Clinical Controversies in Perioperative Medicine
Hugo Quinny Cheng, MD Division of Hospital Medicine University of California, San Francisco
Predicting & Managing Cardiac Risk
A 70-y.o. man with progressive weakness due to cervical myelopathy will have spinal decompression & fusion. He had a drug-eluting stent placed 8 months ago for stable angina. He also has insulin-requiring diabetes and a remote CVA. He uses a walker, needs help with some ADLs.
1. How do you assess his risk for cardiac complications?
2. What about his drug-eluting stent?
3. Should you start a beta-blocker?
70-y.o. with DES placed 8 months ago, IDDM and remote stroke undergoing cervical spine surgery for weakness. How would you estimate this patient’s cardiac risk?
1. I use the Revised Cardiac Risk Index (RCRI), so ~ 10%
2. I use the RCRI, so ~ 5%
3. I use the “NSQIP” prediction tool, so ~ 1%
4. I don’t need a prediction tool; my gut says he’s high risk
Revised Cardiac Risk Index
Predictors: – Ischemic heart disease – Congestive heart failure – Diabetes requiring insulin – Creatinine > 2 mg/dL – Stroke or TIA – “High Risk” operation (intraperitoneal, intrathoracic,
or suprainguinal vascular)
# of RCRI Complications Predictors All
0 0.5%
1 1.3%
2 4%
≥ 3 9%
All: MI, cardiac arrest, complete heart block, pulmonary edema
Devereaux PJ et al. CMAJ 2005; 173:627.
Serious
0.4%
1%
2.4%
5.4%
Serious: MI & cardiac arrest
2007 ACC/AHA Guideline
Good Functional Capacity? Go to OR yes
≥ 3 predictors 1 or 2 predictors no predictors*
no or ?
Control HR & go to OR (IIa)
Vascular surgery?
Consider stress test if results will change management (IIa)
no
yes
or (IIb)
Go to OR
* CAD, CHF, DM, CKD, CVA/TIA
New Cardiac Risk Prediction Tool
Derived from National Surgical Quality Improvement Program (NSQIP) database: • > 400 K patients in derivation & validation cohorts • Wide range of operations • “Complication” = 30-day incidence of MI & cardiac arrest
Independent 1. Type of surgery Predictors 2. Age 3. Serum creatinine > 1.5 mg/dL 4. Functional status (dependency for ADLs) 5. American Society of Anesth (ASA) class
Gupta PK et al. Circulation 2011; 124:681
ASA Class (a brief digression)
American Society of Anesthesiologists Physical Classification 1. Healthy, normal 2. Mild systemic disease 3. Severe systemic disease 4. Severe systemic disease that is a constant threat to life 5. Moribund patient not expected to survive without surgery
• Subjective assessment • Moderate inter-observer variability
NSQIP Cardiac Risk Calculator
www.qxmd.com/calculate-online/cardiology/gupta-perioperative-cardiac-risk
70-y.o. with h/o remote MI, stroke, IDDM undergoing cervical spine surgery. Needs help with some ADLs.
www.qxmd.com/calculate-online/cardiology/gupta-perioperative-cardiac-risk
Age 70
Cr < 1.5
ASA Class 3
Partially dependent
Spine surgery
70-y.o. with h/o remote MI, stroke, IDDM undergoing cervical spine surgery for progressive weakness.
www.qxmd.com/calculate-online/cardiology/gupta-perioperative-cardiac-risk
Other findings: • Excellent performance (AUC = 0.88) • MI/Cardiac arrest strongly predicts mortality (61% vs. 1%)
Caveats: • Didn’t look at all possible variables (e.g., TTE, stress test)
0.72%
Which Prediction Tool is Better?
RCRI NSQIP
Sample size ~4000 ~400,000
# of hospitals 1 > 200
Currency of data ’89 −’94 ’07 − ’08
Screen for MI? CK-MB, ECG No
Changes to Practice & Guideline?
• Suspect new ACC/AHA guideline will still use RCRI • Personal practice: use NSQIP when quantifying risk
70-y.o. with DES placed 8 months ago, IDDM and remote stroke undergoing cervical spine surgery for weakness. What about that stent?
1. Operate now, he can’t wait
2. Operate now only if he can continue antiplatelet therapy
3. Wait until 12 months after stent placement
ACC/AHA Guidelines for PCI
• Avoid PCI if patient may have upcoming surgery that requires stopping dual antiplatelet therapy
• Delay elective surgery in patients with recent PCI – Bare metal stent: 1 month – Drug eluting stent: 1 year
Surgical Outcomes After Stenting
Question: How do stent type and time until surgery affect risk of cardiac complications?
Study Design: Retrospective cohort analysis
• Over 25,000 pts who had noncardiac surgery between 6 weeks & 2 years after BMS or DES placement
• Identify risk factors for cardiac complications (all-cause mortality, MI, revascularization)
Hawn MT et al. JAMA. doi:10.1001/jama.2013.278787
Time Since Stent Placement
Time of surgery after PCI didn’t matter after first 6 months 20%
15%
10%
5%
60 120 180 240 300 360
6 months
Time between PCI & Surgery
Com
plic
atio
ns BMS
DES
Hawn MT et al. JAMA. doi:10.1001/jama.2013.278787
Surgical Outcomes After Stenting
Question: Does holding or continuing antiplatelet drugs affect risk of cardiac complications in patients with stents?
Study Design: Case-control study • 284 patients with stents who had antiplatelet drugs held
for noncardiac surgery matched with patients who had drugs continued
Results: • Holding antiplatelet drugs did not increase risk of cardiac
complications (O.R. for 0.86; 95%CI, 0.57-1.29).
Hawn MT et al. JAMA. doi:10.1001/jama.2013.278787
Guidelines for DES
Guideline Recommendation
ACC / AHA Wait 12 months before elective surgery if it requires stopping dual therapy
ACCP
• Wait 6 months before surgery (strong) • If < 6 months, continue dual therapy (weak)
ESC • 6 - 12 months of dual therapy • Continue ASA in favor of clopidogrel
70-y.o. with DES placed 8 months ago, IDDM and remote stroke undergoing cervical spine surgery for weakness. Would you start a beta-blocker?
1. Yes, I follow the guidelines
2. Maybe, I do this less often now
3. No, I’ve stopped doing this
4. No, I’ve never done this because I don’t trust the Dutch
2009 ACC / AHA Guideline for β-blockers
Definite indications (Class 1): • Already using β-blocker to treat angina, HTN, arrhythmia
Probable indications (Class 2a): • Vascular or intermediate-to-high risk surgery patients with
coronary disease, or more than 1 other risk predictor *
Uncertainty (Class 2b): • Patients undergoing vascular or intermediate risk surgery
without coronary disease but with 1 other predictor *
* CAD, CHF, DM, CKD, CVA/TIA
POISE: Treatment Protocol
1st dose Metoprolol XL 100 mg
2nd dose Metoprolol XL 100 mg
3rd & daily dose Metoprolol XL 200 mg
2-4 h OR 0-6 h 12 h
Patients: 8351 pts with s/f major noncardiac surgery • CAD, CHF, CVA/TIA, CKD, DM, or high-risk surgery • Not already taking β-blocker
Outcome: 30-day cardiac mortality, nonfatal arrest or MI
Devereaux PJ. Lancet. 2008; 371:1839-1847
POISE: Results
Metoprolol XL: Reduced cardiac events (mostly nonfatal MI)
but Increased risk of stroke & total mortality
Devereaux PJ. Lancet. 2008; 371:1839-1847
DECREASE-IV
Patients: 1066 pts with estimated 1-6% risk of postoperative cardiac complications, undergoing elective non-CV surgery
Treatment: 1. Bisoprolol 2.5 mg daily started at randomization; -- dose titrated in hospital by 1.25 - 2.5 mg daily;
-- maximum 10 mg daily; -- target heart rate = 50-70 with SBP >100 2. Fluvastatin XL 80 mg daily 3. Bisoprolol + Fluvastatin 4. Double placebo • Drugs started median 34 days prior to surgery Outcome: 30-day cardiovascular mortality or nonfatal MI
Dunkelgrun, M et al. Ann Surgery, 2009; 249: 921-926.
DECREASE-IV Results
Bisoprolol-treated patients had fewer complications Trend towards benefit with statins No safety issues
* *
Car
diac
Dea
th o
r Non
fata
l MI
Dunkelgrun, M et al. Ann Surgery, 2009; 249: 921-926.
* P = .002
Investigation of possible breaches of academic integrity
Findings regarding DECREASE IV: • Data poorly documented • Inclusion criteria violated • Outcomes not assessed per protocol
Conclusions:
• Cannot vouch for reliability of findings or validity of conclusions from this trial
β-blockers: So Now What?
Meta-analysis of secure β-blocker trials • Reduces perioperative MI (mostly asymptomatic) • Increase in mortality & strokes
Practice & Guideline Changes?
• Uncertain benefit vs. risk, even in high risk patients • Avoid fixed dose (non-titrated) perioperative β-blockade • No good reason to start β-blocker without other indication
Bouri, S et al. Heart 2013;0:1–9. doi:10.1136/heartjnl-2013-304262
Managing Perioperative Anticoagulation
Your orthopedic colleague asks your advice on how to manage anticoagulation in two patients who had hip fractures.
• One has atrial fibrillation due to HTN. • The other has a mechanical AVR. • Neither has any other relevant comorbidity 1. Heparin bridge for AVR only
2. Heparin bridge for AF only
3. Heparin bridge for both
4. Heparin bridge for neither
Thromboembolic Risks with Atrial Fibrillation
Ann
ual S
troke
Ris
k
CHADS-2 Score: 1 point for CHF, HTN, Age>75, DM 2 points for Stroke/TIA Score 0 - 2: < 5% stroke risk / yr Score 3 - 4: 5-10% Score 5 - 6: > 10%
Ansell J. Chest. 2004;126:204S-233S.
Thromboembolic Risks with Mechanical Valves
Ann
ual I
ncid
ence
Cannegieter, et al. Circulation, 1994
Effect of Mechanical Valve Location & Design on Thromboembolic Risk
Valve Location: Aortic RR = 1.0 Mitral RR = 1.8
Valve Design: Caged Ball RR = 1.0 Tilting Disk RR = 0.7 Bi-leaflet RR = 0.6
Cannegieter, et al. Circulation, 1994
Benefits & Harm of Bridging Perioperative Anticoagulation
Death or disability from thromboembolism
averted by bridging
Death or disability from perioperative bleeding
caused by bridging
Benefits & Risks
Randomized trial in progress Review of cohort studies:
Thrombosis Total Bleeding
Serious Bleeding
Bridged 1.1% 11% 3.7%
Not Bridged 0.9% 2% 0.9%
Odds Ratio (95% CI)
0.8 (0.4-1.5)
5.4 (3.0-9.7)
3.6 (1.5-8.5)
Seigal, D et al. Circulation, 2013; 126:1630
Perioperative Anticoagulation: 2012 ACCP Guidelines (9th Edition)
Atrial Fib. Mechanical Valve Recommendation
CHADS2 = 5-6; recent CVA; rheumatic AF
Any MVR; older (caged-ball or tilting disc) AVR; recent CVA
Bridge with heparin
CHADS2 = 3-4 Bileaflet AVR plus other stroke risk factor(s)
???
CHADS2 = 0-2 Bileaflet AVR without AF or other stroke risk factor
No heparin bridge
All recommendations are weak, based on low quality evidence
Are Curbside Consults Safe?
You’re about to leave the hospital when an orthopedic surgeon calls you with “a quick, curbside question” about diabetes management for a “stable” patient.
1. I never do curbside consults
2. Ask questions to determine whether curbside is appropriate
3. No problem! She’s stable.
Curbside Consults
Studied 47 requests for curbside advice to hospitalist • Curbside consultant could ask questions ad lib • Made recommendations without seeing patient or chart • Different hospitalist performed formal, in-person evaluation
Questions:
• Did curbside consultant obtain accurate information? • Did advice and management differ?
Burden, M et al. J Hosp Med, 2013; 8:31–3
Curbside vs. Formal Medicine Consult
Compared to formal consultation, how often did curbside evaluation lead to:
Incomplete clinical information 34% Inaccurate clinical information 28% Different recommendations 55% Any difference in management 60% Major difference in management 36%
Burden, M et al. J Hosp Med, 2013; 8:31–3
Thank You