies CD. Validation of 3-minute diagnostic interview for CAM-defined Delirium to detect postoperative delirium in the recovery room: a prospective diagnostic study. Eur J Anaesthesiol 2019;36:683–7.
32. Sharma PT, Sieber FE, Zakriya KJ, Pauldine RW, Gerold KB, Hang J, et al. Recovery room delirium predicts postoperative delirium after hip-fracture repair. Anesth Analg 2005;101:1215–20.
33. Radtke FM, Franck M, Schneider M, Luetz A, Seeling M, Heinz A, et al. Comparison of three scores to screen for delirium in the recovery room. Br J Anaesth 2008;101:338–43.
34. Radtke FM, Franck M, MacGuill M, Seeling M, Lütz A, Westhoff S, et al. Duration of fluid fasting and choice of analgesic are modifiable factors for early postoperative delirium. Eur J Anaesthesiol 2010;27:411–6.
35. Radtke FM, Franck M, Hagemann L, Seeling M, Wernecke KD, Spies CD. Risk factors for inadequate emergence after anesthesia: emergence delirium and hypoactive emergence. Minerva Anestesiol 2010;76:394–403.
36. Saller T, MacLullich AM, Schäfer ST, Crispin A, Neitzert R, Schüle C, et al. Screening for delirium after surgery: validation of the 4 A's test (4AT) in the post-anaesthesia care unit. Anaesthesia 2019;74:1260–6.
37. Wang CM, Huang HW, Wang YM, He X, Sun XM, Zhou YM, et al. Incidence and risk factors of postoperative delirium in patients admitted to the ICU after elective intracranial surgery: a prospective cohort study. Eur J Anaesthesiol 2020;37:14–24.
38. Wiinholdt D, Eriksen SA, Harms LB, Dahl JB, Meyhoff CS. Inadequate emergence after non-cardiac surgery-A prospective observational study in 1000 patients. Acta Anaesthesiol Scand 2019;63:1137–42.
39. Winter A, Steurer MP, Dullenkopf A. Postoperative delirium assessed by post anesthesia care unit staff utilizing the Nursing Delirium Screening Scale: a prospective observational study of 1000 patients in a single Swiss institution. BMC Anesthesiol 2015;15:184.
40. Wu J, Gao S, Zhang S, Yu Y, Liu S, Zhang Z, et al. Perioperative risk factors for recovery room delirium after elective non-cardiovascular surgery under general anaesthesia. Perioper Med (Lond) 2021;10:3.
41. Xará D, Silva A, Mendonça J, Abelha F. Inadequate emergence after anesthesia: emergence delirium and hypoactive emergence in the postanesthesia care unit. J Clin Anesth 2013;25:439–46.
42. Yu D, Chai W, Sun X, Yao L. Emergence agitation in adults: risk factors in 2,000 patients. Can J Anaesth 2010;57:843–8.
43. Zhang Y, He ST, Nie B, Li XY, Wang DX. Emergence delirium is associated with increased postoperative delirium in elderly: a prospective observational study. J Anesth 2020;34:675–87.
44. Tasbihgou SR, Absalom AR. Postoperative neurocognitive disorders. Korean J Anesthesiol 2021;74:15–22.
45. Abelha FJ, Luís C, Veiga D, Parente D, Fernandes V, Santos P, et al. Outcome and quality of life in patients with postoperative delirium during an ICU stay following major surgery. Crit Care 2013;17:R257.
46. Robinson TN, Raeburn CD, Tran ZV, Angles EM, Brenner LA, Moss M. Postoperative delirium in the elderly: risk factors and outcomes. Ann Surg 2009;249:173–8.
47. Moskowitz EE, Overbey DM, Jones TS, Jones EL, Arcomano TR, Moore JT, et al. Post-operative delirium is associated with increased 5-year mortality. Am J Surg 2017;214:1036–8.
48. Dahmani S, Delivet H, Hilly J. Emergence delirium in children: an update. Curr Opin Anaesthesiol 2014;27:309–15.
49. Geissbühler M, Hincapié CA, Aghlmandi S, Zwahlen M, Jüni P, da Costa BR. Most published meta-regression analyses based on aggregate data suffer from methodological pitfalls: a meta-epidemiological study. BMC Med Res Methodol 2021;21:123.
50. Tolly B, Waly A, Peterson G, Erbes CR, Prielipp RC, Apostolidou I. Adult emergence agitation: a veteran-focused narrative review. Anesth Analg 2021;132:353–64.

Article information Continued

Fig. 1.

PRISMA flow diagram for study selection. POD: postoperative delirium. A detailed list of major exclusions is presented in the Supplement.

Fig. 2.

Forest plot for mortality representing the odds ratio for the effect of normal emergence versus early postoperative neurocognitive disorders (ePND) on all-cause mortality for the included studies. The plot displays the study, sample size, odds ratio, CI, and P value. The size of the squares indicates the weight of the studies (considering sample size and standard deviations); the diamond represents the pooled OR with CI.

Fig. 3.

Forest plot for postoperative delirium representing the odds ratio for the effect of normal emergence versus early postoperative neurocognitive disorders (ePND) on postoperative delirium incidence for the included studies. The plot displays the study, sample size, odds ratio, CI, and P value. The size of the squares indicates the weight of the studies (considering sample size and standard deviations); the diamond represents the pooled OR with CI.

Table 1.

Characteristics and Description of the 22 Studies included in the Meta-Analysis

Study Design ePND criteria Scale type Sample size, ePND (%) Mean age (yr) Sex, male (%) ASA≥ 3, (%) Volatile anesthetics (% of patients) Operation time (min) Anesthesia time (min)
Assefa and Sahile (2019) [22] PO RASS ≥ +1 Motor 306, 97 (32%) 44 47 2.6 N/D 65.5 N/D
Fields et al. (2018) [23]* Matched case-control RASS score of +3 or +4 or the administration of haloperidol Motor 1020, 510 (50%) 63 73 N/D 44 138 N/D
Gutiérrez et al. (2019) [24] PO CAM Cognitive 93, 6 (6%) 73 39 N/D N/D N/D N/D
Hesse et al. (2019) [25] PO CAM-ICU Cognitive 626, 125 (20%) 59 60 43 98 N/D 116
Huang et al. (2020) [26] PO Nu-DESC Cognitive 168, 58 (35%) 72 53 55 100 141 N/D
Kim et al. (2015) [27] RO SAS ≥ 5 Motor 488, 48 (10%) 62 82 6.2 92 109 146
Lepouse et al. (2006) [28] PO SAS ≥ 5 Motor 1359, 64 (5%) 51 54 21 64 76 N/D
Munk et al. (2016) [29] PO RASS ≥ +1 Motor 1970, 73 (4%) 60 46 15 11 80 147
Neufeld et al. (2013) [30] PO DSM-IV Cognitive 91, 41 (45%) 79 42 N/D N/D 186 N/D
Olbert et al. (2019) [31] PO DSM-V Cognitive 176, 16 (9%) 69 54 30 N/D N/D 130
Sharma et al. (2005) [32] PO CAM Cognitive 47, 21 (45%) 77 34 2 100 145 N/D
Radtke et al. (2008) [33] PO DSM-IV Cognitive 154, 21 (14%) 54 60 25 55 75 N/D
Radtke et al. (2010) [34] PO Nu-DESC Cognitive 910, 100 (11%) 53 53 26 53 79 N/D
Radtke et al. (2010) [35] PO RASS ≥ +1 Motor 1868, 93 (5%) 52 51 29 61 84 N/D
Saller et al. (2019) [36] PO DSM-V Cognitive 543, 22 (4%) 52 55 21 41 97 N/D
Wang et al. (2020) [37] PO CAM-ICU Cognitive 800, 209 (26%) 48 41 3.5 N/D 261 N/D
Wiinholdt et al. (2019) [38] PO Nu-DESC Cognitive 1000, 103 (10%) 60 48 20 72 N/D 141
Winter et al. (2015) [39] PO Nu-DESC ˃ 2 Cognitive 1000, 43 (4%) 58 40 9.3 N/D 93 166
Wu et al. (2021) [40] PO CAM-ICU Cognitive 228, 57 (25%) 39 43 13 82 N/D N/D
Xará et al. (2013) [41] PO RASS ≥ +1 Motor 266, 17 (6%) 56 40 16 N/D 90 N/D
Yu et al. (2010) [42] PO Three point scale Motor 2000, 426 (21%) 42 43 0 68 N/D N/D
Zhang et al. (2020) [43] PO СAM-ICU Cognitive 915, 339 (37%) 71 60 0 31 203 N/D

*This case-control study was excluded from the occurrence rate of ePND and mortality rate calculations of the included studies overall. ePND: early postoperative neurocognitive disorders, SAS: Riker Sedation–Agitation Scale, ASA: American Society of Anesthesiologists, RASS: Richmond Agitation-Sedation Scale, CAM: Confusion Assessment Method, CAM-ICU: Confusion Assessment Method for the Intensive Care Unit, Nu-DESC: Nursing-Delirium Screening Scale, DSM: Diagnostic and Statistical Manual of Mental Disorders, RO: retrospective observational, PO: prospective observational, N/D: no data.

Table 2.

Outcomes and Sensitivity Analyses

Outcome Trials included ePND (n) No ePND (n) SMD/OR 95% CI P value for overall effect P value for heterogeneity I2 (%) P value for publication bias (Egger’s test)
All-cause mortality 3 677 2,702 2.7 1.3, 5.8 0.01* 0.95 0 0.81
Postoperative delirium 9 1,301 3,009 15.7 5.7, 43.5 < 0.001* < 0.001 93 0.052
PACU length of stay 16 2,054 11,865 0.6 0.2, 1.0 0.004* < 0.001 98 0.48
All studies with moderate bias 7 1,050 4,204 0.4 0.1, 0.7 0.006* < 0.001 90 0.62
Hospital length of stay 12 1,409 6,043 0.6 0.4, 0.9 < 0.001* < 0.001 93 0.16

ePND: early postoperative neurocognitive disorder, OR: odds ratio, SMD: standardized mean difference, PACU: post anesthesia care unit. *Significant effect, P < 0.05.

Table 3.

Certainty of Evidence for Studied Outcomes (GRADE Approach)

Outcome No. of participants and studies Risk of bias Inconsistency Indirectness Imprecision Publication bias Upgrades Overall quality of evidence
Mortality at longest follow-up 3,379, Very serious (−2) Very serious (−2) Not serious (0) Serious (−1) Not serious (0) Large effect OR > 2 (+1) ⊕ΟΟΟ
Statement: ePND is associated with increased odds of mortality compared to normal emergence. OR: 2.69, 95% CI [1.25, 5.76], RR: 2.64 3 studies Very low
Postoperative delirium 4,310, Very serious (−2) Serious (−1) Not serious (0) Not serious (0) Not serious (0) Large effect OR > 5 (+2) ⊕⊕⊕⊕
Statement: ePND is associated with increased odds of POD compared to normal emergence. OR: 15.7, 95% CI [5.68, 43.53], RR: 9.4 9 studies Plausible confounding* (+1) High
PACU LoS 13,919, Very serious (−2) Serious (−1) Not serious (0) Serious (−1) Not serious (0) None ⊕ΟΟΟ
Statement: ePND is associated with increased PACU stay compared to normal emergence. SMD: 0.59, 95% CI [0.19, 0.98] 16 studies Very low
PACU LoS 5,254, Not serious (0) Serious (−1) Not serious (0) Serious (−1) Not serious (0) None ⊕⊕ΟΟ
Statement: Studies with moderate risk of bias suggest that ePND is associated with increased PACU stay compared to normal emergence. SMD: 0.40, 95% CI [0.12, 0.68] 7 studies Low
Hospital LoS 7,452, Very serious (−2) Very serious (−2) Not serious (0) Serious (−1) Not serious (0) None ⊕ΟΟΟ
Statement: ePND is associated with increased hospital stay compared to normal emergence. SMD: 0.62, 95% CI [0.36, 0.87] 12 studies Very low

ePND: early postoperative neurocognitive disorder, ED: emergence delirium, EA: emergence agitation, POD: postoperative delirium, OR: odds ratio, SMD: standardized mean difference, PACU: post anesthesia care unit, LoS: length of stay. Key; 0: no evidence downgrade, −1: serious limitation, −2: very serious limitation, +1: evidence upgrade. Baseline evidence level for studies of prognostic factors: high. *Therapy for early postoperative neurocognitive disorders in the ePND group could potentially decrease POD rates.