Association Between Perioperative Hypotension and Delirium in Postoperative Critically Ill Patients: A Retrospective Cohort Analysis : Anesthesia & Analgesia

Secondary Logo

Journal Logo

Advanced Search
Featured Articles: Original Clinical Research Report

Association Between Perioperative Hypotension and Delirium in Postoperative Critically Ill Patients: A Retrospective Cohort Analysis

Maheshwari, Kamal MD*,†; Ahuja, Sanchit MD*,‡; Khanna, Ashish K. MD, FCCP, FCCM§; Mao, Guangmei PhD, MPH*,‖; Perez-Protto, Silvia MD, MS*,†; Farag, Ehab MD, FRCA, FASA*,†; Turan, Alparslan MD*,†; Kurz, Andrea MD*,†; Sessler, Daniel I. MD*

Author Information

From the *Department of Outcomes Research, Cleveland Clinic, Cleveland, Ohio

†Department of General Anesthesiology, Cleveland Clinic, Cleveland, Ohio

‡Department of Anesthesiology, Pain Management and Perioperative Medicine, Henry Ford Health System, Detroit, Michigan

§Wake Forest Center for Biomedical Informatics and the Critical Injury Illness and Recovery Research Center, Outcomes Research Consortium, Cleveland, Ohio

‖Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, Ohio.

Published ahead of print 11 November, 2019.

Accepted for publication September 30, 2019.

Funding: None.

The authors declare no conflicts of interest.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website.

Reprints will not be available from the authors.

Address correspondence to Kamal Maheshwari, MD, Department of General Anesthesiology, Anesthesiology Institute, Cleveland Clinic, 9500 Euclid Ave, E-31, Cleveland, OH 44195. Address e-mail to [email protected].

Anesthesia & Analgesia 130(3):p 636-643, March 2020. | DOI: 10.1213/ANE.0000000000004517

Abstract

See Article, p 635

KEY POINTS

  • Question: What is the relationship between perioperative hypotension and delirium in critically ill noncardiac surgery patients?
  • Findings: We observed significant associations between intraoperative and postoperative hypotension with delirium in critically ill noncardiac surgery patients.
  • Meaning: Perioperative hypotension reduction may help reduce delirium in critically ill noncardiac surgery patients.

Delirium is common in surgical intensive care units (ICUs) and is associated with significant morbidity, mortality, and cost.1–3 Postoperative delirium can occur within a few days after surgery and has a reported incidence ranging from 11% to 43%.4,5 The pathophysiology of delirium is largely unknown, and proposed mechanisms involve increase in inflammatory mediators due to surgical stress and cerebral hypoxia caused due to hypoperfusion.6 Blood pressure is an important determinant of organ perfusion, and existing evidence indicates that intraoperative hypotension is associated with perioperative organ injury such as heart, kidney, and brain.7–11 Low blood pressure and resultant hypoperfusion of the cortex and subcortex may cause delirium.12 Specifically, perfusion abnormalities in frontal and parietal lobes are associated with delirium symptoms.13–15 Thus, avoiding hypotension may help maintain optimal cerebral perfusion and reduce delirium risk.16

While major risk factors for delirium include advanced age, baseline cognitive dysfunction, and high-risk surgery, critically ill postoperative surgical patients are especially at high risk.1,17,18 Most of the risk factors are not modifiable. In contrast, both intraoperative and postoperative hypotension are common and amenable to intervention. For example, hypotension in hip fracture patients is associated with postoperative delirium.19,20 We, therefore, sought to determine the association between perioperative hypotension and postoperative delirium. Specifically, we studied the associations between intraoperative and postoperative hypotension and delirium as assessed by Confusion Assessment Method Intensive Care Unit (CAM-ICU) scale.

METHODS

After Cleveland Clinic Institutional Review Board (IRB) approval (IRB 17-812; Principal Investigator: K.M.; date of registration: June 2017) and waived requirement of written informed consent, data from eligible patients were obtained from the Anesthesia Institute’s Perioperative Health Documentation System database and electronic medical record system. This analysis adheres to the guidelines of Strengthening the Reporting of Observational Studies in Epidemiology.

Study Population

We included adults who had noncardiac surgery from 2009 to 2016 at the Cleveland Clinic main campus, who were admitted to the surgical ICU within 8 hours after surgery and remained in the critical care unit for at least 48 hours. Patients admitted to the surgical ICU from the regular nursing floor were not included because the nature and pathophysiology of hypotension and delirium may differ in such patients.

We excluded patients with preexisting dementia or cognitive impairment, neurological surgery, <1 CAM-ICU assessment per ICU day (details below), or inadequate blood pressure measurements defined as <1 measurement per hour or reading gaps longer than 2 hours while in critical care. We also excluded patients with missing baseline variables.

Measurements

We recorded baseline patient’s characteristics including age, sex, race, American Society of Anesthesiologists (ASA) physical status, Charlson comorbidity index, and disease history. Surgical and intraoperative details were also recorded. The exposure variables were mean arterial pressure (MAP) during surgery and during surgical ICU stay. Per routine, blood pressure in ICU is measured frequently and verified by nurse every hour. Only the nurse-verified pressures are recorded. Among our final population, on an average, 2 blood pressure readings were available per patient per hour.

Intraoperative hypotension exposure, time-weighted average (TWA) MAP <65 mm Hg, was calculated as area under the curve (AUC) of MAP <65 mm Hg divided by surgery duration. We originally planned to use TWA MAP <75 mm Hg as our measure for ICU hypotension. However, only 20% of patients had any MAP <75 mm Hg during ICU stay, meaning that about 80% of patients had TWA MAP <75 mm Hg as zero. We, therefore, instead used the lowest MAP to characterize postoperative ICU hypotension as described by Khanna et al.21 The lowest MAP on each day, during the initial 5 ICU days or until delirium was first diagnosed, was summarized as a time-dependent variable.

Delirium was assessed per routine by nurses who were trained to use the CAM-ICU scale at 12-hour intervals while patients remained in ICU. CAM-ICU is a clinically valid and reliable tool with high interrater reliability22 and is a good screening and confirmatory test.23,24 Delirium was defined as a positive CAM-ICU assessment which is accompanied by a Richmond Agitation-Sedation Scale (RASS) of −3 or greater.25 However, assessments on the day of surgery were excluded from analysis because they might be influenced by residual anesthetic effects.

Statistical Analysis

To assess the association between intraoperative MAP and delirium in ICU, we used a Cox proportional hazard survival model with TWA MAP <65 mm Hg and all confounding variables listed in Table 1 as fixed-effect covariates. For patients with delirium in the ICU within 5 days after surgery, we used the earliest time of delirium as their event time and assigned their censor variables as zero; patients without delirium and discharged from ICU in <5 days after surgery were coded as having competing risk at their discharge day; and patients who did not experience delirium and still in ICU after 5 days were censored at day 5. The cause-specific hazard ratio (HR) of TWA MAP <65 mm Hg was reported for the association between intraoperative hypotension and delirium.

Table 1. - Patient Characteristics and Surgery Information
Factor Total (N = 1083) Without Delirium (N = 706) With Delirium (N = 377)
Age, y 62 ± 14 61 ± 14 64 ± 13
Sex, female, n (%) 499 (46) 320 (45) 179 (47)
Race, Caucasian, n (%) 897 (83) 583 (83) 314 (83)
ASA status, n (%)
 I or II 37 (3) 26 (4) 11 (3)
 III 507 (47) 338 (48) 169 (45)
 IV 534 (49) 339 (48) 195 (52)
 V 5 (0.5) 3 (0.4) 2 (0.5)
Charlson comorbidity index 3 (2–5) 3 (1–6) 3 (2–5)
History of, n (%)
 Depression 173 (16) 103 (15) 70 (19)
 Alcohol abuse 67 (6) 35 (5) 32 (8)
 Vision impairment 25 (2) 14 (2) 11 (3)
 Hearing impairment 24 (2) 15 (2) 9 (2)
 Use of benzodiazepines 325 (30) 185 (26) 140 (37)
Surgery duration, h 8 (5–11) 8 (5–11) 7 (4–11)
Emergency surgery, n (%) 227 (21) 122 (17) 105 (28)
Surgery type, n (%)
 Organ transplant 217 (20) 138 (20) 79 (21)
 On digestive system 308 (28) 190 (27) 118 (31)
 On musculoskeletal system 237 (22) 160 (23) 77 (20)
 Other 321 (30) 218 (31) 103 (27)
Blood loss, mL 600 (200–2000) 700 (250–2000) 500 (100–2000)
Transfusion, mL 621 (0–1946) 618 (0–1716) 626 (0–2296)
Intraoperative opioid as IV morphine equivalent, mg 30 (19–60) 35 (20–63) 25 (15–52)
Values presented as mean ± SD, median (Q1, Q3), or N (column %).
Abbreviations: ASA, American Society of Anesthesiologists; IV, intravenous; Q, quartile; SD, standard deviation.

To assess the association between postoperative ICU hypotension and delirium in ICU, we used a Cox proportional hazard survival model with the lowest MAP on each day as a time-varying risk factor, adjusted for TWA MAP <65 mm Hg, all confounding variables listed in Table 1 and mechanical ventilation on arrival to ICU as fixed-effect covariates. Considering that the relationship between lowest MAP and delirium can be nonlinear, we would add a quadratic term of lowest MAP in the model if the added term was significantly associated with the outcome or could improve model fitting based on the Akaike information criterion (AIC). Otherwise, we would report the linear relationship between the lowest MAP and the HR of delirium. When assessing the association between postoperative hypotension and delirium, patients who have delirium earlier during the ICU stay were less likely to experience low blood pressure, since exposure periods were shorter than for patients without delirium.26 To address this time-related bias, we used the lowest MAP on each day during ICU stay as a time-varying covariate in a Cox proportional hazard survival model.27 Patients usually get better because they are discharged alive from ICU; thus, ICU discharge cannot be used as a noninformative censor in a survival analysis.28

Sample Size Calculation

Post hoc, with about 380 patients having delirium and a standard deviation of intraoperative TWA MAP <65 around 1.3, we had 80% power to detect a HR of 1.12 or larger. We used a significance criterion of 0.05 and a corresponding 95% confidence interval (CI) for each analysis. All tests were 2-sided. Statistical Analysis System (SAS) software version 9.4 (SAS Institute, Cary, NC) was used for all analyses.

RESULTS

A total of 1083 patients who had surgeries at Cleveland Clinic main campus from 2009 to 2016 met all inclusion and exclusion criteria (Figure 1). A total of 377 (35%) had delirium assessed by CAM-ICU within 5 days after surgery. Patient baseline characteristics, comorbidities, history of disease, and surgery-related information were summarized in Table 1 and Supplemental Digital Content, Table 1, https://links.lww.com/AA/C957, split by delirium status. We also summarized available postoperative factors by delirium status in Table 2 and reported the standard differences. Patients with delirium had higher Acute Physiology and Chronic Health Evaluation (APACHE) score at ICU admission than patients without it, while the opioid consumption and use of benzodiazepines were similar in 2 groups.

Table 2. - Summary of Postoperative Factors in ICU
Without Delirium (N = 706) With Delirium (N = 377)
Factor Nmiss Summary Nmiss Summary
Mechanical ventilation on arrival to ICU, n (%) 0 544 (77) 0 306 (81)
APACHE score 38 56 ± 21 31 66 ± 24
First pH 203 7.34 ± 0.08 160 7.32 ± 0.08
Fist anion gap, mEq/L 91 12.7 ± 3.9 147 13.1 ± 4.2
First Na, mEq/L 91 140 ± 4 147 140 ± 4
Opioid use per hour, mg as IV morphine equivalent 0 1.9 (0.9–4.1) 0 1.9 (0.7–4.6)
Use of benzodiazepines 0 105 (15) 0 42 (11)
Anemia in ICU 89 344 (56) 147 128 (56)
BUN/creatinine ratio >18 14 396 (57) 21 192 (54)
Use of propofol 0 568 (81) 0 334 (89)
Use of dexmedetomidine 0 68 (10) 0 165 (44)
Summary statistics were presented as mean ± SD, median (Q1–Q3), or N (column %). An absolute standardized difference smaller than 0.1 is considered to be balanced between 2 groups.
Abbreviations: APACHE, acute physiology and chronic health evaluation; BUN, blood urea nitrogen; ICU, intensive care unit; IV, intravenous; Na, sodium level; Nmiss, number of missing data; Q, quartile; SD, standard deviation.

F1
Figure 1.:
Study flow chart. CAM-ICU indicates Confusion Assessment Method Intensive Care Unit; ICU, intensive care unit; SICU, surgical intensive care unit.

Intraoperative hypotension as measured by median TWA MAP <65 mm Hg, which is AUC MAP <65 mm Hg per hour, was 0.39 mm Hg (quartile [Q]1 = 0.09, Q3 = 0.96) among patients without delirium and 0.47 mm Hg (Q1 = 0.12, Q3 = 1.31) among patients with delirium (Table 3). One mm Hg increase in TWA MAP <65 mm Hg was significantly associated with an increased cause-specific hazard of delirium (HR, 1.11; 95% CI, 1.03–1.20; P = .009) shown in Table 3, adjusted for all variables listed in Table 1. We did sensitivity analysis using AUC MAP <65 mm Hg to quantify intraoperative hypotension. Consistent with previous results, 1 hour × mm Hg increase in AUC MAP <65 mm Hg was associated with an increased hazard of delirium (HR, 1.01; 95% CI, 1.00–1.02; P = .033) (Figure 2).

F2
Figure 2.:
Adjusted hazard ratio of delirium in intensive care unit. The circle represents the estimated hazard ratio, and the line represents the 95% CI. AUC indicates area under the curve; CI, confidence interval; ICU, intensive care unit; MAP, mean arterial pressure; TWA, time-weighted average.

The median lowest MAP in ICU on each day was 86 mm Hg (Q1 = 77 and Q3 = 98). We included only a linear term of lowest MAP in the model because adding a quadratic term did not improve model fitting (no decrease in AIC), and the quadratic term was not associated with the hazard of delirium. Considering that the intraoperative hypotension could be a potential confounder for the association between postoperative ICU hypotension and delirium, we added TWA MAP <65 mm Hg in adjustment besides all confounding variables in Table 1 and mechanical ventilation on arrival to ICU. Moreover, the correlation between TWA MAP <65 and ICU lowest MAP was low (Pearson correlation, −0.15), indicating collinearity is not an issue in our model. Estimated from a survival model with lowest MAP on each day as a time-varying covariate, 10 mm Hg decrease in lowest MAP was significantly associated with an increased HR of delirium in ICU (HR, 1.12; 95% CI, 1.04–1.20; P = .003) as represented in Table 3.

Table 3. - Association of Intraoperative and Postoperative MAP to Delirium in ICU
Summary Hazard Ratio (95% CI)Table 3. P Value
Without Delirium With Delirium
Intraoperative period 1 mm Hg higher
TWA of MAP <65 mm Hg 0.39 (0.09, 0.96) 0.47 (0.12, 1.31) 1.11 (1.03–1.20) .009Table 3.
ICU stay 10 mm Hg lower
Lowest MAP on each day 86 (77, 97) 83 (74, 95) 1.12 (1.04–1.20) .003Table 3.
Summary statistics were presented as median (Q1, Q3).
Abbreviations: CI, confidence interval; ICU, intensive care unit; MAP, mean arterial pressure; Q, quartile; TWA, time-weighted average.
aThe hazard ratios for intraoperative TWA of MAP <65 mm Hg were estimated using Cox proportional hazard model, adjusted for all potential confounders listed in Table 1 and Supplemental Digital Content, Table 1, https://links.lww.com/AA/C957. The hazard ratios for 10 mm Hg difference in the lowest MAP on each day were estimated using time-varying covariate Cox proportional hazard model, adjusted for intraoperative TWA of MAP <65 mm Hg and other confounders.
bThe association is statistically significant.

We did a sensitivity analysis by defining postoperative ICU hypotension as any MAP <75 mm Hg on each day during the ICU stay. There were 395 (43.5%) patients who experienced MAP <75 mm Hg during the observation period. Consistent with primary analysis, we found that postoperative ICU hypotension was significantly associated with higher hazard of delirium. The adjusted HR of delirium comparing patient with any MAP <75 mm Hg to patients without it was 1.31 (95% CI, 1.05–1.65; P = .023; Figure 2).

DISCUSSION

The incidence of delirium was 35% during the first 5 postoperative days which is consistent with previous reports.4,5 Confirming our hypothesis, intraoperative and ICU hypotension were independently associated with delirium. Physicians generally try to keep blood pressure above various empirical thresholds, with an assumption that adequate cerebral blood flow is maintained through at least a reasonable degree of hypotension, which may not be true.

For intraoperative hypotension analysis, we used MAP <65 mm Hg as our exposure based on our previous work that identifies this threshold to be associated with myocardial injury,8,10 kidney injury,8,9 and death.7 A randomized trial (N = 92) involving patients who had cardiac surgeries demonstrated increased postoperative delirium in patients who were assigned to relative hypotension MAP 60–70 mm Hg compared to normal MAP 80–90 mm Hg, supports our chosen threshold.29 In contrast, a prospective observational study (N = 594) of noncardiac surgery patients >65 years of age did not identify a relationship between delirium and hypotension (MAP <50 mm Hg).30 However, they measured delirium only once daily for just the first 2 postoperative days and may, therefore, have missed delirium in some patients. We included critically ill patients who are at high risk of delirium and followed them for up to 5 days after surgery with twice daily delirium assessment, reducing the chances of a false-negative outcome assessment.

For postoperative ICU hypotension analysis, the lowest MAP on each day was associated with an increased hazard of delirium. We used lowest MAP in contrast to cumulative exposure below a threshold because the incidence of hypotension was lower than previously reported; only 20% of patients had any MAP <75 mm Hg in our surgical ICU patients. We believe that this is due to our focus on surgical patients in contrast to most ICU studies with mixed population of medical, neurological, cardiac, or surgical patients. In surgical patients, the cardiorespiratory issues are usually optimized before elective surgery. In addition, perioperative stress and pain may promote comparatively higher blood pressures in contrast to septic ICU patients. Our methodology and the finding are consistent with Khanna et al21 who reported similar amount of hypotension and that increasing amounts of hypotension, defined by lowest MAPs per day, were strongly associated with myocardial injury, mortality, and renal injury in postoperative critical care patients.21 Furthermore, in patients with septic shock, resuscitation with higher MAP target (80 and 85 mm Hg) was associated with higher arousal level than a MAP target between 65 and 70 mm Hg, supporting our finding that hypotension may increase the risk of delirium.31

One of the other challenges in our analysis was to disentangle the contributions of intraoperative and postoperative hypotension on delirium because the hypotension in each period is presumably correlated.32 Events happening before the exposure of interest should be considered confounders, whereas events happening after the exposure and in the causal pathway should be considered mediators.33 In our analysis, the correlation between intraoperative hypotension and ICU lowest MAP was low (Pearson correlation, 0.26). While we report a significant association between postoperative ICU hypotension and delirium after adjusting for intraoperative hypotension, the strength of the relationship may be somewhat underestimated.

Our study is limited in being a retrospective analysis and therefore suffers from some degree of confounding and bias. The patients who experienced delirium were elderly (mean age, 64 years) and possibly had other comorbidities that may affect delirium. However, we did adjust for known confounders. While we included patients undergoing diverse surgical procedures and had good power for analysis, this investigation was a single-center study and the results may not be generalizable. An important limitation is that we did not study various types of delirium during assessment that have different pathophysiology. A prospective trial is needed to confirm the relationship between different types of delirium and hypotension. We did not include delirium on the day of surgery and possibly missed some patients with hypoactive delirium, which led to an underestimation of incidence of delirium. Future studies should use specific tools to identify various types of delirium and their pathophysiology. We used cumulative exposure for intraoperative analysis and lowest MAP per day for postoperative analysis because there were different amount of hypotension in the 2 periods. Both the threshold to define hypotension and the method chosen to model hypotension may affect the association of hypotension with outcome.34 Last, this study was not designed to identify a specific optimal threshold of blood pressure in operating rooms or the ICU. However, this study highlights that increasing amount of intraoperative and postoperative hypotension is associated with delirium.

In summary, both intraoperative and postoperative hypotension were associated with delirium in noncardiac surgical patients admitted to a critical care unit. Unlike other risk factors for delirium, hypotension is potentially preventable. Randomized trials to confirm causality and the potential benefit of intervening to minimize hypotension seem well worth conducting.

DISCLOSURES

Name: Kamal Maheshwari, MD.

Contribution: This author helped design the study, review and analyze the data, write the article, and review and approve the final manuscript.

Name: Sanchit Ahuja, MD.

Contribution: This author helped design the study, review and analyze the data, write the article, and review and approve the final manuscript.

Name: Ashish K. Khanna, MD, FCCP, FCCM.

Contribution: This author helped design the study, review and analyze the data, write the article, and review and approve the final manuscript.

Name: Guangmei Mao, PhD, MPH.

Contribution: This author helped review the data, write the article, and review and approve the final manuscript.

Name: Silvia Perez-Protto, MD, MS.

Contribution: This author helped review and approve the final manuscript.

Name: Ehab Farag, MD, FRCA, FASA.

Contribution: This author helped review and approve the final manuscript.

Name: Alparslan Turan, MD.

Contribution: This author helped review and approve the final manuscript.

Name: Andrea Kurz, MD.

Contribution: This author helped review and approve the final manuscript.

Name: Daniel I. Sessler, MD.

Contribution: This author helped design the study, review and analyze the data, write the article, and review and approve the final manuscript.

This manuscript was handled by: Avery Tung, MD, FCCM.

FOOTNOTES

REFERENCES

1. Marcantonio ER, Goldman L, Orav EJ, Cook EF, Lee TH. The association of intraoperative factors with the development of postoperative delirium. Am J Med. 1998;105:380–384.
2. Robinson TN, Raeburn CD, Tran ZV, Angles EM, Brenner LA, Moss M. Postoperative delirium in the elderly: risk factors and outcomes. Ann Surg. 2009;249:173–178.
3. Rudolph JL, Marcantonio ER. Review articles: postoperative delirium: acute change with long-term implications. Anesth Analg. 2011;112:1202–1211.
4. Aldecoa C, Bettelli G, Bilotta F, et al. European Society of Anaesthesiology evidence-based and consensus-based guideline on postoperative delirium. Eur J Anaesthesiol. 2017;34:192–214.
5. Bryson GL, Wyand A. Evidence-based clinical update: general anesthesia and the risk of delirium and postoperative cognitive dysfunction. Can J Anaesth. 2006;53:669–677.
6. Hála M. Pathophysiology of postoperative delirium: systemic inflammation as a response to surgical trauma causes diffuse microcirculatory impairment. Med Hypotheses. 2007;68:194–196.
7. Monk TG, Bronsert MR, Henderson WG, et al. Association between intraoperative hypotension and hypertension and 30-day postoperative mortality in noncardiac surgery. Anesthesiology. 2015;123:307–319.
8. Salmasi V, Maheshwari K, Yang D, et al. Relationship between intraoperative hypotension, defined by either reduction from baseline or absolute thresholds, and acute kidney and myocardial injury after noncardiac surgery: a retrospective cohort analysis. Anesthesiology. 2017;126:47–65.
9. Sun LY, Wijeysundera DN, Tait GA, Beattie WS. Association of intraoperative hypotension with acute kidney injury after elective noncardiac surgery. Anesthesiology. 2015;123:515–523.
10. Walsh M, Devereaux PJ, Garg AX, et al. Relationship between intraoperative mean arterial pressure and clinical outcomes after noncardiac surgery: toward an empirical definition of hypotension. Anesthesiology. 2013;119:507–515.
11. Wesselink EM, Kappen TH, Torn HM, Slooter AJC, van Klei WA. Intraoperative hypotension and the risk of postoperative adverse outcomes: a systematic review. Br J Anaesth. 2018;121:706–721.
12. Whitlock EL, Vannucci A, Avidan MS. Postoperative delirium. Minerva Anestesiol. 2011;77:448–456.
13. De Raedt R, D’haenen H, Everaert H, Cluydts R, Bossuyt A. Cerebral blood flow related to induction of a depressed mood within and out of the realm of attention in normal volunteers. Psychiatry Res. 1997;74:159–171.
14. Fong TG, Bogardus ST Jr, Daftary A, et al. Cerebral perfusion changes in older delirious patients using 99mTc HMPAO SPECT. J Gerontol A Biol Sci Med Sci. 2006;61:1294–1299.
15. Gani H, Domi R, Kodra N, et al. The incidence of postoperative delirium in elderly patients after urologic surgery. Med Arch. 2013;67:45–47.
16. Tognoni P, Simonato A, Robutti N, et al. Preoperative risk factors for postoperative delirium (POD) after urological surgery in the elderly. Arch Gerontol Geriatr. 2011;52:e166–e169.
17. Deiner S, Silverstein JH. Postoperative delirium and cognitive dysfunction. Br J Anaesth. 2009;103(suppl 1):i41–i46.
18. Litaker D, Locala J, Franco K, Bronson DL, Tannous Z. Preoperative risk factors for postoperative delirium. Gen Hosp Psychiatry. 2001;23:84–89.
19. Neerland BE, Krogseth M, Juliebø V, et al. Perioperative hemodynamics and risk for delirium and new onset dementia in hip fracture patients; A prospective follow-up study. PLoS One. 2017;12:e0180641.
20. Wang NY, Hirao A, Sieber F. Association between intraoperative blood pressure and postoperative delirium in elderly hip fracture patients. PLoS One. 2015;10:e0123892.
21. Khanna AK, Maheshwari K, Mao G, et al. Association between mean arterial pressure and acute kidney injury and a composite of myocardial injury and mortality in postoperative critically ill patients: a retrospective cohort analysis. Crit Care Med. 2019;47:910–917.
22. Khan BA, Perkins AJ, Gao S, et al. The confusion assessment method for the ICU-7 delirium severity scale: a novel delirium severity instrument for use in the ICU. Crit Care Med. 2017;45:851–857.
23. Gusmao-Flores D, Salluh JI, Chalhub RÁ, Quarantini LC. The confusion assessment method for the intensive care unit (CAM-ICU) and intensive care delirium screening checklist (ICDSC) for the diagnosis of delirium: a systematic review and meta-analysis of clinical studies. Crit Care. 2012;16:R115.
24. Nishimura K, Yokoyama K, Yamauchi N, et al.; TMAD investigators. Sensitivity and specificity of the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) and the Intensive Care Delirium Screening Checklist (ICDSC) for detecting post-cardiac surgery delirium: a single-center study in Japan. Heart Lung. 2016;45:15–20.
25. Sessler CN, Gosnell MS, Grap MJ, et al. The Richmond Agitation-Sedation Scale: validity and reliability in adult intensive care unit patients. Am J Respir Crit Care Med. 2002;166:1338–1344.
26. Shintani AK, Girard TD, Eden SK, Arbogast PG, Moons KG, Ely EW. Immortal time bias in critical care research: application of time-varying Cox regression for observational cohort studies. Crit Care Med. 2009;37:2939–2945.
27. Dekker FW, de Mutsert R, van Dijk PC, Zoccali C, Jager KJ. Survival analysis: time-dependent effects and time-varying risk factors. Kidney Int. 2008;74:994–997.
28. Dignam JJ, Zhang Q, Kocherginsky M. The use and interpretation of competing risks regression models. Clin Cancer Res. 2012;18:2301–2308.
29. Siepe M, Pfeiffer T, Gieringer A, et al. Increased systemic perfusion pressure during cardiopulmonary bypass is associated with less early postoperative cognitive dysfunction and delirium. Eur J Cardiothorac Surg. 2011;40:200–207.
30. Hirsch J, DePalma G, Tsai TT, Sands LP, Leung JM. Impact of intraoperative hypotension and blood pressure fluctuations on early postoperative delirium after non-cardiac surgery. Br J Anaesth. 2015;115:418–426.
31. Jouan Y, Seegers V, Meziani F, et al. Effects of mean arterial pressure on arousal in sedated ventilated patients with septic shock: a SEPSISPAM post hoc exploratory study. Ann Intensive Care. 2019;9:54.
32. Maheshwari K, Nathanson BH, Munson SH, et al. The relationship between ICU hypotension and in-hospital mortality and morbidity in septic patients. Intensive Care Med. 2018;44:857–867.
33. Mascha EJ, Dalton JE, Kurz A, Saager L. Statistical grand rounds: understanding the mechanism: mediation analysis in randomized and nonrandomized studies. Anesth Analg. 2013;117:980–994.
34. Vernooij LM, van Klei WA, Machina M, Pasma W, Beattie WS, Peelen LM. Different methods of modelling intraoperative hypotension and their association with postoperative complications in patients undergoing non-cardiac surgery. Br J Anaesth. 2018;120:1080–1089.

Supplemental Digital Content

Copyright © 2019 International Anesthesia Research Society