10/15/19

1

Sleep and Postoperative Delirium-New Avenues

Balachundhar Subramaniam MD, MPH, FASAEllisson C. “Jeep” Pierce Endowed Chair of AnesthesiaAssociate Professor of Anesthesiology, HarvardDirector, Center for Anesthesia Research ExcellenceBeth Israel Deaconess Medical Center, Boston, MAOctober 20th 2019

Disclosures

– NIH R01 funding as PI (NIGMS, NIA)– Mallinckrodt (Investigator initiated research, Consultant)– Edward Life Sciences (Investigator initiated research, Consultant)

2

10/15/19

2

Postoperative Delirium

– Current Postoperative Delirium Burden– Sleep and Delirium- What is the connection– Enhancing Natural Sleep- Current work

3

Topic Outline

– Current Postoperative Delirium Burden– Sleep and Delirium- What is the connection– Enhancing Natural Sleep- Current work

4

10/15/19

3

Association of any Adverse Hospital Outcomes by Complication and Delirium Statusa

– Effect of Delirium and Other Major Complications on Outcomes After Elective Surgery in Older Adults. JAMA Surg. 2015;150(12):1134-1140.

5

Copyright © 2018 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.

Online Special Article

Critical Care Medicine www.ccmjournal.org e825

1School of Pharmacy, Northeastern University, Boston, MA.2Division of Pulmonary, Critical Care and Sleep Medicine, Tufts Medical Center, Boston, MA.

3Faculty of Medicine, McGill University, Montreal, QC, Canada.4Regroupement de Soins Critiques Respiratoires, Réseau de Santé Res-piratoire, Montreal, QC, Canada.

5Ingram School of Nursing, McGill University, Montreal, QC, Canada.6Division of Pulmonary and Critical Care Medicine, Department of Physi-cal Medicine and Rehabilitation, School of Medicine, Johns Hopkins Uni-versity, Baltimore, MD.

7Department of Intensive Care Medicine, Brain Center Rudolf Magnus, University Medical Center, Utrecht University, Utrecht, The Netherlands.

8Department of Anesthesiology, Division of Anesthesiology Critical Care Medicine, Vanderbilt University Medical Center, Nashville, TN.

9Division of Sleep Medicine, Vanderbilt University Medical Center, Nash-ville, TN.

10 Division of Pulmonary and Critical Care, Brigham and Women’s Hospital and School of Medicine, Harvard University, Boston, MA.

11 Division of Anesthesiology, Perioperative Care and Pain Medicine, New York University Langone Health, New York, NY.

12 Division of Medicine, New York University Langone Health, New York, NY.

13 Division of Neurology, New York University Langone Health, New York, NY.

14Division of Surgery, New York University Langone Health, New York, NY.15 Department of Medicine (Critical Care), McMaster University, Hamilton,

ON, Canada.16 Department of Health Research Methods, Impact and Evidence, McMas-

ter University, Hamilton, ON, Canada.17 The Ohio State University, College of Nursing, Center of Excellence in

Critical and Complex Care, Columbus, OH.18The Ohio State University Wexner Medical Center, Columbus, OH.19 Department of Intensive Care Medicine, Radboud University Medical

Center, Nijmegen, The Netherlands.20 Division of Critical Care, London Health Sciences Centre, London, ON,

Canada.21 Schulich School of Medicine & Dentistry, University of Western Ontario,

London, ON, Canada.22 Center for Quality Aging, Division of Allergy, Pulmonary and Critical Care

Medicine, Department of Medicine, Vanderbilt University Medical Center, Nashville, TN.

23 Center for Health Services Research, Vanderbilt University Medical Cen-ter, Nashville, TN.DOI: 10.1097/CCM.0000000000003299

Clinical Practice Guidelines for the Prevention and Management of Pain, Agitation/Sedation, Delirium, Immobility, and Sleep Disruption in Adult Patients in the ICU

John W. Devlin, PharmD, FCCM (Chair)1,2; Yoanna Skrobik, MD, FRCP(c), MSc, FCCM (Vice-Chair)3,4; Céline Gélinas, RN, PhD5; Dale M. Needham, MD, PhD6; Arjen J. C. Slooter, MD, PhD7; Pratik P. Pandharipande, MD, MSCI, FCCM8; Paula L. Watson, MD9; Gerald L. Weinhouse, MD10; Mark E. Nunnally, MD, FCCM11,12,13,14; Bram Rochwerg, MD, MSc15,16; Michele C. Balas, RN, PhD, FCCM, FAAN17,18; Mark van den Boogaard, RN, PhD19; Karen J. Bosma, MD20,21; Nathaniel E. Brummel, MD, MSCI22,23; Gerald Chanques, MD, PhD24,25; Linda Denehy, PT, PhD26; Xavier Drouot, MD, PhD27,28; Gilles L. Fraser, PharmD, MCCM29; Jocelyn E. Harris, OT, PhD30; Aaron M. Joffe, DO, FCCM31; Michelle E. Kho, PT, PhD30; John P. Kress, MD32; Julie A. Lanphere, DO33; Sharon McKinley, RN, PhD34; Karin J. Neufeld, MD, MPH35; Margaret A. Pisani, MD, MPH36; Jean-Francois Payen, MD, PhD37; Brenda T. Pun, RN, DNP23; Kathleen A. Puntillo, RN, PhD, FCCM38; Richard R. Riker, MD, FCCM29; Bryce R. H. Robinson, MD, MS, FACS, FCCM39; Yahya Shehabi, MD, PhD, FCICM40; Paul M. Szumita, PharmD, FCCM41; Chris Winkelman, RN, PhD, FCCM42; John E. Centofanti, MD, MSc43; Carrie Price, MLS44; Sina Nikayin, MD45; Cheryl J. Misak, PhD46; Pamela D. Flood, MD47; Ken Kiedrowski, MA48; Waleed Alhazzani, MD, MSc (Methodology Chair)16,49

Copyright © 2018 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.

6

Answer Options Rating Average

Non-pharmacologic prevention: Early mobility 4.50

Non-pharmacologic prevention: Environment (e.g., noise, light) 4.50

Non-pharmacologic prevention: Sleep protocols 4.50

Non-pharmacologic prevention: Music therapy 4.50

Predictors for ICU delirium (i.e., modifiable vs. non-modifiable) 4.33

Pharmacologic prevention: choice of sedative 4.33

Pharmacologic prevention: haloperidol 4.17

Pharmacologic prevention: atypical antipsychotics 4.17

Pharmacologic treatment: dexmedetomidine 4.17

Non-Pharmacologic treatment: early mobility 4.00

Non-Pharmacologic treatment: environment (e.g. noise, light etc) 4.00

Non-Pharmacologic treatment: sleep protocols 4.00

Non-Pharmacologic treatment: music therapy 4.00

Pharmacologic treatment: haloperidol 4.00

Crit Care Med. 2018 Sep;46(9):e825-e873.

10/15/19

4

7

Complete consensus for delirium treatment

Intervenable causes

– Sleep, Pain and Cognition

8

Main Outcome and Measures—Major postoperative complications, defined as life altering or threatening events (Accordion Severity ≥ grade 2), were identified by expert panel adjudication. Delirium was measured daily with the Confusion Assessment Method and a validated chart review method. Four subgroups were analyzed: (1) no complications, no delirium; (2) complications alone; (3) delirium alone; and (4) both complications and delirium. Adverse outcomes included length of stay (LOS) > 5 days, institutional discharge, and rehospitalization within 30 days of discharge.

Results—Of 566 participants, mean age (±SD) was 76.7± 5.2 years, 42% male and 92% white. Forty-seven (8%) developed major complications, and 135 (24%) developed delirium. When compared to no complications, no delirium as the reference group, major complications alone contributed only to prolonged LOS (RR 2.8, 95% CI 1.9–4.0); by contrast, delirium alone significantly increased all adverse outcomes, including prolonged LOS (RR 1.9, 95% CI 1.4–2.7), institutional discharge (RR 1.5, 95% CI 1.3–1.7), and 30-day readmission (RR 2.3, 95% CI 1.4–3.7). The subgroup with both complications and delirium had the highest rates of all adverse outcomes, including prolonged LOS (RR 3.4, 95% CI 2.3–4.8), institutional discharge (RR 1.8, 95% CI 1.4–2.5) and 30-day readmission (RR 3.0, 95% CI 1.3–6.8). Delirium exerted the highest attributable risk at a population level (5.8%, 95% CI 4.7–6.8) compared with all other adverse events (prolonged LOS, institutional discharge, or readmission).

Conclusions and Relevance—Major postoperative complications and delirium are separately associated with adverse events and demonstrate a strong combined effect. Delirium occurs more frequently, and has greater impact at the population level than other major complications.

KeywordsDelirium; postoperative complications; Confusion Assessment Method; adverse surgical outcomes

IntroductionThirty-six percent of inpatient operations were performed in patients ≥65 years old.1 The number of patients is projected to increase as the population ages.1,2 Understanding the risks for adverse outcomes in the aging surgical population is essential to implement programs with the potential to decrease morbidity, mortality, costs, and safety. Postoperative complications increase with advancing age.3,4 These complications, occurring in 10–25% older persons,3–7 can lead to adverse outcomes such as disability, loss of independence, diminished quality of life, high health care costs, and increased mortality.8 Physiologic changes across organ systems, cardiovascular, neurological, and pulmonary systems, contribute to the increasing risk of postoperative complications with advancing age.6,8

Postoperative delirium as the leading complication of major surgery in older persons with adverse consequences has been well documented across many studies.9–12 It is associated with higher in-hospital and 6-month mortality, functional decline, greater rates of institutional discharge, longer lengths of stay (LOS), increased use of hospital resources and higher healthcare costs.9–14 Delirium rates following surgery range from 5%–50%.12,14,15

Its impact on annual health care costs is estimated at over $182 billion per year in the U.S. (2011 USD),16 and thus, has assumed increasing attention as a public health and patient

Gleason et al. Page 2

JAMA Surg. Author manuscript; available in PMC 2016 December 01.

Author Manuscript

Author Manuscript

Author Manuscript

Author Manuscript

Effect of Intravenous Acetaminophen vs Placebo CombinedWith Propofol or Dexmedetomidine on PostoperativeDelirium Among Older Patients Following Cardiac SurgeryThe DEXACET Randomized Clinical TrialBalachundhar Subramaniam, MD, MPH; Puja Shankar, MD; Shahzad Shaefi, MD, MPH; Ariel Mueller, MA; Brian O’Gara, MD;Valerie Banner-Goodspeed, MPH; Jackie Gallagher, MS; Doris Gasangwa, BS; Melissa Patxot, BS; Senthil Packiasabapathy, MBBS; Pooja Mathur, BS;Matthias Eikermann, MD, PhD; Daniel Talmor, MD, MPH; Edward R. Marcantonio, MD, SM

IMPORTANCE Postoperative delirium is common following cardiac surgery and may beaffected by choice of analgesic and sedative.

OBJECTIVE To evaluate the effect of postoperative intravenous (IV) acetaminophen(paracetamol) vs placebo combined with IV propofol vs dexmedetomidine on postoperativedelirium among older patients undergoing cardiac surgery.

DESIGN, SETTING, AND PARTICIPANTS Randomized, placebo-controlled, factorial clinical trialamong 120 patients aged 60 years or older undergoing on-pump coronary artery bypass graft(CABG) surgery or combined CABG/valve surgeries at a US center. Enrollment was September2015 to April 2018, with follow-up ending in April 2019.

INTERVENTIONS Patients were randomized to 1 of 4 groups receiving postoperative analgesiawith IV acetaminophen or placebo every 6 hours for 48 hours and postoperative sedationwith dexmedetomidine or propofol starting at chest closure and continued for up to 6 hours(acetaminophen and dexmedetomidine: n = 29; placebo and dexmedetomidine: n = 30;acetaminophen and propofol: n = 31; placebo and propofol: n = 30).

MAIN OUTCOMES AND MEASURES The primary outcome was incidence of postoperativein-hospital delirium by the Confusion Assessment Method. Secondary outcomes includeddelirium duration, cognitive decline, breakthrough analgesia within the first 48 hours, andICU and hospital length of stay.

RESULTS Among121patientsrandomized(medianage,69years;19women[15.8%]),120completedthe trial. Patients treated with IV acetaminophen had a significant reduction in delirium (10% vs 28%placebo;difference,−18%[95%CI,−32%to−5%];P = .01;HR,2.8[95%CI,1.1-7.8]).Patientsreceivingdexmedetomidine vs propofol had no significant difference in delirium (17% vs 21%; difference, −4%[95%CI,−18%to10%];P = .54;HR,0.8[95%CI,0.4-1.9]).Thereweresignificantdifferencesfavoringacetaminophen vs placebo for 3 prespecified secondary outcomes: delirium duration (median, 1 vs2 days; difference, −1 [95% CI, −2 to 0]), ICU length of stay (median, 29.5 vs 46.7 hours; difference,−16.7[95%CI,−20.3to−0.8]),andbreakthroughanalgesia(median,10 082.5vs12 609.0µgmorphineequivalents; difference, −2530 [95% CI, −5064 to −22]). For dexmedetomidine vs propofol, onlybreakthrough analgesia was significantly different (median, 10 110.0 vs 12 612.5 µg; difference,−2567 [95% CI, −5094 to −26]; P = .03). Fourteen patients in both the placebo-dexmedetomidineandacetaminophen-propofolgroups(46%and45%)and7intheacetaminophen-dexmedetomidineand placebo-propofol groups (24% and 23%) had hypotension.

CONCLUSIONS AND RELEVANCE Among older patients undergoing cardiac surgery, postoperativescheduled IV acetaminophen, combined with IV propofol or dexmedetomidine, reduced in-hospitaldelirium vs placebo. Additional research, including comparison of IV vs oral acetaminophen and otherpotentially opioid-sparing analgesics, on the incidence of postoperative delirium is warranted.

TRIAL REGISTRATION ClinicalTrials.gov Identifier: NCT02546765

JAMA. 2019;321(7):686-696. doi:10.1001/jama.2019.0234Corrected on July 16, 2019.

Video and Supplementalcontent

CME Quiz atjamanetwork.com/learning

Author Affiliations: Department ofAnesthesia, Critical Care, and PainMedicine, Beth Israel DeaconessMedical Center, Harvard MedicalSchool, Boston, Massachusetts(Subramaniam, Shankar, Shaefi,Mueller, O’Gara, Banner-Goodspeed,Gasangwa, Patxot, Packiasabapathy,Mathur, Eikermann, Talmor);Department of Medicine, Division ofGerontology, Beth Israel DeaconessMedical Center, Harvard MedicalSchool, Boston, Massachusetts(Gallagher, Marcantonio).

Corresponding Author:Balachundhar Subramaniam, MD,MPH, Beth Israel Deaconess MedicalCenter, 375 Longwood Ave,W/MS-414, Boston, MA 02215(bsubrama@bidmc.harvard.edu).

Section Editor: Derek C. Angus, MD,MPH, Associate Editor, JAMA(angusdc@upmc.edu).

Research

JAMA | Preliminary Communication | CARING FOR THE CRITICALLY ILL PATIENT

686 (Reprinted) jama.com

© 2019 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 10/03/2019

10/15/19

5

9

Copyright © 2018 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.

Devlin et al

e854 www.ccmjournal.org September 2018 • Volume 46 • Number 9

Evidence Gaps: Studies using questionnaires and interviews, while patient-centered, are subject to recall bias and exclude patients who are not able to self-report due to sedation, delir-ium, dementia, or acute brain injury. Furthermore, patients may have sleep that is severely fragmented with microarous-als, but the patients may not be able to identify the disrup-tive factors because they were not fully awakened from sleep. Studies using polysomnography are limited to those that can be analyzed by standard criteria and exclude highly abnormal electroencephalograms or those with poor-quality electroen-cephalogram signals. Studies correlating various factors with impaired sleep on polysomnography do not prove causation, only association, and were largely weak associations in univari-ate analysis.

OutcomesQuestion: Do sleep and circadian rhythm alterations “during” an ICU admission affect outcomes during and/or after the ICU stay in critically ill adults?

Ungraded Statements: Although an association between sleep quality and delirium occurrence exists in critically ill adults, a cause-effect relationship has not been established.

An association between sleep quality and duration of mechanical ventilation, length of ICU stay, and ICU mortality in critically ill adults remains unclear.

The effects of sleep quality and circadian rhythm alterations on outcomes in critically ill patients after ICU discharge are unknown.

Rationale: A handful of studies help answer these ques-tions (Supplemental Table 45, Supplemental Digital Content 55, http://links.lww.com/CCM/D813). Poor sleep quality is often assumed to be a potentially modifiable risk factor for ICU delirium; several studies have evaluated this relationship. Critically ill adults who are severely sleep deprived are 30% more likely to have mental status changes (441). Subsequent polysomnography studies further supported this association (442, 451). Critically ill adults with severe REM deprivation (442, 451) and circadian sleep-cycle disruption (as evidenced by a greater proportion of daytime sleep) are more likely to experience delirium (451). Poor sleep quality has also been found to be an independent risk factor (496) for postcardiac surgery ICU delirium. Additionally, before-and-after observa-tional studies of multidisciplinary bundles that include sleep enhancement protocols have been shown to decrease delirium prevalence (312, 454), although in only one study did sleep efficiency improve with the intervention (312). Although an association between sleep quality and delirium occurrence exists, it remains unknown if poor sleep is a cause for delirium.

Use of a multicomponent delirium prevention protocol that incorporated a nonpharmacologic sleep enhancement protocol

TABLE 2. List of Factors That Patients Report as Disruptive to SleepEnvironmental Physiologic and Pathophysiologic

Noise (447, 453, 454, 480, 483–488, 490, 491) Pain (454, 483–486, 488, 490, 491)

Light (241, 453, 454, 480, 482–484, 486–488) Discomfort (454, 483, 486, 488, 490)

Comfort of bed (483, 486–488) Feeling too hot or too cold (484, 486, 488)

Activities at other bedsides (483, 486, 487) Breathing difficulty (484, 491)

Visitors (clinician or family) (483) Coughing (484, 491)

Room ventilation system (483) Thirst (484, 486) and hunger (486, 488)

Hand washing by clinicians (483) Nausea (484, 488)

Bad odor (486, 488) Needing to use bedpan/urinal (486, 488)

Care Related Psychologic

Nursing care (447, 453, 480, 482–484, 486, 488, 491) Anxiety/worry/stress (483, 484, 486, 489–491)

Patient procedures (447, 453, 480, 482, 483, 487, 488) Fear (485, 486, 489)

Vital sign measurement (442, 448, 475, 477, 481, 483) Unfamiliar environment (485, 488, 491)

Diagnostic tests (447, 453, 480, 483) Disorientation to time (454, 486)

Medication administration (447, 453, 480, 482) Loneliness (488, 491)

Restricted mobility from lines/catheters (454, 486, 488) Lack of privacy (485, 488)

Monitoring equipment (454, 486, 488) Hospital attire (486, 488)

Oxygen mask (486, 488) Missing bedtime routine (483)

Endotracheal tube (491) Not knowing nurses’ names (486)

Urinary catheters (486) Not understanding medical terms (486)

Abnormal Sleep patterns in ICU

Freedman, Gazendam, LeVan,

et al.

: Environmental Noise and Sleep in the ICU 453

(total time and percentage of total sleep time), and (

5

) total sleep time(TST).

Pearson’s correlation analysis and one-way analysis of variancewere used to determine the relationship of patient age, duration ofstay, and APACHE III score to the following factors: sleep stages (1,2, 3/4, and REM), arousal indexes (noise, nonnoise, and total), day-time sleep (total time and percentage of total sleep time), nighttimesleep (total time and percentage of total sleep time), and total sleeptime (TST). Spearman’s correlation analysis was used to confirm thePearson’s analysis. Only the Pearson’s correlation coefficients werereported if there was an agreement between these latter two analyses.

RESULTS

Demographics

A total of 24 ICU patients were enrolled between March 1997and February 1999. Two patients withdrew from the studyprior to its initiation at the request from the patient’s familymembers after the initial consent was given. Twenty-two pa-tients completed the study for a total of 30 24-h periods (8 pa-tients were studied for 48 h continuously and the remaining 14patients were studied for 24 h). The study population wascomprised of 12 males and 10 females with a mean age of61

6

16 yr (range 20–83), mean APACHE III score of 57

6

28 (range 7–132), and a mean duration of ICU stay prior tothe study of 18

6

20 d (range 3–80). Twenty of the 22 patientswere mechanically ventilated at the time of the study, and re-mained mechanically ventilated for the entire study period.Primary reasons for mechanical ventilation included pneumo-nia (7 patients), sepsis (5 patients), chronic obstructive pulmo-nary disease (COPD) exacerbation (3 patients), acute respira-tory distress syndrome (ARDS) (3 patients), and myastheniagravis (2 patients). Fourteen patients received no sedationduring the study period. Of the remaining 8 patients, 4 pa-tients received intermittent, as needed, intravenous lorezepamand 4 patients received intermittent, as needed, intravenousdoses of fentanyl. No patients were on a combination of ben-zodiazepines and narcotics during the study period. No pa-tients received tricyclic or other types of antidepressant medi-

cations during the study period. All patients had a Glasgowcoma score of 14 or greater upon entrance into the study.

Sleep Architecture and Sleep Distribution

Seventeen (77.3%) of the 22 patients had scorable EEG dataaccording to the criteria of Rechtschaffen and Kales (18).From this group, there were a total of 21 separate 24-h day/night periods (13 patients with 24 h of scorable data and 4 pa-tients with 48 h of scorable data). The other 5 patients (22.7%)demonstrated evidence of septic encephalopathy throughoutthe majority or all of the study period, and were therefore un-able to be scored according to standard criteria of sleep versuswake (

see

section on sepsis and sleep). Of the 17 ICU patientswith scorable sleep–wake polysomnograms, all demonstratedabnormal sleep architecture. Although the mean total sleeptime per 24 h period was within the normal range (8.8

6

5.0 h)(17) there were large individual variations in total sleep time(range 1.7 to 19.4 h).

There was a predominance of stage 1 sleep (mean 59

6

33%) with decreased or absent stages 2 (mean 26

6

28%), 3/4(mean 9

6

18%), and REM sleep (mean 6

6

9%). Twelve ofthese 17 patients demonstrated no REM sleep. Of the 8 pa-tients who were studied continuously for 48 h, 5 had scorableEEG data. There were no significant differences betweenstudy Day 1 and 2 with respect to mean total time spent insleep stages 1, 3/4, and REM, day versus night total sleeptimes, or the overall arousal index. There were no significantdifferences (p

.

0.05) between sexes or window status with re-spect to TST/24-h period in time in any sleep stage. Therewere no significant correlations (p

.

0.05) between TST/24-hperiod or time in any sleep stage with age, duration of ICUstay, or APACHE III score.

Sleep Disruption and Arousal Index

All 17 patients with scorable sleep–wake stages demonstratednonconsolidated sleep that was distributed throughout the 24-hday–night study period (

see

Figure 2). Fifty-seven percent

6

Figure 2. Schematic representation of theredistribution of sleep and wake in five sub-jects over the 24-h period. Black areas rep-resent episodes of sleep and white areasrepresent wakefulness.

10

10/15/19

6

ICU Sleep- Key features– Total sleep time (TST) and sleep efficiency are often normal– Sleep fragmentation, the proportion of time spent in light sleep (stages

N1 + N2), and time spent sleeping during the day (vs. night) are higher– The proportion of time spent in deep sleep (stage N3 sleep and REM) is

lower– Subjective sleep quality is reduced– NO CAUSE AND EFFECT between sleep and delirium– Sleep measurements in ICU is challenging– Is sleep different in critically ill adults if delirium is present?

• Greater disruption of circadian rhythm; REM sleep is lower

11

Postoperative Delirium

– Current Postoperative Delirium Burden– Sleep and Delirium- What is the connection– Enhancing Natural Sleep- Current work

12

10/15/19

7

13

A Systematic Review and Meta-Analysis Examining the Impact of Sleep Disturbance on Postoperative Delirium

Fadayomi, Ayòtúndé B.; Ibala, Reine; Bilotta, Federico; Westover, Michael B.; Akeju, OluwaseunCritical Care Medicine46(12):e1204-e1212, December 2018.

A Systematic Review and Meta-Analysis Examining the Impact of Sleep Disturbance on Postoperative Delirium

Forest plot for all studies and subgroup analyses by study characteristics. A, Forest plot showing pooled analysis of studies categorized into obstructive sleep apnea and unspecified types of sleep disturbance. B, Forest plot showing pooled analysis of studies categorized into preoperative sleep disturbance and postoperative sleep prospective.

14

Fadayomi, Ayòtúndé B.; Ibala, Reine; Bilotta, Federico; Westover, Michael B.; Akeju, Oluwaseun. Critical Care Medicine46(12):e1204-e1212, December 2018.

10/15/19

8

Forest plot for all studies and subgroup analyses by study characteristics. A, Forest plot showing pooled analysis of studies categorized into prospective and retrospective study design. B, Forest plot showing pooled analysis of studies categorized into studies with median/mean age less than 65 yr and greater than or equal to 65 yr.

Copyright © 2019 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All rights reserved. 15

Fadayomi, Ayòtúndé B.; Ibala, Reine; Bilotta, Federico; Westover, Michael B.; Akeju, Oluwaseun

Critical Care Medicine46(12):e1204-e1212, December 2018.

A Systematic Review and Meta-Analysis Examining the Impact of Sleep Disturbance on Postoperative Delirium

A Systematic Review and Meta-Analysis Examining the Impact of Sleep Disturbance on Postoperative Delirium

Forest plot for all studies and subgroup analyses by study characteristics. A, Forest plot showing pooled analysis of studies with crude odds ratio. B, Forest plot showing pooled analysis after restricting to those with sleep disturbance before onset of delirium.

Copyright © 2019 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All rights reserved. 16

Fadayomi, Ayòtúndé B.; Ibala, Reine; Bilotta, Federico; Westover, Michael B.; Akeju, Oluwaseun

Critical Care Medicine46(12):e1204-e1212, December 2018.

10/15/19

9

Sleep Restriction on Circadian secretory pattern

17

The Journal of Clinical Endocrinology & Metabolism, Volume 89, Issue 5, 1 May 2004, Pages 2119–2126,

TNF α IL-6

Neuroinflammation and Perioperative NeuroCognitive Disorders

18

Subramaniyan S, Terrando N. Neuroinflammation and Perioperative Neurocognitive Disorders. Anesth Analg. 2019;128(4):781–788.

10/15/19

10

Novel risk markers and long-term outcomes of delirium: The Successful Aging after Elective Surgery (SAGES) Study Design and MethodsSchmitt EM,, et al. Novel risk markers and long-term outcomes of delirium: the successful aging after elective surgery (SAGES) study design and methods. J Am Med Dir Assoc. 2012;13(9):818.

19

Postoperative Delirium

– Current Postoperative Delirium Burden– Sleep and Delirium- What is the connection– Enhancing Natural Sleep- Current work

20

10/15/19

11

Brain age from the electroencephalogram of sleep

Haoqi Sun a, Luis Paixao a, Jefferson T. Oliva a,b, Balaji Goparaju a, Diego Z. Carvalho a,Kicky G. van Leeuwen a,c, Oluwaseun Akeju d, Robert J. Thomas e, Sydney S. Cash a,Matt T. Bianchi a,1, M. Brandon Westover a,*,1aDepartment of Neurology, Massachusetts General Hospital, Boston, MA, USAbBioinspired Computing Laboratory, Computer Science Department, University of São Paulo, BrazilcUniversity of Twente, Enschede, the NetherlandsdDepartment of Anesthesiology, Critical Care and Pain Medicine, Massachusetts General Hospital, Boston, MA, USAeDepartment of Medicine, Division of Pulmonary, Critical Care & Sleep, Beth Israel Deaconess Medical Center, Boston, MA, USA

a r t i c l e i n f o

Article history:Received 24 January 2018Received in revised form 12 October 2018Accepted 14 October 2018Available online 19 October 2018

Keywords:Brain ageSleepEEGMachine learning

a b s t r a c t

The human electroencephalogram (EEG) of sleep undergoes profound changes with age. These changescan be conceptualized as “brain age (BA),” which can be compared to chronological age to reflect thedegree of deviation from normal aging. Here, we develop an interpretable machine learning model topredict BA based on 2 large sleep EEG data sets: the Massachusetts General Hospital (MGH) sleep labdata set (N ¼ 2532; ages 18e80); and the Sleep Heart Health Study (SHHS, N ¼ 1974; ages 40e80). Themodel obtains a mean absolute deviation of 7.6 years between BA and chronological age (CA) in healthyparticipants in the MGH data set. As validation, a subset of SHHS containing longitudinal EEGs 5.2 yearsapart shows an average of 5.4 years increase in BA. Participants with significant neurological or psy-chiatric disease exhibit a mean excess BA, or “brain age index” (BAI ¼ BA-CA) of 4 years relative tohealthy controls. Participants with hypertension and diabetes have a mean excess BA of 3.5 years. Thefindings raise the prospect of using the sleep EEG as a potential biomarker for healthy brain aging.

! 2018 Elsevier Inc. All rights reserved.

1. Introduction

Human sleep undergoes robust and predictable changes withage, reflected in both overall sleep architecture and electroen-cephalogram (EEG) oscillations/waveforms. At the level of sleeparchitecture (macrostructure), older participants sleep earlier andwake earlier, have shorter sleep duration, increased sleep frag-mentation, and reduced percentages of rapid eye movement sleep,as well as (at least in males) deep non-rapid eye movement sleep(Mander et al., 2017; Scullin, 2017). At the level of EEG micro-structure, older participants exhibit reduced slow waves duringdeep sleep (Carrier et al., 2001; Larsen et al., 1995); decreased sleepspindle amplitude, density, and duration (Purcell et al., 2017); andless phase coupling between slow oscillations and sleep spindles(Helfrich et al., 2017). However, no study has yet addressed theinverse problem: how accurately can a person’s age be predictedfrom the sleep EEG? What factors make a person’s “BA” older oryounger than their chronological age (CA)?

BA serves as a potential aging biomarkerwhere the variation of BAbetween individuals of the same chronological age may carry impor-tant information about the risk of cognitive impairment, neurologicalorpsychiatricdisease, ordeath.Variousbiomarkersof aginghavebeenproposed to better predict lifespan and functional capability, rangingfrom molecular and cellular levels to structural level biomarkers. Atthe molecular and cellular level, aging-related biomarkers includeleukocyte telomere length (Kruketal.,1995), Ink4/Arf locusexpression(Krishnamurthyet al., 2004), N-glycanprofile (Vanhooren et al., 2010),mitochondrial DNA deletions (Eshaghian et al., 2006), and DNA-methylation status at CpG sites across the genome (Petkovich et al.,2017) among others. At the structural level, brain anatomy changesdramatically throughout life (Raz and Rodrigue, 2006). For example,cortical volume (Cole et al., 2017), thalamic volume (Redline et al.,2004), and white matter integrity (Mander et al., 2017), each de-creases with aging. Using magnetic resonance images (MRI), chrono-logical age can be predicted with mean absolute deviation (MAD) of5 years in healthy participants (Franke et al., 2010). Various diseases,including Alzheimer’s disease, schizophrenia, epilepsy, traumaticbrain injury, bipolardisorder,majordepression, cognitive impairment,diabetesmellitus, andHIV, are associatedwith excess BA, that is, olderMRI BA than chronological age (Cole, 2017; Cole et al., 2018; Cole andFranke, 2017; Franke et al., 2010, 2012).

* Corresponding author at: Department of Neurology Massachusetts GeneralHospital, 55 Fruit Street, Boston, MA 02114. Tel.: þ1 650-862-1154; fax: þ1 617-726-3311.

E-mail address: mwestover@mgh.harvard.edu (M.B. Westover).1 These authors contributed equally to the work as co-senior authors.

Contents lists available at ScienceDirect

Neurobiology of Aging

journal homepage: www.elsevier .com/locate/neuaging

0197-4580/$ e see front matter ! 2018 Elsevier Inc. All rights reserved.https://doi.org/10.1016/j.neurobiolaging.2018.10.016

Neurobiology of Aging 74 (2019) 112e120

21

Brain age from the electroencephalogram of sleep

Haoqi Sun a, Luis Paixao a, Jefferson T. Oliva a,b, Balaji Goparaju a, Diego Z. Carvalho a,Kicky G. van Leeuwen a,c, Oluwaseun Akeju d, Robert J. Thomas e, Sydney S. Cash a,Matt T. Bianchi a,1, M. Brandon Westover a,*,1aDepartment of Neurology, Massachusetts General Hospital, Boston, MA, USAbBioinspired Computing Laboratory, Computer Science Department, University of São Paulo, BrazilcUniversity of Twente, Enschede, the NetherlandsdDepartment of Anesthesiology, Critical Care and Pain Medicine, Massachusetts General Hospital, Boston, MA, USAeDepartment of Medicine, Division of Pulmonary, Critical Care & Sleep, Beth Israel Deaconess Medical Center, Boston, MA, USA

a r t i c l e i n f o

Article history:Received 24 January 2018Received in revised form 12 October 2018Accepted 14 October 2018Available online 19 October 2018

Keywords:Brain ageSleepEEGMachine learning

a b s t r a c t

The human electroencephalogram (EEG) of sleep undergoes profound changes with age. These changescan be conceptualized as “brain age (BA),” which can be compared to chronological age to reflect thedegree of deviation from normal aging. Here, we develop an interpretable machine learning model topredict BA based on 2 large sleep EEG data sets: the Massachusetts General Hospital (MGH) sleep labdata set (N ¼ 2532; ages 18e80); and the Sleep Heart Health Study (SHHS, N ¼ 1974; ages 40e80). Themodel obtains a mean absolute deviation of 7.6 years between BA and chronological age (CA) in healthyparticipants in the MGH data set. As validation, a subset of SHHS containing longitudinal EEGs 5.2 yearsapart shows an average of 5.4 years increase in BA. Participants with significant neurological or psy-chiatric disease exhibit a mean excess BA, or “brain age index” (BAI ¼ BA-CA) of 4 years relative tohealthy controls. Participants with hypertension and diabetes have a mean excess BA of 3.5 years. Thefindings raise the prospect of using the sleep EEG as a potential biomarker for healthy brain aging.

! 2018 Elsevier Inc. All rights reserved.

1. Introduction

Human sleep undergoes robust and predictable changes withage, reflected in both overall sleep architecture and electroen-cephalogram (EEG) oscillations/waveforms. At the level of sleeparchitecture (macrostructure), older participants sleep earlier andwake earlier, have shorter sleep duration, increased sleep frag-mentation, and reduced percentages of rapid eye movement sleep,as well as (at least in males) deep non-rapid eye movement sleep(Mander et al., 2017; Scullin, 2017). At the level of EEG micro-structure, older participants exhibit reduced slow waves duringdeep sleep (Carrier et al., 2001; Larsen et al., 1995); decreased sleepspindle amplitude, density, and duration (Purcell et al., 2017); andless phase coupling between slow oscillations and sleep spindles(Helfrich et al., 2017). However, no study has yet addressed theinverse problem: how accurately can a person’s age be predictedfrom the sleep EEG? What factors make a person’s “BA” older oryounger than their chronological age (CA)?

BA serves as a potential aging biomarkerwhere the variation of BAbetween individuals of the same chronological age may carry impor-tant information about the risk of cognitive impairment, neurologicalorpsychiatricdisease, ordeath.Variousbiomarkersof aginghavebeenproposed to better predict lifespan and functional capability, rangingfrom molecular and cellular levels to structural level biomarkers. Atthe molecular and cellular level, aging-related biomarkers includeleukocyte telomere length (Kruketal.,1995), Ink4/Arf locusexpression(Krishnamurthyet al., 2004), N-glycanprofile (Vanhooren et al., 2010),mitochondrial DNA deletions (Eshaghian et al., 2006), and DNA-methylation status at CpG sites across the genome (Petkovich et al.,2017) among others. At the structural level, brain anatomy changesdramatically throughout life (Raz and Rodrigue, 2006). For example,cortical volume (Cole et al., 2017), thalamic volume (Redline et al.,2004), and white matter integrity (Mander et al., 2017), each de-creases with aging. Using magnetic resonance images (MRI), chrono-logical age can be predicted with mean absolute deviation (MAD) of5 years in healthy participants (Franke et al., 2010). Various diseases,including Alzheimer’s disease, schizophrenia, epilepsy, traumaticbrain injury, bipolardisorder,majordepression, cognitive impairment,diabetesmellitus, andHIV, are associatedwith excess BA, that is, olderMRI BA than chronological age (Cole, 2017; Cole et al., 2018; Cole andFranke, 2017; Franke et al., 2010, 2012).

* Corresponding author at: Department of Neurology Massachusetts GeneralHospital, 55 Fruit Street, Boston, MA 02114. Tel.: þ1 650-862-1154; fax: þ1 617-726-3311.

E-mail address: mwestover@mgh.harvard.edu (M.B. Westover).1 These authors contributed equally to the work as co-senior authors.

Contents lists available at ScienceDirect

Neurobiology of Aging

journal homepage: www.elsevier .com/locate/neuaging

0197-4580/$ e see front matter ! 2018 Elsevier Inc. All rights reserved.https://doi.org/10.1016/j.neurobiolaging.2018.10.016

Neurobiology of Aging 74 (2019) 112e120

participants) was around 0 years, whereas BAI for the group withsignificant neurological and psychiatric disease (167 participants)increased up to 4 years with a significant difference (t-test p-value< 0.01). Therefore, at the population level, participants with sig-nificant neurological or psychiatric disease had on average BA4 years older than their chronological age. Furthermore, in thehealthy group, the MAD was 7.6 years, and correlation between CAand BA was 0.83; while in the group with significant neurologicaland psychiatric disease, the MAD was larger at 9.5 years and cor-relation smaller at 0.74.

Although often not associated with any formal neurological orpsychiatric diagnosis, there is biological plausibility and accumu-lating evidence for the idea that hypertension and diabetes mellitusaccelerate brain aging (Gorelick et al., 2011, 2017). We thereforeexplored whether EEG BA is sensitive to hypertension and diabetes.These participants were identified using the diagnosis from 5 yearsbefore to 1 year after the PSG recording, as well as the self-reportedmedication intake form collected at the time of PSG recording. Ascompared in Fig. 4B, the control group was defined as participantstaking no medications, having no diagnoses of hypertension ordiabetes, and no significant neurological or psychiatric disease. Thecontrol group had a mean excess BA of BAI ¼ "1.1 years (n ¼ 472).The “contrast” group was defined as participants with both hy-pertension and diabetes, which had BAI ¼ 2.4 years (n ¼ 137). Theattributable excess BA was 3.5 years (p ¼ 0.0002).

3.6. Model-derived age norm

The parametric formulation of our BA model allows studyingindividual differences at the level of EEG features. To do this, wefirst derived an “age norm”, that is, typical EEG feature values atdifferent CAs with matched BAs. Specifically, the age norm of age xwas obtained by averaging EEG features from all healthy partici-pants who had both CA and BA within [x-5, xþ5] years.

The age norm showed that when following a normal agingprocess, the features most negatively correlated with age weredelta band (<4 Hz) powers in stage N3 sleep in the occipital andcentral areas; and features that most positively correlate with agewere theta band (4e8 Hz) powers in stage N1 sleep in frontal, oc-cipital, and central areas. The details are described in Table S6.

Using this model-derived age norm, we could examine reasonsfor deviation from CA quantitatively. For example, Fig. 5A shows theEEG spectrogram of a 29-year-old female with multiple medicalproblems, including diabetes mellitus, obesity, smoking, a cere-brovascular accident, congestive heart failure, and acute renal fail-ure. Her brain activity resembled that of a much older adult. Her BAwas 63 years. As shown in Fig. 5B, when compared to the age norm,her sleep EEG exhibited (1) less delta to theta ratio during deep N3sleep; and (2) less theta and alpha band bursts, that is, morecontinual theta and alpha, during N2 sleep.

Note that our results only showed that BAI reflects age and pa-thology at the population level. Nevertheless this example sug-gested the possibility that BAI may, with further studies, serve as apersonalized biomarker of brain health. Other examples are shownin Figures S5eS7 in the supplementary material.

4. Discussion

The concept of differential agingdthe idea that a 50-year-oldcan “have the heart of a 20-year-old” or that the lungs of a 30-year-old smoker “work like they’re 80”dis useful clinically and is gainingscientific traction. Similarly, while cognitive decline is a “normal”part of aging, some individuals clearly age better than others. In thiswork, we have developed a machine learning model that leveragesstatistical changes across the adult lifespan in the pattern of brainactivity during sleep. We call the predicted age "brain age" or BA.Model predictions are highly correlated with CA (r ¼ 0.83, MAD ¼7.6 years) for healthy participants, and at the population level, themodel accurately tracks advancing age. More interestingly, thisstudy presents preliminary evidence that excess BA (EEG BA inexcess than chronological age) reflects underlying brain pathology.In particular, our findings show that significant neurological orpsychiatric disease accelerate brain aging, as do hypertension anddiabetes. These findings suggest that features of brain activityduring sleep, reflected in EEG, can be harnessed using machinelearning to provide an easily accessible and low-cost biomarker ofbrain health.

One strength of our study is the use of large data sets: 2532 sleepEEGs from the MGH Sleep Lab, and 1974 sleep EEGs from the SHHSdata set (Dean et al., 2016; Quan et al., 1997; Redline et al., 1998).

A B

Fig. 4. (A) Histograms of BAIs for participants with (red) and without (blue) significant neurological or psychiatric disease defined in Table S1. The dashed lines show the Gaussian fitto these distributions. The difference of group means is 4 years. *p-value ¼ 6.1 $10"7 < 0.05. (B) Histograms of BAIs for participants with (red) and without (blue) hypertension anddiabetes. The difference of group means is 3.5 years. *p-value ¼ 1.8 $ 10"4 < 0.05. (For interpretation of the references to color in this figure legend, the reader is referred to the Webversion of this article.)

H. Sun et al. / Neurobiology of Aging 74 (2019) 112e120 117

Meditation and Sleep

– ↑ sleep quality, efficiency Khalsa S. Appl Psychophysiol Biofeedback, 2004 (Yoga)

– Cancer patients (Mindfulness)• Reduce total wake time• Improve Sleep Quality

– ↑Sleep Quality in old (Tai Chi)

– Cognitive Based Therapy-Insomnia

– ↓GABA and inflammatory markers- Streeter CC. J AlternComplement Med, 2010

– Higher Melatonin levels

– Mimic Natural Sleep

10/15/19

12

begins with stage N1, transitions to stage N2 followed bystage N3 and on to REM sleep. The cycle then beginsagain and is repeated approximately every 90–120 min.Each of these sleep stages, or a combination, may beimportant for specific health benefits [7]. N1 sleep ischaracterized by the loss of occipital alpha (8–12 Hz)oscillations that are associated with relaxed wakefulness,and decreased beta (13–25 Hz) oscillation power(Figure 1a) [7]. N2 sleep is characterized by slow-deltaoscillations, K-complexes, and sleep spindle oscillations(12–16 Hz) (Figure 1b) [7,25!!]. K-complexes are tran-sient low-frequency oscillations that reflect reducedcortical neuronal activity [7,25!!]. Spindle oscillationsreflect rebound bursting in thalamocortical neurons[7,25!!]. Slow-delta oscillations, which are larger inamplitude than N2 sleep slow oscillations, dominateN3 sleep (Figure 1c) [7,25!!]. Slow-delta oscillationsare a result of profound cortical and thalamic hyperpo-larization [7,25]. They likely result from decreasedbrainstem inputs to the thalamus and cortex due toGABAergic and galanergic inhibition of brainstemarousal nuclei [7–9,10!!]. REM sleep isassociated with an activated EEG pattern that iscomprised of mixed frequencies, and the absence ofK-complexes or sleep spindles (Figure 1d) [7,25!!].

EEG Oscillations during general anesthesiaSlow-delta oscillationsSlow-delta oscillations are a shared feature of sleep and GA(Figures 1 and 2). Recent studies suggest that slow-deltaoscillations identical to those observed in vivo require anintact thalamocorticalnetwork[26!].This is consistentwiththe slow-delta oscillations that were first observed in corti-cal EEG by Frederic Bremer in the 1930s when he per-formed a surgical transection at the level of the mesen-cephalon in cats [27]. This brain preparation, which severedexcitatory brainstem inputs to the thalamus and cortex, wastermed cerveau isole (isolated cerebrum) [27].

Slow-delta oscillations are also observed when GA isinduced or maintained with anesthetics (propofol, ethers)that facilitate inhibitory post-synaptic currents (IPSCs) byenhancing GABAA receptor decay-time and conductance(Figure 2a and b) [3,4,19,22!!,28!,29–33]. Similar to sleep,a mechanism to explain these GABAergic slow-deltaoscillations likely involves loss of excitatory inputs frombrainstem arousal centers to the cortex [7,10!!]. However,numerous lines of evidence suggest that inhibition ofbrainstem arousal nuclei that send excitatory inputs to thecortex is not sufficient to explain GA-induced slow-deltaoscillations.

General anesthesia and sleep neural oscillations Akeju and Brown 179

Figure 1

Non-REM stage N1 Non-REM stage N2

Non-REM stage N3

0

102030

10 20 300

102030

10 20 30

-15

15

0

102030

0

102030

REM sleep

10 20 3010 20 30Time (sec) Time (sec)

Time (sec)Time (sec)

Freq

uenc

y (Hz

)Fr

eque

ncy (

Hz)

Freq

uenc

y (Hz

)Fr

eque

ncy (

Hz)

Powe

r (dB

)

1sec50m

v

(c) (d)

(b)(a)

Current Opinion in Neurobiology

Sleep stages have distinct EEG signatures that result from differences in the neural circuits that are involved in their generation and maintenance.The spectrogram, which is the decomposition of the EEG signal by frequency as a function of time, makes these differences clear. Thesesignatures are also visible in the raw EEG signal (black traces represent first 10 s of data shown in spectrogram). (a) EEG slowing and the loss ofthe awake state alpha oscillations are distinguishing features of N1 sleep. (b) Slow-delta (0.1–4 Hz) oscillations, K-complexes (black arrow onspectrogram and raw EEG), and spindle oscillations (12–16 Hz, red arrow on spectrogram and raw EEG) are distinguishing features of N2 sleep.(c) The predominance of slow-delta oscillations is a distinguishing feature of N3 sleep. (d) Activated ‘saw-tooth’ EEG without the awake-statealpha oscillations are distinguishing features of REM sleep.dB, decibels; EEG, electroencephalogram; Hz, Hertz; N1, non-rapid eye movement stage 1 sleep; N2, non-rapid eye movement stage 3 sleep; N3,non-rapid eye movement stage 3 sleep; REM, rapid eye movement.

www.sciencedirect.com Current Opinion in Neurobiology 2017, 44:178–185

Sleep slow-delta oscillations are synchronous with rela-tively brief periods of inactivity in cortical neurons [34].In contrast, slow-delta oscillations observed under pro-pofol-induced slow-delta oscillations are asynchronouswith prolonged periods of inactivity in cortical neurons[19]. A source localization study of propofol GA and sleepdescribed more putative cortical sources for propofol GAslow-delta oscillations compared to sleep [31]. This sug-gests that significantly greater enhancement of IPSCs incortical circuits could in part explain the functional dif-ferences between sleep and GA-induced slow-deltaoscillations.

An empirical clinical observation supports the notion thatGA slow-delta oscillations result from significantenhancement of IPSCs in cortical circuits. The amplitudeof GA slow-delta oscillation increases in a dose dependentfashion until a plateau is reached [32], after which,increased drug administration results in burst suppres-sion, which is characterized by brief bursts of spikes, sharpwaves, or slow waves of relatively high amplitude alter-nating with periods of relatively flat EEG, termed iso-electricity [3,20,22!!]. Further drug increases results in a

brain state that is characterized by only isoelectric periods(Figure 2c) [3,20,22!!]. Notably, burst suppression andisoelectricity are brain states of relative cortical quies-cence that are not observed during normal sleep or incerveau isole preparations [27]. Rather, they are closelyassociated with cortical pathologies such as diffuse anoxicbrain injury, hypothermia and Ohtahara syndrome [3,35].

Alpha oscillationsAlpha oscillations are not a characteristic feature of non-REM sleep [7]. This is highlighted by the fact that HansBerger, the inventor of electroencephalogram, first dis-tinguished an ‘awake restful with eyes closed’ state fromsleep by associating occipital alpha oscillations (8–12 Hz)with wakefulness [36]. Although not a part of the empiri-cal characterization of sleep, short bursts of alpha power,which likely reflect micro-arousals or complete arousals,have been described during REM sleep [25!!,37,38].Anesthetics that enhance IPSCs through GABAA receptortargeting induce and maintain strong alpha oscillations(Figure 2a and b). However, in contrast to the occipitalalpha oscillations that are associated with wakefulness,GABAA associated alpha oscillations are frontally

180 Neurobiology of sleep

Figure 2

Propofol General Anesthesia Sevoflurane General Anesthesia

Sevoflurane General Anesthesia - isoelectricity Ketamine General Anesthesia

0

20

40

Freq

uenc

y (Hz

)

0

20

40

Freq

uenc

y (Hz

)

0

20

40

Freq

uenc

y (Hz

)

0

20

40

Freq

uenc

y (Hz

)

10 20 30Time (sec) 10 20 30

Time (sec)

10 20 30Time (sec)

10 20 30Time (sec)

0

-25

20

Powe

r (dB

)

0

-25

5

Powe

r (dB

)

1sec50m

v

(c) (d)

(b)(a)

Current Opinion in Neurobiology

General anesthesia. Each anesthetic drug has a different EEG signature that results from differences in the neural circuits that are involved in stategeneration and maintenance. The spectrogram, which is the decomposition of the EEG signal by frequency as a function of time, makes thesedifferences clear. These signatures are also visible in the raw EEG signal (black traces represent first 10 s of data shown in spectrogram). (a)Slow-delta (0.1–4 Hz) and alpha (8–12 Hz) oscillations are the predominant EEG signatures of propofol-general anesthesia. This finding isconsistent with the EEG signatures of other intravenous GABAA receptor targeting anesthetics (i.e., benzodiazepines, etomidate) during generalanesthesia. (b) Slow-delta oscillations, theta (4–8 Hz), and alpha oscillations are the predominant EEG signatures of sevoflurane-generalanesthesia. This finding is consistent with the EEG signatures of other modern day derivatives of ether during general anesthesia (desflurane,isoflurane). The close similarities between the EEG signatures of propofol and modern day derivatives ether anesthesia has been suggested toresult from enhancement of GABAA receptor IPSCs. (c) Isoelectricity is observed when high doses of anesthetics such as sevoflurane andpropofol are administered. Significantly enhancement of IPSCs in cortical circuits is a mechanism to explain isoelectricity. (d) Gamma oscillations("30–45 Hz) that are interspersed with slow-delta (black arrow on spectrogram and raw EEG) oscillations are the predominant EEG signatures ofgeneral anesthesia maintained with the NMDA receptor antagonist ketamine.dB, decibels; EEG, electroencephalogram; GABAA, gamma amino butyric acid A; Hz, Hertz; IPSCs, inhibitory post synaptic currents; NMDA, N-methyl-D-aspartate.

Current Opinion in Neurobiology 2017, 44:178–185 www.sciencedirect.com

10/15/19

13

dominant, larger in amplitude, and highly coherent[18,21,29,30,33,39,40].

A model to explain the generation of anesthesia-inducedcoherent alpha oscillations suggests that an increase inGABAA decay-time and conductance results in corticalalpha oscillatory dynamics [18]. At the same time,increased GABAA decay-time and conductance in thethalamus leads to enhanced rebound spiking of relaycells, strengthening the intrinsic alpha oscillatorydynamic of the thalamus [18]. The net result is reciprocalthalamic-cortical alpha oscillation coupling [18]. It isimportant to note that GA-induced alpha oscillationsoccur in a frequency range and spatial distribution thatis similar to sleep spindles (12–16 Hz)[18,21,29,30,33,39,40]. However, sleep spindles have atransient envelope with a refractory period of severalseconds [7,25], whereas GA-induced alpha oscillationsdo not exhibit such a refractory period[18,21,29,30,33,39,40]. Further, GA alpha oscillationshave only been found to occur with slow-delta oscillations[[21,28!,29,30,32,33,40]21,22,28!,29,30,32,33,40], and theamplitude of GA alpha oscillations is modulated signifi-cantly by the phase of the slow-delta oscillations [33].

Gamma oscillationsGamma oscillations (>30 Hz) or ‘active EEG’ are a char-acteristic feature of REM sleep (Figure 2d) [7,25!!].Ketamine, which is an N-methyl-D-aspartate (NMDA)receptor antagonist commonly administered to produceGA, is also associated with gamma oscillations [41]. How-ever, REM sleep gamma oscillations are distinct fromketamine-GA gamma oscillations. Ketamine effectivelyblocks excitatory NMDA receptors on fast-spiking corti-cal interneurons to down regulate interneuron activity,

and decrease GABA release at the interneuron–pyramidalneuron junction [42,43]. The net result of decreasedcortical inhibitory tone is markedly excited pyramidalneuronal activity [42,43]. A prediction of the ketamine-induced ‘excited’ cortical state is high-frequency neuraloscillations. Consistent with this prediction gamma oscil-lations, which are tightly constrained to a narrow fre-quency band ("30–45 Hz), are a characteristic featureof ketamine-GA is (Figure 2d) [41!]. Ketamine-GAgamma oscillations alternate with slow-oscillations in acharacteristic ‘gamma-burst’ pattern (Figure 2d) [41!,44].With decreasing drug concentrations, the slow-delta oscil-lations are first to dissipate [41!]. It is possible that theketamine slow-delta oscillations result in part fromreduced activity of glutamatergic projections from theparabrachial nucleus to the basal forebrain and the thala-mus [10!!], or by direct drug action in the thalamus[45,46]. However, mechanisms to explain the dynamicsof ketamine-induced slow and gamma oscillations are anarea of active investigation.

EEG oscillations during anesthesia-inducedbrain states of sedationCompared to GA, sedation is a closer behavioral approxi-mation to sleep because arousal to consciousness can beachieved from sedated states. However, sedatives do notproduce physiological sleep.

Beta oscillationsBeta oscillations are associated with sedative statesinduced by anesthetics that enhance IPSCs throughGABAA receptor targeting (Figure 3a) [33,47,48]. Incontrast, non-REM sleep is associated with a decreaseof beta oscillations (Figure 1) [49]. A model to explainsedation-associated beta oscillations suggests that they

General anesthesia and sleep neural oscillations Akeju and Brown 181

Figure 3

0102030

Freq

uenc

y (Hz

)0

102030

Freq

uenc

y (Hz

)

10 20 30Time (sec) 10 20 30Time (sec)

Propofol Sedation Dexmedetomidine Sedation

1sec

50m

v

0

-25

5

Powe

r (dB

)

(b)(a)

Current Opinion in Neurobiology

Sedation states. Each anesthetic drug has a different EEG signature that results from differences in the neural circuits that are involved in stategeneration and maintenance. The spectrogram, which is the decomposition of the EEG signal by frequency as a function of time, makes thesedifferences clear. These signatures are also visible in the raw EEG signal (black traces represent first 10 s of data shown in spectrogram). (a) Beta("13–30 Hz) oscillations are the predominant EEG signature of sedation maintained by propofol and other medications that enhance GABAA

receptor IPSCs (i.e., ether anesthetics, benzodiazepines, zolpidem). (b) Slow-delta and spindle (12–16 Hz; red arrow on spectrogram and raw EEG)oscillations are the predominant EEG signatures of dexmedetomidine-sedation. These dexmedetomidine-induced EEG signatures very closelyapproximate the EEG signatures of N2 sleep (Figure 1b).dB, decibels; EEG, electroencephalogram; GABAA, gamma amino butyric acid A; Hz, Hertz; IPSCs, inhibitory post synaptic currents.

www.sciencedirect.com Current Opinion in Neurobiology 2017, 44:178–185

182 Neurobiology of sleep

Figure 4

.. .. . .. ..NE

Ach

Glu

GluAch 5HT

DA

His

Ox

Thal

Interneuron Pyramidal neurons

.. .. .. .. Ach5HT

DAHis

Ox

Thal

Interneuron Pyramidal neurons

Glu

Ach

.Thal

InterneuronPyramidal neuronsKetamine

Glu

Glu

.........

Locus ceruleusPreoptic areaTuberomamillary nucleus

Perifornical HypothalmusVentral periaqueductal grayRaphe nucleus

Parabrachial nucleus

Pedunculopontine/Lateral dorsal tegmentum

Basal forebrain

(b) GABAergic GA: large amplitude slow-delta and coherent frontal alpha oscillations

(c) Ketamine GA: slow-delta interpersed with gamma oscillations (d) Dexmedetomidine sedation: slow-delta and spindle oscillations

(e)

.. ......

......+

Propofol&Ethers

(a) Non-REM sleep: slow-delta and spindle oscillations

+++++

GA

Sedation

N1 N2 Dex

N3 REM

Sleep medicines (i.e. zolpidem, benzodiazepines)

Propofol Ether anesthetics

Ether anesthetics Propofol Ketamine

Phys

iolog

ical

oscil

lation

s No

nphy

siolog

ical

oscil

lation

s

Arousable Unarousable

Sleep

GABA and galaninefferent projections (active)

Norepinephrine efferent projections (inactive)

Anesthetic drug effectDecreased cortical andsubcortical excitatoryneurotransmitter

Sedation

N1 N2 22222222222Dex

N3 REM

Sleep medicines (i.e. zolpidem, benzodiazepines)

Propofol Ether anesthetics

EPropofol K

Sleep REM

... .. . .. ..

NE

Ach

Glu

GluAch 5HT

DA

His

Ox

Thal

Interneuron Pyramidal neurons..Dexmedetomidine

Net cortical Effect Net cortical Effect

Net cortical Effect Net cortical Effect

NEGlu

Current Opinion in Neurobiology

Anesthetics produce unconsciousness by acting on subcortical and cortical circuits. (a) A mechanism to explain non-REM sleep is GABA- andgalanin-mediated inhibition of brainstem arousal nuclei from sleep active cells in the preoptic area of the hypothalamus. Decreased firing rate oflocus ceruleus cells and norepinephrine release from efferent projections to the preoptic area is a putative mechanism to explain activation ofsleep active cells and non-REM sleep initiation. Decreased norepinephrine to the cortex, thalamus, and basal forebrain also contribute todecreased arousal. Other sites that are not depicted (i.e., cells in the parafacial zone) may play important roles in non-REM sleep. (b) Propofol,ether anesthetics, and other medications that significantly enhance GABAA receptor IPSCs decrease arousal by enhancing the inhibitory activity ofGABAergic interneurons in the cortex and thalamus. They also inhibit the major excitatory brainstem nuclei that project directly and indirectly tothe cortex. (c) Ketamine likely decreases the level of arousal by directly targeting the thalamus and also by blocking glutamatergic projections

Current Opinion in Neurobiology 2017, 44:178–185 www.sciencedirect.com

10/15/19

14

Low-dose nocturnal dexmedetomidine prevents ICU delirium. A Randomized, Placebo-controlled Trial. Am J Respir Crit Care Med 2018

during their ICU stay, is unlikely to havebeen influenced by the level of sedationmaintained during the study (5). Low-dosedexmedetomidine was safe. The way in

which dexmedetomidine may haveinfluenced sleep quality, and therelationship between sleep quality anddelirium incidence, remain unclear.

To date, systematic reviews ofrandomized controlled trials have failed toidentify a safe and effective pharmacologicstrategy to either prevent or treat deliriumin critically ill adults (8, 14, 15). Themethodological face-value challengeswithin these studies are numerous. In thestudies describing interventions such ashaloperidol or simvastatin administration,both delirious and nondelirious patientswere enrolled, blurring the differencesbetween prevention and treatment (30, 31).Moreover, some publications combinecoma and delirium as a spectrum of“brain failure,” combining iatrogenicand potentially preventable drug-inducedexcessive sedation with delirium symptoms(32). Studies with antipsychotics havebeen disappointing, not only for theirineffectiveness but also for their failure todemonstrate any impact on outcomesconsidered meaningful by patients, suchas mortality or discharge destination (33).Finally, no clinical interventional studiesevaluating a pharmacologic preventionstrategy, ours included, have consideredpatient perceptions of the often nightmare-like symptoms associated with delirium

Placebo

Dexmed.

0.0

0

50 20 4 3 1 1 1 01122482450

7 14 21 28 35 42 49

0.2

0.4

0.6

0.8

1.0

Number of patients remaining in the ICU still free of delirium

Cum

ulat

ive

prop

ortio

n of

pat

ient

s w

ithou

t del

irium

Follow-up days

Dexmedetomidine

Log-rank

p=0.0063

Placebo

Figure 2. Kaplan-Meier curve for the time to the first occurrence of delirium between dexmedetomidine(dexmed.) and placebo groups during the ICU stay for those patients still at risk for developingdelirium each day for the first time (log rank P value = 0.006). Patients with persistent coma onany given day were deemed not to have delirium.

0

5

10

15

20

25

30

35

40

1 2 3 4 5 6 7 8 9 10

Pre

vale

nce

of d

eliri

um (%

)

Days After Enrollment

Dexmedetomidine

Placebo

Number of patientsremaining in the ICU

Dexmedetomidine group 50 48 46 43 41 39 36 32 29 25

Placebo group 50 49 46 42 38 36 34 31 25 23

Figure 3. Daily prevalence of delirium between dexmedetomidine and placebo groups over the first 10 study days. All patients were free of delirium atbaseline. The number of patients evaluated at each time point differs from the Kaplan-Meier analysis, as patients were not removed from the risk pool atfirst delirium onset. This figure is presented for descriptive purposes only; delirium prevalence on individual days was not compared statistically, given theconcerns associated with multiple testing.

ORIGINAL ARTICLE

1152 American Journal of Respiratory and Critical Care Medicine Volume 197 Number 9 | May 1 2018

STUDY PROTOCOL Open Access

The prevention of delirium in elderly withobstructive sleep apnea (PODESA) study:protocol for a multi-centre prospectiverandomized, controlled trialJean Wong1* , David Lam1, Stephen Choi2, Mandeep Singh1,3, Naveed Siddiqui4, Sanjeev Sockalingam5

and Frances Chung1

Abstract

Background: Delirium is a common problem that occurs in 5–50% of elderly individuals following surgery. Patientswho develop delirium after surgery are at increased risk for serious complications. Recent studies suggest that patientswith obstructive sleep apnea (OSA), a sleep disorder characterized by repeated episodes of complete or partial blockageof the upper airway – are at greater risk to develop delirium. OSA is more common in elderly individuals but is oftenundiagnosed. Identification and treatment of unrecognized OSA may reduce the incidence of postoperative delirium.However, few studies have investigated the effect of perioperative treatment of OSA to prevent postoperative delirium.

Methods: This multi-centre randomized controlled trial will enrol 634 elderly patients undergoing elective hip/kneereplacement surgery. The study has been approved by the Research Ethics Boards of the three participating institutions.Patients will be screened with the STOP-Bang questionnaire. Those with a score of 3 or greater will have a portable homesleep study using the ApneaLink™ Air device. Patients identified to have OSA will be randomized to 1) Auto-titratingcontinuous positive airway pressure (APAP) applied during sleep for 72 h after surgery or until discharge if they aredischarged before 72 h or 2) Control group – routine care, no APAP. All patients will be evaluated for delirium for 72 hafter surgery or until discharge if they are discharged before 72 h. The primary outcome is the occurrence of delirium –assessed twice daily using the Confusion Assessment Method for 72 h or until discharge if the hospital stay is <72 h.

Discussion: Delirium is associated with increased morbidity and mortality, and higher healthcare costs. With the agingpopulation, the incidence of postoperative delirium will likely increase as the number of elderly individuals undergoingsurgery rises. The results of our study will be published in a peer-reviewed journal and presented at local and internationalmedical conferences. Our study findings may lead to improved surgical outcomes, enhanced patient safety and reducedhealthcare costs.

Trial registration: This study was retrospectively registered at clinicaltrials.gov NCT02954224 on November 3, 2016.

Keywords: Aged, Continuous positive airway pressure, Delirium, Sleep apnea, Obstructive, Perioperative period

* Correspondence: jean.wong@uhn.ca1Department of Anesthesia, Toronto Western Hospital, University HealthNetwork, University of Toronto, 399 Bathurst Street, Toronto, ON M5T 2S8,CanadaFull list of author information is available at the end of the article

© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Wong et al. BMC Anesthesiology (2018) 18:1 DOI 10.1186/s12871-017-0465-5

Sleep/Wake Protocol Implementation to Improve Sleep Quality in the ICU (NCT03313115)1.Contributors: Joseph TonnaDate created: 2018-08-21 03:01 PM | Last Updated: 2019-05-11 04:47 PM

Suvorexant and Sleep/Delirium in ICU PatientsEikermann M et al. Orexin receptor antagonist -blocks awake state

Original Article

A randomised trial of peri-operative positive airway pressure forpostoperative delirium in patients at risk for obstructive sleepapnoea after regional anaesthesia with sedation or generalanaesthesia for joint arthroplasty*J. W. Nadler,1 J. L. Evans,2 E. Fang,3 X. A. Preud’Homme,4 R. L. Daughtry,5 J. B. Chapman,6

M. P. Bolognesi,7 D. E. Attarian,8 S. S. Wellman9 and A. D. Krystal10,11

1 Assistant Professor of Anaesthesiology, University of Rochester, Rochester, NY, USA2 Resident, University of Pittsburgh Medical Center, Pittsburgh, PA, USA3 Resident, Singapore General Hospital, Singapore4 Assistant Professor of Psychiatry, 5 Technician Coordinator, 6 Certified Registered Nurse Anaesthetist, 7 Professor ofOrthopaedic Surgery, 8 Professor of Orthopaedic Surgery, 9 Assistant Professor of Orthopaedic Surgery, 10 AdjunctProfessor of Psychiatry, Duke University, Durham, NC, USA11 Professor of Psychiatry, Executive Vice Chair for the Langley Porter Psychiatric Institute (LPPI), University ofCalifornia at San Francisco, San Francisco, CA, USA

SummaryPrevious pilot work has established an association between obstructive sleep apnoea and the development of acutepostoperative delirium [1–3], but it remains unclear to what extent this risk factor is modifiable in the ‘real world’peri-operative setting. In a single-blind randomised controlled trial, 135 elderly surgical patients at risk for obstruc-tive sleep apnoea were randomly assigned to receive peri-operative continuous positive airway pressure (CPAP) orroutine care. Of the 114 patients who completed the study, 21 (18.4%) experienced delirium. Delirium was equallycommon in both groups: 21% (12 of 58 subjects) in the CPAP group and 16% (9 of 56 subjects) in the routine caregroup (OR = 1.36 [95%CI 0.52–3.54], p = 0.53). Delirious subjects were slightly older – mean (SD) age 68.9 (10.7)vs. 64.9 (8.2), p = 0.07 – but had nearly identical pre-operative STOP-Bang scores (4.19 (1.1) versus 4.27 (1.3),p = 0.79). Subjects in the CPAP group used their devices for a median (IQR [range]) of 3 (0.25–5 [0–12]) nightspre-operatively (2.9 (0.1–4.8 [0.0–12.7]) hours per night) and 1 (0–2 [0–2]) nights postoperatively (1.4 (0.0–5.1 [0.0–11.6]) hours per night). Among the CPAP subjects, the residual pre-operative apnoea–hypopnea index had a signifi-cant effect on delirium severity (p = 0.0002). Although we confirm that apnoea is associated with postoperative delir-ium, we did not find that providing a short-course of auto-titrating CPAP affected its likelihood or severity.Voluntary adherence to CPAP is particularly poor during the initiation of therapy..................................................................................................................................................................

Correspondence to: J. W. NadlerEmail: jacob_nadler@urmc.rochester.eduAccepted: 17 January 2017Keywords: obstructive sleep apnoea; postoperative apnoea; mask CPAP; clinical trial*Presented in part at SLEEP 2014, the 28th Annual Meeting of the Associated Professional Sleep Societies Minneapolis,MN, USA, May-June 2014.

© 2017 The Association of Anaesthetists of Great Britain and Ireland 729

Anaesthesia 2017, 72, 729–736 doi:10.1111/anae.13833

Original Article

A randomised trial of peri-operative positive airway pressure forpostoperative delirium in patients at risk for obstructive sleepapnoea after regional anaesthesia with sedation or generalanaesthesia for joint arthroplasty*J. W. Nadler,1 J. L. Evans,2 E. Fang,3 X. A. Preud’Homme,4 R. L. Daughtry,5 J. B. Chapman,6

M. P. Bolognesi,7 D. E. Attarian,8 S. S. Wellman9 and A. D. Krystal10,11

1 Assistant Professor of Anaesthesiology, University of Rochester, Rochester, NY, USA2 Resident, University of Pittsburgh Medical Center, Pittsburgh, PA, USA3 Resident, Singapore General Hospital, Singapore4 Assistant Professor of Psychiatry, 5 Technician Coordinator, 6 Certified Registered Nurse Anaesthetist, 7 Professor ofOrthopaedic Surgery, 8 Professor of Orthopaedic Surgery, 9 Assistant Professor of Orthopaedic Surgery, 10 AdjunctProfessor of Psychiatry, Duke University, Durham, NC, USA11 Professor of Psychiatry, Executive Vice Chair for the Langley Porter Psychiatric Institute (LPPI), University ofCalifornia at San Francisco, San Francisco, CA, USA

SummaryPrevious pilot work has established an association between obstructive sleep apnoea and the development of acutepostoperative delirium [1–3], but it remains unclear to what extent this risk factor is modifiable in the ‘real world’peri-operative setting. In a single-blind randomised controlled trial, 135 elderly surgical patients at risk for obstruc-tive sleep apnoea were randomly assigned to receive peri-operative continuous positive airway pressure (CPAP) orroutine care. Of the 114 patients who completed the study, 21 (18.4%) experienced delirium. Delirium was equallycommon in both groups: 21% (12 of 58 subjects) in the CPAP group and 16% (9 of 56 subjects) in the routine caregroup (OR = 1.36 [95%CI 0.52–3.54], p = 0.53). Delirious subjects were slightly older – mean (SD) age 68.9 (10.7)vs. 64.9 (8.2), p = 0.07 – but had nearly identical pre-operative STOP-Bang scores (4.19 (1.1) versus 4.27 (1.3),p = 0.79). Subjects in the CPAP group used their devices for a median (IQR [range]) of 3 (0.25–5 [0–12]) nightspre-operatively (2.9 (0.1–4.8 [0.0–12.7]) hours per night) and 1 (0–2 [0–2]) nights postoperatively (1.4 (0.0–5.1 [0.0–11.6]) hours per night). Among the CPAP subjects, the residual pre-operative apnoea–hypopnea index had a signifi-cant effect on delirium severity (p = 0.0002). Although we confirm that apnoea is associated with postoperative delir-ium, we did not find that providing a short-course of auto-titrating CPAP affected its likelihood or severity.Voluntary adherence to CPAP is particularly poor during the initiation of therapy..................................................................................................................................................................

Correspondence to: J. W. NadlerEmail: jacob_nadler@urmc.rochester.eduAccepted: 17 January 2017Keywords: obstructive sleep apnoea; postoperative apnoea; mask CPAP; clinical trial*Presented in part at SLEEP 2014, the 28th Annual Meeting of the Associated Professional Sleep Societies Minneapolis,MN, USA, May-June 2014.

© 2017 The Association of Anaesthetists of Great Britain and Ireland 729

Anaesthesia 2017, 72, 729–736 doi:10.1111/anae.13833

10/15/19

15

Melatonin and Sleep in Preventing HospitalizedDelirium: A Randomized Clinical TrialStuti J. Jaiswal, MD, PhD,a,b Thomas J. McCarthy, MD,b Nathan E. Wineinger, PhD,a Dae Y. Kang, PhD,c Janet Song,a

Solana Garcia,a Christoffel J. van Niekerk, MD,b Cathy Y. Lu,a Melissa Loeks, MPH,d Robert L. Owens, MDc

aThe Scripps Research Institute, La Jolla, Calif; bDepartment of Internal Medicine, Scripps Clinic/Scripps Green Hospital, La Jolla, Calif;cDivision of Pulmonary, Critical Care & Sleep Medicine, University of California San Diego School of Medicine, La Jolla, Calif;dRespironics, Inc., a Philips Healthcare Company, Murrysville, Pa.

ABSTRACT

PURPOSE: Studies suggest that melatonin may prevent delirium, a condition of acute brain dysfunctionoccurring in 20%-30% of hospitalized older adults that is associated with increased morbidity and mortal-ity. We examined the effect of melatonin on delirium prevention in hospitalized older adults while measur-ing sleep parameters as a possible underlying mechanism.METHODS: This was a randomized clinical trial measuring the impact of 3 mg of melatonin nightly on inci-dent delirium and both objective and subjective sleep in inpatients age !65 years, admitted to internalmedicine wards (non-intensive care units). Delirium incidence was measured by bedside nurses using theconfusion assessment method. Objective sleep measurements (nighttime sleep duration, total sleep timeper 24 hours, and sleep fragmentation as determined by average sleep bout length) were obtained via actig-raphy. Subjective sleep quality was measured using the Richards Campbell Sleep Questionnaire.RESULTS: Delirium occurred in 22.2% (8/36) of subjects who received melatonin vs in 9.1% (3/33) whoreceived placebo (P = .19). Melatonin did not significantly change objective or subjective sleep measure-ments. Nighttime sleep duration and total sleep time did not differ between subjects who became deliriousvs those who did not, but delirious subjects had more sleep fragmentation (sleep bout length 7.0 § 3.0 vs9.5 § 5.3 min; P = .03).CONCLUSIONS: Melatonin given as a nightly dose of 3 mg did not prevent delirium in non-intensive careunit hospitalized patients or improve subjective or objective sleep.! 2018 Elsevier Inc. All rights reserved. ! The American Journal of Medicine (2018) 131:1110"1117

KEYWORDS: Delirium; Sleep; Melatonin

INTRODUCTIONDelirium occurs in an estimated 20%-30% of elderlypatients admitted to hospital wards.1 Individuals !65 years,especially those with dementia, evidence of prior cognitivedecline, or prior neurologic insult such as stroke, are atincreased risk. Major consequences of delirium includeincreased in-hospital2 and 1-year mortality rates,3 as wellas worsened neurocognitive outcomes.4-6 Delirious patientsstay hospitalized longer7,8 and are more likely to dischargeto a facility instead of home.9 No US Food and DrugAdministration"approved therapies exist for the treatmentor prevention of delirium, although clinicians frequentlyprescribe antipsychotics and anxiolytics to control associ-ated agitation, despite data clearly demonstrating their lackof efficacy and risk of serious side effects.10 Sleep depriva-tion has been hypothesized as a cause of hospital delirium,

Funding: This research was funded in part by a National Institutes of

Health/National Center for Advancing Translational Sciences flagship Clinicaland Translational Science Award Grant (KL2 TR001112 and UL1 TR001114),

and a Scripps ClinicMedical Group Research and EducationGrant.

Conflicts of Interest: None. Although ML is employed by Philips

Respironics, the study design and interpretation of all data were done bythe investigators.

Authorship: All of the authors contributed to this article by participa-

tion in the research and preparation of the manuscript. Additionally, allauthors had access to the data and a role in writing the manuscript.

Trial Registration: Registered at CLINICALTRIALS.GOV:

NCT02597231.

Data Statement: The de-identified data for this study are availableupon reasonable request to the authors via e-mail to jaiswal.stuti@scripp-

shealth.org.

Requests for reprints should be addressed to Stuti J. Jaiswal, MD, PhD,

The Scripps Research Institute, Scripps Translational Science Institute,3344 N Torrey Pines Ct, Suite 300, La Jolla, CA 92037.

E-mail address: jaiswal.stuti@scrippshealth.org

0002-9343/© 2018 Elsevier Inc. All rights reserved.https://doi.org/10.1016/j.amjmed.2018.04.009

CLINICAL RESEARCH STUDY

Not effective

Impact of Melatonin on Sleep and Pain After Total Knee ArthroplastyUnder Regional Anesthesia With Sedation: A Double-Blind, Randomized,Placebo-Controlled Pilot Study

Meghan A. Kirksey, MD, PhD a,1,2,3, Daniel Yoo, MB a,4,5,6, Thomas Danninger, MD a,b,7,8,9,Ottokar Stundner, MD a,b,10,11, Yan Ma, PhD c,12,13,14, Stavros G. Memtsoudis, MD, PhD a,15,16,17

a Hospital for Special Surgery, New York, New Yorkb Paracelsus Medical University, Salzburg, Austriac George Washington University, Washington, DC

a b s t r a c ta r t i c l e i n f o

Article history:Received 26 February 2015Accepted 18 June 2015

Keywords:melatoninsleeparthroplastypainactigraphy

This pilot study explores sleep disruption after total knee arthroplasty and the impact of melatonin on sleep andpostoperative pain. Sleep time was decreased on the last preoperative night and first two postoperative nights.Sleep efficiency was decreased on all three postoperative nights. Compared to placebo, melatonin increasedsleep efficiency by 4.4% (mean; 95% CI −1.6, 10.4; P = 0.150) and sleep time by 29 min (mean; 95% CI −2.0,60.4; P= 0.067). Melatonin appeared to have no effect on subjective sleep quality or daytime sleepiness, pain atrest or painwith standardized activity. In conclusion, sleep quality is impaired after total knee arthroplasty and ex-ogenous melatonin does not appear to improve postoperative sleep or pain to a significant degree.

© 2015 Elsevier Inc. All rights reserved.

Sleep disruption is a challenge commonly faced by patients and careproviders in the perioperative period [1,2] and has been shown to affectpostoperative performance after total knee arthroplasty [3]. Postoperativesleepdisruption is likely influencedby environmental factors [4] and anes-thetic exposure [5,6] and is known to be exacerbated by postoperativepain [7]. In a reciprocal manner, sleep disruption has been shown to exac-erbate pain perception [8,9].

Melatonin is an inexpensive over-the-counter dietary supplementwith an established safety profile [10] that has shown promise in man-aging sleep disorders and amelioration of chronic and acute pain. Evi-dence suggests that exogenous melatonin can be efficacious inimproving sleep disruption in tracheostomized patients in the ICU [11]as well as those experiencing jet lag [12,13].

Previous studies have found conflicting results regarding the poten-tial for melatonin to improve sleep and pain in the perioperative period[14]. These discrepanciesmay result fromdifferences in surgical and an-esthetic conditions, differences in melatonin dose and administrationregimens, variations in study quality, different methods of assessingpain and sleep quality, and different patient populations. There is noconsensus as to what dose, duration, and timing of melatonin adminis-tration in the perioperative period are most likely to improve sleepquality or quantity. This study was designed to explore the effect of astable regimen of exogenous perioperative melatonin, administeredover 6 consecutive nights, on postoperative pain, sleep quality, andsleep efficiency in patients undergoing total knee arthroplasty under re-gional anesthesiawith sedation. To our knowledge, this study is the firstto examine perioperative subjected sleep quality as well as sleep timeand efficiency as measured by the validated objective tool of wristactigraphy in this population.

The Journal of Arthroplasty 30 (2015) 2370–2375

One ormore of the authors of this paper have disclosed potential or pertinent conflictsof interest, which may include receipt of payment, either direct or indirect, institutionalsupport, or association with an entity in the biomedical field which may be perceived tohave potential conflict of interest with this work. For full disclosure statements refer tohttp://dx.doi.org/10.1016/j.arth.2015.06.034.

Institution: Hospital for Special Surgery, New York, NY Weill College of Medicine atCornell University, New York, NY.

Funding: no external funding. The authors gratefully acknowledge funding from theDepartment of Anesthesiology at the Hospital for Special Surgery.

Reprint requests: Meghan A Kirksey, MD, PhD, Hospital for Special Surgery, Departmentof Anesthesiology, Hospital for Special Surgery, 535 E. 70th Street, New York, NY 10021.

1 Role: This author helped design the study, conduct the study, analyze the data, andwrite the manuscript.

2 Conflicts: Meghan A. Kirksey reported no conflicts of interest.3 Attestation:Meghan A. Kirksey has seen the original study data, reviewed the analysis

of the data, and approved the final manuscript.4 Role: This author helped conduct the study and write the manuscript.5 Conflicts: Daniel Yoo reported no conflicts of interest.6 Attestation:Daniel Yoohas seen the original studydata, reviewed the analysis of the data,

approved the final manuscript, and is the author responsible for archiving the study files.7 Role: This author helped conduct the study.8 Conflicts: Thomas Danninger reported no conflicts of interest.9 Attestation: Thomas Danninger has seen the original study data, reviewed the analysis

of the data, and approved the final manuscript.10 Role: This author helped conduct the study.11 Attestation: Ottokar Stundner has seen the original study data, reviewed the analysisof the data, and approved the final manuscript.12 Role: This author helped analyze the data and write the manuscript.13 Conflicts: Yan Ma reported no conflicts of interest.14 Attestation: YanMa has seen the original study data, reviewed the analysis of the data,and approved the final manuscript.15 Role: This author helped design the study, conduct the study, and write the manuscript.16 Conflicts: Stavros Memtsoudis reported no conflicts of interest.17 Attestation: Stavros Memtsoudis has seen the original study data, reviewed the analy-sis of the data, and approved the final manuscript.

http://dx.doi.org/10.1016/j.arth.2015.06.0340883-5403/© 2015 Elsevier Inc. All rights reserved.

Contents lists available at ScienceDirect

The Journal of Arthroplasty

j ourna l homepage: www.ar throp lasty journa l .o rg

Step count on Postoperative Delirium Following Cardiac Surgery

Enrolling… from Clinicaltrials.gov

Thank you

30