Transcript Obstructive Sleep Apnea - UM Anesthesiology

Obstructive Sleep Apnea
Perioperative Implications
From Mechanisms to Risk Modification
Satya Krishna Ramachandran MD FRCA
Assistant Professor of Anesthesiology
University of Michigan Medical School, Ann Arbor
[email protected]
Disclosures
• Paid scientific advisory consultant
– Galleon Pharmaceuticals
– Merck, Sharp & Dohme
• Funding
– PSA with MSD for 2014
– MiCHR CTSA PGP UL1TR000433 for 2014
The material of this talk is independent of
these disclosures
This is a confidential Quality Improvement and Assurance/peer review document of the University of Michigan Hospitals and Health Centers.
Unauthorized disclosure or duplication is absolutely prohibited. This document is protected from disclosure pursuant to the provisions of MCL 333.20175;
MCL 333.21515; MCL 331.531; MCL 331.533 or such other statutes that may be applicable
Goals & Objectives
• To describe the relationship between OSA and
early postoperative respiratory failure
• To review mechanisms of unanticipated early
postoperative respiratory failure
• To critically evaluate methods of riskmodification of early postoperative respiratory
failure
Obstructive Sleep Apnea and
Respiratory Failure
Evidence in the surgical population
• Retrospective studies: associations
–
–
–
–
Gupta – more complications, ICU admissions
Hwang – more morbidity
Memtsoudis – independent increase in morbidity
Mokhlesi – Increased respiratory failure
• Prospective evidence: associations
– Chung – more postoperative desaturation episodes
– Gali – more morbidity with postoperative episodic desat.
• Sudden death – case reports
Gupta. Mayo Clin Proc. 2001;76:897-905
Hwang. Chest. 2008;133:1128-34
Memtsoudis. Anesth Analg. 2011;112:113-21
Gali B. Anesthesiology 2009;110:869-77
Ostermeier. Anesth Analg. 1997;85:452-60
AHI and outcome
Gami. N Engl J Med. 2005;352:1206-14.
Nocturnal pattern in sudden death
Gami. N Engl J Med. 2005;352:1206-14.
Severity of
OSA and
nocturnal
variation in
sudden death
Gami. N Engl J Med. 2005;352:1206-14.
If they are prone to sudden death
during sleep,
is the risk of
postoperative sudden death
increased
in patients with OSA?
Nocturnal Variation In Outcome Of ARE
Postoperative ARE
from RM database
Number of cases (n=32)...
12
Irreversible
Reversible
10
8
6
4
2
0
06:00-11:59
12:00-17:59
18:00-23:59
00:00-05:59
35 cases – 5 deaths
/ 6 years
History or known
risk factors for OSA
present in ~40%
cases
Time of day (24 hour clock)..
Ramachandran SK. J Clin Anesth 2011;23:207-13
Mechanisms of Perioperative AE?
•
•
•
•
•
•
Hypoxia
Sympathetic activation
Cardiovascular variability
Inflammation
Comorbid disease
Chemoceptor hypersensitivity
Somers et al. Circulation. 2008;118:1080-1111
Mechanisms of Perioperative AE?
•
•
•
•
•
•
Hypoxia
Sympathetic activation
Cardiovascular variability
Inflammation
Comorbid disease
Chemoceptor hypersensitivity
Hypoxia and Arrhythmia/Conduction
• Nocturnal ventricular arrhythmias
– Min SpO2<60%
– AHI >65.hr-1
• QRS prolongation
– Min SpO2<90%
– AHI >30.hr-1
• Heart Block
– Min SpO2<90%
– Obesity
Sheppard. Chest. 1985 Sep;88(3):335-40
Valencia-Flores. Obes Res. 2000 May;8(3):262-9.
Ramachandran – unpublished data
Mechanisms of Perioperative AE?
•
•
•
•
•
•
Hypoxia
Sympathetic activation
Cardiovascular variability
Inflammation
Comorbid disease
Chemoceptor hypersensitivity
MSNA and OSA
Somers et al. J Clin Invest. 1995;96:1897-904
MSNA and Sleep Stage
Somers et al. J Clin Invest. 1995;96:1897-904
MSNA In Awake State
Somers et al. J Clin Invest. 1995;96:1897-904
Mechanisms of Perioperative AE?
•
•
•
•
•
•
Hypoxia
Sympathetic activation
Cardiovascular variability
Inflammation
Comorbid disease
Chemoceptor hypersensitivity
Cardiovascular variability
Narkiewicz et al. Circulation. 1998;98:1071-1077
Intrathoracic Pressure Changes
• Repeated Mueller maneuvers during OSA
– Intrathoracic pressures approach -65 mmHg
• ?Increased risk of postoperative pulmonary edema
• Increased transmural gradient across atria
and ventricles
– Increased wall stress and afterload
– Diastolic dysfunction
– Atrial remodeling
Mechanisms of Perioperative AE?
•
•
•
•
•
•
Hypoxia
Sympathetic activation
Cardiovascular variability
Inflammation
Comorbid disease
Chemoceptor hypersensitivity
OSA and Inflammation
• Selective activation of inflammatory pathways
– Hypoxemia
– Sleep deprivation/fragmentation
• Increased levels in OSA
– Cytokines, adhesion molecules, serum amyloid
– C-reactive protein - ?obesity related
– TNF
• May influence postoperative mortality and morbidity
Mechanisms of Perioperative AE?
•
•
•
•
•
•
Hypoxia
Sympathetic activation
Cardiovascular variability
Inflammation
Comorbid disease
Chemoceptor hypersensitivity
Unanticipated Postoperative
Respiratory Failure
• Prediction model in 222,094 patients from the NSQIP
dataset.
• Overall, 49.4% unanticipated tracheal intubations
occurred within first three days after surgery.
• The incidence of unanticipated early postoperative
intubation (UEPI) was 0.83-0.9%
Ramachandran SK et al. Anesthesiology 2011;115:44-53
UEPI Independent Predictors
•
•
•
•
•
•
•
Surgical Type
Current Ethanol Use
Current Smoker
Dyspnea
COPD
Diabetes Mellitus
Active Congestive Heart
Failure
• Hypertension Requiring
Medication
• Abnormal Liver Function
• Cancer
• Prolonged Hospitalization
• Recent Weight Loss
• Body Mass Index < 18.5 Or
≥ 40 Kg/m2
• Sepsis
Ramachandran SK et al. Anesthesiology 2011;115:44-53
UEPI Independent Predictors
•
•
•
•
•
•
•
Surgical Type
Current Ethanol Use
Current Smoker
Dyspnea
COPD
Diabetes Mellitus
Active Congestive Heart
Failure
• Hypertension Requiring
Medication
• Abnormal Liver Function
• Cancer
• Prolonged Hospitalization
• Recent Weight Loss
• Body Mass Index < 18.5 Or
≥ 40 Kg/m2
• Sepsis
Ramachandran SK et al. Anesthesiology 2011;115:44-53
Mechanisms of Perioperative AE?
•
•
•
•
•
•
Hypoxia
Sympathetic activation
Cardiovascular variability
Inflammation
Comorbid disease
Chemoceptor hypersensitivity
OSA and chemoreceptor sensitivity
• Limited adult data
• Postoperative ARE outcomes unrelated to dose
• Opioid consumption lower in patients who died
Ramachandran SK et al. J Clin Anesth 2011;23:207-13
Metabolic Disease and RD?
RISK MODIFICATION
Baseline Risk Reduction Strategies
• Preoperative CPAP
• Opioid sparing techniques
– Regional anesthesia/analgesia
– Non-opioid adjuncts
– Minimal access surgery
• Continuous pulse oximetry monitoring
• Postoperative CPAP
Expert Opinion
PREoperative CPAP
• No RCT guided evidence of perioperative benefit
• Possible mechanisms:
– Less severe nocturnal desaturation
– More dependable postoperative CPAP usage
• Challenges:
– Majority of patients are undiagnosed
– Adherence with therapy is low
– Timely preoperative testing/fitting
Preop CPAP Benefit - MSNA, MAP
Somers et al. J Clin Invest. 1995;96:1897-904
CPAP and QTc Dispersion
• Longitudinal 6-month study of CPAP
• 12-lead ECG data analysis
Dursunoglu et al. Sleep Medicine 2007;8:478–483
CPAP and Arrhythmia in CHF
Ryan et al. Thorax 2005;60:781–785
Cessation of CPAP and MSNA
Somers et al. J Clin Invest. 1995;96:1897-904
UM Model
for Fast
Track PSG
MSQC study
• Introduced a new concept – Preoperative PAP
treatment for OSA
– Implies diagnosis of OSA
– Compliance generally ~50%
• MSQC nurse abstractors collect data from 56
hospitals in Michigan
– Risk adjusted for surgery, comorbid conditions and
intraoperative characteristics
Frequency Tables
Entire Cohort
Sleep Apnea
Freq.
None
32,148
Untreated
1,769
Treated
1,446
Total
35,363
(%)
90.91
5
4.09
100
General Surgery
Sleep Apnea
Freq.
None
20,873
Untreated
1,226
Treated
1,013
Total
23,112
(%)
90.31
5.3
4.38
100
MSQC Analysis
Entire Cohort
Morbidity
Sleep Apnea
None
Untreated
Treated
1.00
1.26
0.87
Pulmonary
Occurence
Sleep Apnea
None
Untreated
Treated
1.00
1.14
0.60
Mortality
Adjusted
Odds Ratio
Sleep Apnea
None
Untreated
Treated
1.00
1.11
0.69
p Value
[95% Conf. Interval]
0.008
0.115
(ref)
1.06 - 1.50
0.72 - 1.04
0.334
0.007
(ref)
0.87 - 1.48
0.42 - 0.87
0.692
0.237
(ref)
0.66 - 1.86
0.37 - 1.28
Multivariate Analysis
Entire Cohort
Morbidity
Sleep Apnea
None
Untreated
Treated
1.00
1.26
0.87
Pulmonary
Occurence
Sleep Apnea
None
Untreated
Treated
1.00
1.14
0.60
Mortality
Adjusted
Odds Ratio
Sleep Apnea
None
Untreated
Treated
1.00
1.11
0.69
p Value
[95% Conf. Interval]
0.008
0.115
(ref)
1.06 - 1.50
0.72 - 1.04
0.334
0.007
(ref)
0.87 - 1.48
0.42 - 0.87
0.692
0.237
(ref)
0.66 - 1.86
0.37 - 1.28
Baseline Risk Reduction Strategies
• Preoperative CPAP
• Opioid sparing techniques
– Regional anesthesia/analgesia
– Non-opioid adjuncts
– Minimal access surgery
• Continuous pulse oximetry monitoring
• Postoperative CPAP
Expert Opinion
Baseline Risk Reduction Caveats
• Opioid sparing techniques
– Reduce opioid consumption
– May not modify respiratory risk
Blake et al. Anesthes Int Care. 2009;37:720-725
Baseline Risk Reduction Strategies
• Preoperative CPAP
• Opioid sparing techniques
– Regional anesthesia/analgesia
– Non-opioid adjuncts
– Minimal access surgery
• Postoperative CPAP
• Continuous pulse oximetry monitoring
Expert Opinion
UM model for Postop CPAP
Risk Modification – Postop CPAP
• Robust evidence for early treatment of hypoxia
– Randomized Controlled Trial of CPAP vs. O2
– Major elective abdominal surgery
• CPAP associated with
– lower intubation rate (1% vs 10%)
– lower occurrence rate of pneumonia (2% vs 10%), infection
(3% vs 10%), and sepsis (2% vs 9%).
• No RCT evidence of benefit of postoperative CPAP in
OSA patients
Squadrone V. JAMA 2005;293:589-595
Baseline Risk Reduction Strategies
• Preoperative CPAP
• Opioid sparing techniques
– Regional anesthesia/analgesia
– Non-opioid adjuncts
– Minimal access surgery
• Postoperative CPAP
• Continuous pulse oximetry monitoring
Expert Opinion
Postoperative Monitoring Overview
• Outcome studies – monitoring success is
limited to recent, small single center studies,
majority evidence points to no benefit.
• Limitations of current state of alarm
technology
• Why universal monitoring may be a problem
Outcome Studies
• 3 tiers of monitoring conceptually:
– Spot monitoring
– Continuous bedside monitoring
– Integrated monitoring /surveillance systems
• Largest studies are of bedside devices
• Majority of current evidence around IM/SS
• Direct comparative effectiveness trials are
impossible in the current climate
Surveillance Systems
Unanswered Questions
• What were the monitoring signatures of
“MET/RRT events”?
• What were the sensitivity and positive
predictive value of the system?
• Did the treatment change the outcome?
Integrated Monitoring System
• An IMS (BioSign; OBS Medical, Carmel, Indiana) used
heart rate, blood pressure, respiratory rate, and
peripheral oxygen saturation by pulse oximetry to
develop a single neural networked signal, or BioSign
INDEX (BSI)
• Data were analyzed for cardiorespiratory instability
according to BSI trigger value and local MET
activation criteria.
Does IMS Prevent Instability?
Does IMS Reduce Frequency
of Instability Events?
Does IMS Reduce
Duration of Instability?
Unanswered Questions
• What were the monitoring signatures of
“MET/RRT events”?
• What is the sensitivity and positive predictive
value of IMS/SS?
• Did the treatment change the outcome?
What is the Sensitivity of EWS?
Unanswered Questions
• What were the monitoring signatures of
“MET/RRT events”?
• What is the sensitivity and positive predictive
value of IMS/SS?
• Did the treatment change the outcome?
– NNT/NNP
– NNH
Does Monitoring Change Outcomes?
• For outcome modification, two things need to
happen:
– The IMS event changes treatment
– The treatment changes the outcome
• Neither was tested in Hravnak’s or Taenzer’s
study
• Both studies used MET/RRT as escalation step
Are we monitoring the right patient
at the right time?
Relationship Between Desaturation &
Unanticipated Respiratory Failure
100
90
80
70
60
AHI
50
Lowest SaO2(%)
40
30
20
10
0
Preop
Night 1
Night 3
Relationship Between Desaturation &
Unanticipated Respiratory Failure
100
90
80
70
60
Unplanned intubation
50
AHI
40
Lowest SaO2(%)
30
20
10
0
Preop
Night 1
Night 3
Can Monitoring Harm?
Summary
• It is possible to predict need for MET/RRT
fairly accurately using advanced monitoring
• MET/RRT intervention does not change
mortality risk
• Risk periods for desaturation and unplanned
intubation are not congruent
• Postoperative monitoring is associated with
increased technological intensification, alarm
fatigue and risk of harm in CURRENT STATE
Future State of Monitoring
• Can only be effective in pathology that is
responsive to treatment
• Shift away from threshold based event
recognition
• Identification of “state change” from healthy
to at-risk state
• Needs to address poor PPV and sensitivity
Conclusions
• OSA is associated with increased risk of early
postoperative respiratory failure
• PREoperative CPAP is associated with
significant physiological benefit
– Compliant PAP therapy is associated with outcome
benefit
• Postoperative monitoring is of unknown value
in OSA patients