Epidural analgesia for adults undergoing cardiac surgery with or without cardiopulmonary bypass

Joanne Guay; Sandra Kopp

doi:10.1002/14651858.CD006715.pub3

. 2019 Mar 1;2019(3):CD006715. doi: 10.1002/14651858.CD006715.pub3

Epidural analgesia for adults undergoing cardiac surgery with or without cardiopulmonary bypass

Joanne Guay ^1,^2,^3,^✉, Sandra Kopp ⁴

Editor: Cochrane Anaesthesia Group

PMCID: PMC6396869 PMID: 30821845

Abstract

Background

General anaesthesia combined with epidural analgesia may have a beneficial effect on clinical outcomes. However, use of epidural analgesia for cardiac surgery is controversial due to a theoretical increased risk of epidural haematoma associated with systemic heparinization. This review was published in 2013, and it was updated in 2019.

Objectives

To determine the impact of perioperative epidural analgesia in adults undergoing cardiac surgery, with or without cardiopulmonary bypass, on perioperative mortality and cardiac, pulmonary, or neurological morbidity.

Search methods

We searched CENTRAL, MEDLINE, and Embase in November 2018, and two trial registers up to February 2019, together with references and relevant conference abstracts.

Selection criteria

We included all randomized controlled trials (RCTs) including adults undergoing any type of cardiac surgery under general anaesthesia and comparing epidural analgesia versus another modality of postoperative pain treatment. The primary outcome was mortality.

Data collection and analysis

We used standard methodological procedures as expected by Cochrane.

Main results

We included 69 trials with 4860 participants: 2404 given epidural analgesia and 2456 receiving comparators (systemic analgesia, peripheral nerve block, intrapleural analgesia, or wound infiltration). The mean (or median) age of participants varied between 43.5 years and 74.6 years. Surgeries performed were coronary artery bypass grafting or valvular procedures and surgeries for congenital heart disease. We judged that no trials were at low risk of bias for all domains, and that all trials were at unclear/high risk of bias for blinding of participants and personnel taking care of study participants.

Epidural analgesia versus systemic analgesia

Trials show there may be no difference in mortality at 0 to 30 days (risk difference (RD) 0.00, 95% confidence interval (CI) −0.01 to 0.01; 38 trials with 3418 participants; low‐quality evidence), and there may be a reduction in myocardial infarction at 0 to 30 days (RD −0.01, 95% CI −0.02 to 0.00; 26 trials with 2713 participants; low‐quality evidence). Epidural analgesia may reduce the risk of 0 to 30 days respiratory depression (RD −0.03, 95% CI −0.05 to −0.01; 21 trials with 1736 participants; low‐quality evidence). There is probably little or no difference in risk of pneumonia at 0 to 30 days (RD −0.03, 95% CI −0.07 to 0.01; 10 trials with 1107 participants; moderate‐quality evidence), and epidural analgesia probably reduces the risk of atrial fibrillation or atrial flutter at 0 to 2 weeks (RD −0.06, 95% CI −0.10 to −0.01; 18 trials with 2431 participants; moderate‐quality evidence). There may be no difference in cerebrovascular accidents at 0 to 30 days (RD −0.00, 95% CI −0.01 to 0.01; 18 trials with 2232 participants; very low‐quality evidence), and none of the included trials reported any epidural haematoma events at 0 to 30 days (53 trials with 3982 participants; low‐quality evidence). Epidural analgesia probably reduces the duration of tracheal intubation by the equivalent of 2.4 hours (standardized mean difference (SMD) −0.78, 95% CI −1.01 to −0.55; 40 trials with 3353 participants; moderate‐quality evidence). Epidural analgesia reduces pain at rest and on movement up to 72 hours after surgery. At six to eight hours, researchers noted a reduction in pain, equivalent to a reduction of 1 point on a 0 to 10 pain scale (SMD −1.35, 95% CI −1.98 to −0.72; 10 trials with 502 participants; moderate‐quality evidence). Epidural analgesia may increase risk of hypotension (RD 0.21, 95% CI 0.09 to 0.33; 17 trials with 870 participants; low‐quality evidence) but may make little or no difference in the need for infusion of inotropics or vasopressors (RD 0.00, 95% CI −0.06 to 0.07; 23 trials with 1821 participants; low‐quality evidence).

Epidural analgesia versus other comparators

Fewer studies compared epidural analgesia versus peripheral nerve blocks (four studies), intrapleural analgesia (one study), and wound infiltration (one study). Investigators provided no data for pulmonary complications, atrial fibrillation or flutter, or for any of the comparisons. When reported, other outcomes for these comparisons (mortality, myocardial infarction, neurological complications, duration of tracheal intubation, pain, and haemodynamic support) were uncertain due to the small numbers of trials and participants.

Authors' conclusions

Compared with systemic analgesia, epidural analgesia may reduce the risk of myocardial infarction, respiratory depression, and atrial fibrillation/atrial flutter, as well as the duration of tracheal intubation and pain, in adults undergoing cardiac surgery. There may be little or no difference in mortality, pneumonia, and epidural haematoma, and effects on cerebrovascular accident are uncertain. Evidence is insufficient to show the effects of epidural analgesia compared with peripheral nerve blocks, intrapleural analgesia, or wound infiltration.

Plain language summary

Epidural analgesia for heart surgery with or without the heart lung machine in adults

Review question

We set out to determine from randomized controlled trials the effect of epidural pain relief on the number of deaths following surgery and risk of heart, lung, or nerve complications in adults undergoing heart surgery.

This review was first published in 2013, and it was updated in 2019.

Background

For epidural pain relief, a local anaesthetic, an opioid, or a mixture of both drugs is given through a catheter in the epidural space, which is the space immediately outside the membrane surrounding the cord. Epidural analgesia could reduce the risk of complications after surgery, such as lung infections including pneumonia, difficulty in breathing (respiratory failure), heart attack, and irregular heart rhythm caused by atrial fibrillation. A concern is that for cardiac surgery, the blood has to be thinned to reduce blood clotting, which may increase the chance of bleeding around the spinal cord. The collection of blood puts pressure on the spinal cord and can cause permanent nerve damage and disability.

Study characteristics

We included randomized controlled trials involving adults undergoing any type of cardiac surgery under general anaesthesia with or without cardiopulmonary bypass where researchers compared epidural pain relief around the time of surgery against other forms of pain relief. Surgeries performed were coronary artery bypass grafting or valvular procedures and surgeries for congenital heart disease. The average age of participants was between 43 and 75 years. Outcomes were measured up to one year after surgery.

We included 69 studies with 4860 participants. Where stated, the studies were funded by governmental resources (five studies), charity (eight), institutional resources (23), or in part by the industry (two). In all, 31 trials did not mention the source of funding. The evidence is current to November 2018.

Key results

When researchers compared epidural analgesia versus systemic pain relief (e.g. by an analgesic given directly into a vein), they could not detect any difference in the number of deaths in the first 30 days after surgery (38 studies, 3418 participants). There might be a difference in the number of people experiencing heart attacks (26 studies, 2713 participants). These findings were supported by low‐quality evidence. We found a small reduction in the risk of respiratory depression with epidural pain relief (21 studies, 1736 participants), but not in the risk of pneumonia (10 studies, 1107 participants) (low‐ or moderate‐quality evidence). The reduced risk of respiratory depression was more obvious when cardiopulmonary bypass was needed for cardiac surgery. Epidural analgesia reduced the risk of atrial fibrillation or atrial flutter early in recovery at zero to two weeks (18 studies, 2431 participants; moderate‐quality evidence). The number of cerebrovascular accidents was not clearly different (18 studies, 2232 participants), and no lasting neurological complications or epidural haematomas were reported (53 studies, 3982 participants; very low‐ or low‐quality evidence). Although epidural analgesia may have reduced the duration of tracheal intubation, this was noted mainly in older studies, and clinical practices have changed since that time (40 trials, 3353 participants; moderate‐quality evidence).

We found only six studies that compared epidural pain relief versus application of local anaesthetic on the body surface to produce peripheral nerve blocks directly into the space around the lungs (intrapleural analgesia) and onto the surgical wound (wound infiltration). These studies provided low‐ or very low‐quality evidence and did not report on many of the outcomes for this review. Study authors reported no heart attacks and no epidural haematomas.

Quality of the evidence

We rated the quality of evidence as moderate, low, or very low. We included too few participants in our review to rule out any differences in the number of patient deaths between epidural analgesia and systemic analgesia, nor to see any increase in epidural haematomas.

Summary of findings

Background

Description of the condition

The addition of thoracic epidural analgesia to general anaesthesia has been suggested to benefit patients after cardiac surgery (Svircevic 2013). However, this regional anaesthetic technique is controversial because the insertion of an epidural catheter in patients requiring full heparinization for cardiopulmonary bypass may lead to an epidural haematoma. The benefits of practicing off‐pump surgery instead of operating with the aid of cardiopulmonary bypass are not recognized by everyone, except perhaps for decreased risk of cerebrovascular accident and for high‐risk patients (Kowalewski 2016). Some clinicians argue that cardiopulmonary bypass induces a more severe inflammatory response. Also, using cardiopulmonary bypass usually requires more complete heparinization than off‐pump surgery. For this reason, we decided to evaluate all our outcomes while subgrouping the data by with or without cardiopulmonary bypass.

Description of the intervention

Epidural analgesia is a technique by which a local anaesthetic or an opioid or a mixture of both drugs is given in the epidural space (Guay 2016a; Guay 2016b; Salicath 2018). Epidural analgesia produces a superior quality of analgesia and may reduce the risk of postoperative complications such as pneumonia, respiratory failure, and myocardial infarction (Guay 2006; Guay 2014; Guay 2016a; Guay 2016b). Epidural analgesia may also shorten the duration of tracheal intubation as well as the time spent in an intensive care unit, which could have economic benefits (Guay 2016b).

How the intervention might work

High thoracic epidural analgesia may provide cardioprotective effects. High thoracic epidural analgesia increases myocardial oxygen availability, as reported in Lagunilla 2006, and reduces myocardial oxygen consumption (Hutchenson 2006). The latter is attributed to an attenuation of sympathetic response to the surgical stimuli (Kirno 1994). An influence on inflammatory response to the surgical stress and/or the cardiopulmonary bypass has also been reported (Volk 2003).

Why it is important to do this review

A possible complication of epidural analgesia includes spinal cord compression caused by a haematoma, which can result in paraplegia (Bos 2018). Systemic anticoagulation is needed for cardiac surgery and may increase the incidence of epidural haematoma related to the use of an epidural catheter (Horlocker 2018). While reviewing the literature, Landoni and colleagues found 25 cases of epidural haematoma out of 88,820 positioned epidurals in patients undergoing cardiac surgery, for an estimated risk of catheter‐related epidural haematoma of 3 per 10,000 procedures (95% confidence interval (CI) 2 to 4 per 10,000 procedures) (Landoni 2015). For the general population, the incidence of haematoma related to an epidural would be 1 per 10,000 procedures (95% CI 0 to 6 per 10,000 procedures) (Moen 2004). Although the incidence found by Landoni and colleagues may seem relatively low, the consequences of this complication may sometimes be catastrophic. In their large trial, Moen and colleagues reported 33 spinal haematomas related to neuraxial blocks. Only 6 of 33 patients made a full recovery, and 27 suffered permanent neurological damage (Moen 2004). It is therefore mandatory to have a clear view of the real benefits of epidural analgesia in cardiac surgery patients, so that patients and clinicians can make an informed decision when choosing the mode of postoperative analgesia.

This is an update of a previously published Cochrane review (Svircevic 2013).

Objectives

Methods

Criteria for considering studies for this review

Types of studies

We included randomized controlled trials (RCTs). We excluded observational studies, quasi‐randomized trials, cross‐over trials, and cluster‐randomized trials. We did not exclude studies on the basis of language of publication or publication status.

Types of participants

We included adult participants undergoing general anaesthesia for all types of cardiac surgery with or without cardiopulmonary bypass.

Types of interventions

We included trials that compared cardiac surgery including one group of participants with and one group of participants without epidural analgesia (Table 5). We excluded studies that compared cardiac surgery with participants with and participants without spinal anaesthesia. We included studies in which investigators administered epidural analgesia as a single shot block or as a continuous infusion for any duration and containing a local anaesthetic alone (extended duration or not), or in combination with an opioid (extended duration or not), or an opioid alone. We did not exclude studies in which trialists added an adjuvant other than an opioid to the solution. We excluded trials comparing nerve blocks versus systemic analgesia. For the comparator, we included all other modes of analgesia and divided them into:

1. Postoperative analgesia.

Study	Regional blockade	Comparator
Aguero‐Martinez 2012	TEA (T3‐T4) with 10 mL bupivacaine 0.5% and morphine 5 mg administered at least 1 hour before IV heparin	Low doses of opioids
Bach 2002	TEA (T12‐L1) inserted the evening before surgery Bupivacaine 0.25% 10 mL Bupivacaine 0.25% ((body height (cm) − 100) × 10^‐1 = mL/h) for 18 hours Catheter removed on the second or third day after surgery when coagulation parameters had returned to normal range	Not reported
Bakhtiary 2007	TEA (T1‐T3; soft multi‐port) inserted the day before surgery 6 mL ropivacaine 0.16% plus sufentanil 1 mcg/mL Ropivacaine 0.16% plus sufentanil 1 mcg/mL at 2 to 5 mL/h started before surgery and continued for 3 days after surgery	Metamizole and piritramide
Barrington 2005	TEA (T1‐T3) (20‐gauge; Portex, Hythe, Kent, UK) inserted 4 cm cephalad the day before surgery using a midline approach and a loss of resistance to saline technique Ropivacaine 1% 5 mL and fentanyl 50 mcg (adjusted for T1 to T6 sensory block) Ropivacaine 0.2% and fentanyl 2 mcg/mL 5 mL/h started 1 hour after induction and continued until morning of postoperative day 3 (adjusted on pain scores)	IV morphine infusion and infiltration of chest drain sites
Bektas 2015	TEA (T2‐T4) inserted 5 cm into the epidural space 1 day before surgery Lidocaine 60 mg Levobupivacaine 0.25% 0.1 mL/kg/min and fentanyl 2 mcg/kg/min bolus for T1‐L2 sensory block Levobupivacaine 0.25% 0.1 mL/kg/h and fentanyl 2 mcg/mL	IV PCA with morphine for 24 hours
Berendes 2003	TEA (C7‐T1) with a median approach and a hanging drop technique inserted the day before surgery 2 mL of 0.5% bupivacaine with epinephrine Bupivacaine 0.5% at 6 to 12 mL/h plus sufentanil 15 to 25 mcg started just before surgery and kept for 4 days	Not reported
Brix‐Christensen 1998	TEA (T3‐T4) inserted at least 12 hours before surgery Bupivacaine 0.5% 8 mL 30 minutes before induction of anaesthesia Continuous infusion with bupivacaine 2 mg/mL and fentanyl 5 mcg/mL at 5 mL/h during and after surgery until the second postoperative day	IV morphine
Caputo 2011	TEA (T2‐T4) inserted before surgery Bupivacaine 0.5% 5 + 5 mL Bupivacaine 0.125% plus clonidine 0.0003% 10 mL/hour started after induction and continued for 72 hours (adjusted for T1 to T10 sensory block and on pain scores)	IV PCA with morphine
Celik 2015	TEA (T5‐T6) inserted the day before surgery Levobupivacaine 2 mcg/mL and fentanyl 10 mcg/mL started at ICU admission at 5 mL/h and maintained for 24 hours	IV fentanyl infusion at 8 mcg/kg/h for 24 hours
Cheng‐Wei 2017	TEA PCEA with 0.075% bupivacaine and 2 mcg/mL fentanyl	Wound infusion with 0.15% bupivacaine infused continuously at 2 mL/h through a catheter embedded in the wound plus IV PCA
de Vries 2002	TEA (T3‐T4) placed immediately before induction of anaesthesia  Test dose with 3 to 4 mL of lidocaine 2% with epinephrine 1:200,000 8 to 10 mL bupivacaine 0.25% with sufentanil 25 mcg/10 mL Bupivacaine 0.125% and sufentanil 25 mcg/50 mL given at 8 to 10 mL/h	Piritramide 0.2 mg/kg intramuscularly on request
Dohle 2001	TEA (T4‐T5) 18G, midline approach, catheter advanced 3 cm past the needle tip Test dose with 3 mL 2% lidocaine Loading with 8 mL 0.5% bupivacaine injected through the catheter, followed by infusion of 0.25% bupivacaine at the rate of 6 mL/h	Paravertebral blockade, left T4 to T5, loss of resistance with saline, catheter advanced 3 cm past the needle tip Test dose with 3 mL 2% lidocaine Loading with 8 mL 0.5% bupivacaine injected through the catheter, followed by an infusion of 0.25% bupivacaine at the rate of 6 mL/h
El‐Baz 1987	TEA (T3‐T4), epidural catheter (American Pharmaseal Labs, Glendale, CA, USA) inserted by lateral approach Position of the catheter in the epidural space was confirmed by the catheter advancement test (El‐Baz 1984; "After eliciting a lack of resistance to the injection of air through the epidural needle, the ability to advance 20 cm of a soft epidural catheter, without stylet, beyond the vertebral lamina with minimal resistance was indicative of a successful epidural catheterization. After a successful advancement with minimal resistance, the epidural catheter was withdrawn 17‐18 cm leaving 2 ‐3 cm of the catheter in the epidural space and the tip near the spinal segment (T4 ‐ 5) that corresponded to the site of surgical incision. Subdural and intravascular catheterization were excluded by placing the proximal end of the epidural catheter below the site of injection for gravity drainage to assure the absence of cerebrospinal fluid or blood flow through the catheter") Morphine 0.1 mg/h started in ICU	IV morphine on request
El‐Morsy 2012	TEA (T3‐T4) inserted at least 2 hours before heparinization (change of level if blood in the needle or catheter) Test dose with 3 mL 1.5% lidocaine 0.125% bupivacaine with 1 mcg/mL fentanyl at 5 mL/h and continued until 24 hours postoperatively	IV tramadol on demand
El‐Shora 2018	TEA (T6‐T7) catheter inserted through a 17G Tuohy needle with loss of resistance technique Bupivacaine 0.125% plus fentanyl 1 mcg/mL 12 mL followed by 12 mL/h for 48 hours and started after surgery	Ultrasound‐guided bilateral paravertebral blockade at T6‐T7 Bupivacaine 0.125% plus fentanyl 1 mcg/mL 6 mL per side followed by 6 mL/h for 48 hours and started after surgery
Fawcett 1997	TEA (T2‐T4) inserted in operating room 15 mL bupivacaine 0.5% after CPB Bupivacaine 0.375% at 5 to 8 mL/h for 24 hours	IV morphine infusion for 24 hours
Fillinger 2002	TEA (T3‐T10), catheter inserted before induction of anaesthesia through an 18G Hustead needle using loss of resistance to saline technique and leaving 3 cm of catheter in the epidural space Test dose with 3 mL 1.5% lidocaine with 1:200,000 epinephrine Loading with morphine 20 mcg/kg and 0.5% bupivacaine in 5‐mg increments, to a total loading dose of 25 to 35 mg bupivacaine 0.5% bupivacaine with morphine 25 mcg/mL at 4 to 10 mL/h beginning after induction of anaesthesia (adjusted on haemodynamic parameters) Epidural catheters removed on the first postoperative day	Intravenous morphine, intravenous meperidine, and oral oxycodone
Greisen 2012	TEA (T2‐T4) inserted the day before surgery 5 to 7 mL 5.0 mg/mL bupivacaine (Marcaine, Astra, Södertälje, Sweden) together with sufentanil 2.5 mcg/mL Bupivacaine 2.5 mg/mL and sufentanil 1 mcg/mL 4 to 6 mL/h, by discretion of the attending anaesthesiologist, until end of surgery Changed to bupivacaine 1 mg/mL together with sufentanil 1 mcg/mL in ICU and continued after discharge from ICU until second postoperative day	Not reported
Gurses 2013	CEA (C6‐C7) (Braun Perifix 20 G) inserted 3 to 4 cm caudally (T2‐T4) at least 1 hour before heparin injection 0.075 mg/kg levobupivacaine hydrochloride (Chirocaine 5 mg/mL, Abbott Lab, Istanbul, Turkey) + 2 mcg/kg fentanyl (fentanyl citrate 50 mcg/mL, Abbott Lab, Istanbul, Turkey) in total 10 mL bolus 0.0375 mg/kg/h levobupivacaine + 0.5 mcg/kg/h fentanyl epidural infusion started with patient‐controlled analgesia instrument (Abbott Pain Management  Provider, Abbott Laboratoires, North Chicago, IL, USA)	Intramuscular diclofenac sodium (Dikloron 75 mg 10 amp,  Mefar Drug Ltd, Istanbul, Turkey)
Hansdottir 2006	TEA (T2‐T5) inserted the day before surgery using median hanging drop or loss of resistance technique, 3 to 5 cm into the epidural space Test dose with 4 mL lidocaine 1% PCEA with bupivacaine 0.1% and fentanyl 2 mcg/mL	IV PCA with morphine
Heijmans 2007	TEA (C7‐T1) by median approach and hanging drop technique Test dose of 2 mL lidocaine 2% Loading dose of 10 mL bupivacaine 0.25% with 2.5 mg morphine infused over 1 hour Bupivacaine 0.125% and morphine 0.2 mg/mL at 1.5 mL/h for 48 hours	IV piritramide 0.15 mg/kg
Huh 2004	TEA (T4‐T5) inserted the day before surgery Test dose with 3 mL lidocaine 2% and epinephrine 5 to 7 mL bupivacaine 0.15% and fentanyl 50 mcg before skin incision Bupivacaine 0.15% and fentanyl 10 mcg/mL through PCEA for 3 days after surgery	IV meperidine, tramadol, and NSAIDs
Hutchenson 2006	TEA (T2‐T4) inserted 3 cm the day before surgery with fluoroscopic guidance Bupivacaine 0.5% 200 mcg/cm body height Bupivacaine 0.25% 200 mcg/cm body height per hour	Not reported
Jakobsen 2012	TEA (T3‐T4) Test dose of 3 mL 2% lidocaine Bolus dose of 5 to 7 mL, guided by primary patient heights, of 0.5% bupivacaine (Marcaine; Astra, Södertälje, Sweden) and sufentanil 2.5 mcg/mL Bupivacaine 2.5 mg/mL/sufentanil 1 mcg/mL, 4 to 6 mL/h during surgery Bupivacaine 1 mg/mL and sufentanil 1 mcg/mL postoperatively and continued after discharge from ICU until second postoperative day	Participants in both groups received intravenous morphine or alfentanil according to the department’s general guidelines (i.e. morphine 0.05 mg/kg, or alfentanil 25 mcg, if rapid pain relief was needed) All participants in both groups received additional oral or intravenous paracetamol 1 g every 6 hours
Kendall 2004	TEA (T1‐T4) inserted after induction through a paramedian approach and loss of resistance technique 2 mL 0.5% bupivacaine plus epinephrine 0.1 mL/kg 0.1% bupivacaine plus fentanyl 5 mcg/mL followed by infusion at 0.1 mL/kg/h kept for 48 hours	IV PCA with morphine
Kilickan 2006	TEA (T1‐T5) inserted the day before surgery (3 attempts only) Test dose with 3 to 4 mL 2% lidocaine, position confirmed with injection of contrast material and X‐ray Bupivacaine 20 mg after anaesthesia induction Bupivacaine 0.125% 4 to 10 mL/h intraoperatively and postoperatively for 3 days, adjusted for a sensory blockade from T1 to T10	IV PCA with morphine
Kilickan 2008	TEA (T1‐T5) inserted the day before surgery (3 attempts only) Test dose with 3 to 4 mL 2% lidocaine, position confirmed with injection of contrast material and X‐ray Bupivacaine 20 mg 60 minutes before induction of anaesthesia Bupivacaine 20 mg/h intraoperatively and postoperatively for 3 days	IV PCA with Dolantin
Kirno 1994	TEA (T3‐T4; Perifix, B. Braun, Melsungen AG, Germany) at least 12 hours before surgery Mepivacaine 20 mg/mL (Carbocain, Astra, Södertälje, Sweden) was injected to achieve a T1‐T5 block	Not reported
Kirov 2011	TEA (T2‐T4) Test dose of 1 mL 2% lidocaine Ropivacaine 0.75% 1 mg/kg and fentanyl 1 mcg/kg for surgery Ropivacaine 0.2% and fentanyl 2 mcg/mL at 3 to 8 mL/h (VAS score < 30 mm at rest) or via PCEA after surgery	IV fentanyl 10 mcg/mL at 3 to 8 mL/h
Konishi 1995	TEA (T7‐T10) inserted the day before surgery Butorphanol 0.5 to 1.0 mg or Morphine 2.5 mg	Fentanyl, pentazocine, and minor tranquillizers
Kundu 2007	TEA (C7‐T2) inserted 3 to 4 cm cephaladly before anaesthesia induction with hanging drop technique in left lateral decubitus position Lidocaine 1% 5 mL Bupivacaine 0.25% 5 mL plus fentanyl 10 mcg Bupivacaine 0.25% 5 mL plus fentanyl 10 mcg every 2 hours	Not reported
Kunstyr 2001	TEA (T1‐T5) inserted at least 60 minutes before heparinization 10 mL bupivacaine 0.5% Bupivacaine 0.125% plus sufentanil 1 mcg/mL infused at 3 to 8 mL/h after surgery	Postoperative analgesia with a mixture of ketamine 400 mg and sufentanil 100 mcg in 50 mL syringe, administered in a continuous infusion; rate of infusion 0.5 to 3.5 mL/h Nurse administered morphine on request IV PCA with morphine
Lenkutis 2009	TEA (T1‐T2) Lidocaine 2% 7 to 8 mL Bupivacaine 0.25% at 8 mL/h during surgery Bupivacaine 0.25% and fentanyl 5 mcg/mL at 5 to 7 mL/h for at least 84 hours postoperatively	IM/IV pethidine 0.1 to 0.4 mg/kg
Liem 1992	TEA (T1‐T2) inserted the day before surgery by paramedian approach and hanging drop technique  Test dose with 2 mL 2% lidocaine Loading with 0.375% bupivacaine plus sufentanil 5 mcg/mL at a dose of 0.05 mL/cm body length administered over a 10‐minute period 0.125% bupivacaine plus sufentanil 1 mcg/mL at 0.05 mL/cm body length/h started before induction and continued for 72 hours	IV nicomorphine
Loick 1999	TEA (C7‐T1) inserted the day before surgery by median approach and hanging drop technique Test dose with 2 mL bupivacaine 0.5% with adrenaline Loading before induction with 8 to 12 mL bupivacaine 0.375% and 16 to 24 mcg sufentanil into the epidural space in increments to block the somatosensory level C7‐T6 PCEA with bupivacaine 0.75% plus sufentanil 1 mcg/mL if < 65 years of age, and without adjuvant if ≥ 65 years (duration unclear, possibly 48 hours)	PCA with piritramide
Lundstrom 2005	TEA (T1‐T3) inserted the day before surgery by median approach using hanging drop technique Test dose with 2 mL 2% lidocaine Loading with 8 to 10 mL bupivacaine 0.5% (adjusted for sensory block T1‐T8) before induction Bupivacaine 0.125% and morphine 25 mcg/mL at 5 mL/h plus 4 mL every hour started after induction Bupivacaine 0.25% 4 mL on request after surgery (adjusted for T1‐T8) Catheters removed on day 4 or 5	Morphine IV for 24 hours, then orally
Lyons 1998	TEA (C7‐T1) Bupivacaine 0.5% 0.1 mL/kg Bupivacaine 0.1% and fentanyl 2 mcg/mL, infusion for 72 hours	Not reported
Mehta 1998	TEA (T4‐T5 or T5‐T6) 16G, median approach, loss of resistance to saline, catheter inserted 3 to 4 cm past the needle tip On first demand for pain relief, participants in the TEA group received 8 mL 0.25% bupivacaine hydrochloride Maximum of 3 doses was given over the next 12 hours, if required	Intrapleural catheter: 16G epidural catheter inserted in intercostal space 6 to 7 cm in left anterior axillary line by the operating surgeon, 6 to 8 cm in intrapleural space, directed posteriorly and anchored with a skin suture before thoracotomy closure On first demand for pain relief, participants in the intrapleural group received 20 mL 0.25% bupivacaine hydrochloride Before injection of intrapleural bupivacaine, participants were positioned supine with a one‐third left lateral tilt and with the intercostal chest tube clamped after exclusion of any air leak. The chest tube was kept clamped for 20 minutes after the injection Maximum of 3 doses was given over the next 12 hours, if required
Mehta 2008	TEA (C7‐T1) hanging drop technique in the sitting position, catheter inserted 4 cm beyond needle tip Lidocaine 2% 3 mL Bupivacaine 0.5% 8 mL Bupivacaine 0.25% at 0.1 mL/kg/h	Paravertebral blockade Loss of resistance to saline at left T4‐T5 Lidocaine 2% 3 mL Bupivacaine 0.5% 8 mL Bupivacaine 0.25% at 0.1 mL/kg/h
Mehta 2010	TEA (C7‐T1) using hanging drop technique in sitting position inserted at least 2 hours before heparinization; intervention postponed in cases of bloody tap 3 mL 2% lidocaine without epinephrine; adequacy and level of the block established by confirming loss of pin‐prick sensation and warm/cold discrimination 8 to 10 mL 0.25% bupivacaine (aim at T4 sensory block) Bupivacaine infusion (0.125%) with fentanyl citrate (1 mcg/mL) at the rate of 5 mL/h was commenced and continued until postoperative day 3 to provide intraoperative and postoperative analgesia	Not reported
Mishra 2004	No details available	Not reported
Moore 1995	TEA (T1‐T5) Bupivacaine 0.5% in 2 mL increments for sensory block from T1 to L2 Bupivacaine 0.375% at 5 to 8 mL/h started before induction Bupivacaine 0.25% at 5 to 8 mL/h for at least 24 hours	IV papaveretum
Nagaraja 2018	TEA (C7‐T1) inserted (3 to 4 cm caudally) the day before surgery through an 18G Tuohy needle 0.25% plain bupivacaine 15 mL before surgery followed by 0.125% plain bupivacaine at 0.1 mL/kg/h for 48 hours post extubation	Ultrasound‑guided (in‐plane) erector spinae plane lock Catherer inserted 5 cm cephaladly the day before surgery through an 18G Tuohy needle. 3 cm lateral to T6 spinous process (T5 transverse process) with hydrodissection below the erector spinae muscle with 5 mL normal saline, 0.25% plain bupivacaine, 15 mL in each catheter before surgery, followed by 0.125% plain bupivacaine at 0.1 mL/kg/h for 48 hours post extubation, through each catheter
Neskovic 2013	TEA (T2‐T4) inserted 30 minutes before surgery and at least 2 hours before the first dose of heparin Test dose 10 to 15 mL 0.125 or 0.25% bupivacaine with fentanyl 0.125 or 0.25% bupivacaine with fentanyl at 5 to 10 mL/h	Not reported
Nygard 2004	TEA (T1‐T3) inserted the day before surgery by the median approach and hanging drop technique Test dose with 2 mL 2% lidocaine Loading with 8 to 10 mL bupivacaine 05% before induction (adjusted for T1 to T8) Bupivacaine 0.125% with morphine 25 mcg/mL at 5 mL/h started after induction and continued for 4 days  Additional bolus doses of 4 mL bupivacaine 0.5% hourly during the operation	Morphine IV for 24 hours, then orally
Obersztyn 2018	TEA (T1‐T3) with hanging drop technique, catheters inserted 3 to 4 cm into the epidural space at least 6 hours before surgery Before surgery: 9 to 11 mL 0.25% bupivacaine with fentanyl in a concentration of 10 mcg/mL, followed by 0.19% (more exactly, 0.1875%) bupivacaine and fentanyl at 6 mL/h during surgery and 0.125% bupivacaine plus fentanyl 6.25 mcg/mL at 2 to 8 mL/h after surgery until discharge fro the ICU (mean 18.8 hours)	IV morphine
Onan 2011	TEA (T2‐T4; side‐holed 18 G epidural catheter) by using a median approach and a loss of resistance technique with saline solution Test dose with 3 to 4 mL 2% lidocaine 20 mg bolus 0.25% bupivacaine through the epidural catheters 1 hour before surgery 0.25% bupivacaine infused at a rate of 20 mg/h during surgery 0.125% bupivacaine at 4 to 10 mL/h after surgery (adjusted for T1‐T10) Epidural catheters removed at 24 hours postoperatively	Not reported
Onan 2013	TEA (T1‐T5) inserted the night before surgery 3 cm into epidural space Test dose with 3 to 4 mL 2% lidocaine Sensory blockade tested with ice plus X‐ray after injection of contrast material Bolus of 20 mg 0.25% bupivacaine 1 hour before surgery 20 mg/h 0.25% bupivacaine intraoperatively Bupivacaine 0.25% 10 to 20 mL/h during first 24 hours after surgery (adjusted according to pain scores)	Acetaminophen (500 mg) and tramadol  (1 mg/kg) used as rescue medications
Palomero 2008	TEA (T3‐T6) inserted the day before surgery Bolus of 6 to 8 mL 0.33% bupivacaine 0.175% bupivacaine 6 to 8 mL/h for 48 hours Catheter withdrawn after check of coagulation status	Morphine 0.5 to 1 mL/h
Petrovski 2006	TEA; no details	Not reported
Priestley 2002	TEA (T1‐T4; 18G side‐holed epidural catheter) inserted the evening before surgery Test dose with 2% lidocaine 3 to 4 mL Loading with 4 mL ropivacaine 1% and fentanyl 100 mcg (adjusted for T1‐T6) Ropivacaine 1% and fentanyl 5 mcg/mL at 3 to 5 mL/h started before induction and continued for 48 hours	Continuous morphine infusion for 24 hours, followed by PCA with morphine
Rein 1989	TEA (T4‐T5) Bupivacaine 0.5% 10 mL at induction of anaesthesia and 4 mL every hour during surgery Bupivacaine 0.5% at 4 mL/h for 24 hours	Morphine
Royse 2003	TEA (T1‐T3) inserted the night before the operation 8 mL bupivacaine 0.5% with fentanyl 20 mcg before induction Ropivacaine 0.2% with fentanyl 2 mcg/mL at 5 to 14 mL/h (for T1‐T10 sensory block) and continued until postoperative day 3, 6H00 AM	PCA with morphine
Scott 2001	TEA (T2‐T4) inserted before induction Loading with bupivacaine 0.5% 2 boluses of 5 mL (for T1‐T10) Bupivacaine 0.125% and 0.0006% clonidine at 10 L/h started after induction and continued for 96 hours (adjusted on pain scores and sensory block)	Target controlled infusion of alfentanil for 24 hours  followed by PCA with morphine for another 48 hours (adjusted on pain scores)
Sen 2017	TEA (T2‐T4) inserted 4 to 6 cm into epidural space the day before surgery Lidocaine 2% with epinephrine 5 mcg/mL 3 mL Bupivacaine 0.1% and fentanyl 2 mcg/mL at 0.1 mL/kg/h started after induction	IV fentanyl 0.5 to 2 mcg/kg/h IV tramadol 100 mg as rescue analgesia
Sharma 2010	TEA (C7‐T2) inserted at least 2 hours before heparinization and using hanging drop technique via midline approach Test dose 3 mL 2% lignocaine without epinephrine Loading with 8 to 10 mL bupivacaine 0.25% (for sensory block until at least T4) before induction Bupivacaine 0.125% with 1 mcg/mL fentanyl citrate at 5 mL/h started after induction and continued until third postoperative day	IV continuous infusion of tramadol
Stenseth 1994	TEA (T4‐T6) inserted the day before surgery Test dose with lidocaine 10 mL bupivacaine 0.5% before induction 4 mL bupivacaine 0.5% hourly during surgery Bupivacaine 0.5% at 3 mL/h plus 4 mL every 4 hours after surgery	IV morphine on request
Stenseth 1996	TEA (T4‐T6) inserted the day before surgery Test dose with lidocaine 10 mL bupivacaine 0.5% before induction (for at least T1‐T2 block) 4 mL bupivacaine 0.5% hourly during surgery Bupivacaine 0.5% at 3 mL/h plus 4 mL every 4 hours after surgery Morphine epidurally 4 to 6 mg 3 to 4 times a day for the next 2 days, supplemented with bupivacaine 5 mg/mL when needed until third postoperative day	IV morphine on request
Stritesky 2006	TEA (T2‐T4) 1 hour before surgery with an 18G Tuohy needle and hanging drop or loss of resistance technique, with catheter inserted 4 cm past the needle tip 10 mL bupivacaine 0.25% plus fentanyl 100 mcg for loading (half through the needle and half through the catheter) Bupivacaine 0.25% and fentanyl 1 mcg/mL at 8 to 12 mL/h during surgery and for 48 hours	Not reported
Svircevic 2011	TEA (T2‐T4) at least 4 hours before heparinization Test dose with lidocaine 2% 3 mL 0.1 mL/kg administered of a solution of 0.08 mg/mL morphine and 0.125% bupivacaine, followed by continuous infusion of 4 to 8 mL/h of the same solution started before induction Epidural catheter removed before transfer to the general ward (median 22 hours)	Morphine IV infusion
Tenenbein 2008	TEA (T2‐T5) inserted at least 4 hours before systemic heparinization 2.5 mL test dose of 2% lidocaine, with 1:200,000 epinephrine on insertion 3 mL test dose of 2% lidocaine before surgery 0.75% ropivacaine 5 mL with hydromorphone 200 mcg followed by an infusion of ropivacaine 0.75% at 5 mL/h during surgery 0.2% ropivacaine with hydromorphone 15 mcg/mL for 48 hours after surgery	IV PCA with morphine Indomethacin suppositories (100 mg) postoperatively, and twice‐daily naproxen (500 mg)
Tenling 1999	TEA (T3‐T5; 16G), inserted the day before surgery through the lateral approach and loss of resistance technique with saline 0.9% Test dose of 2 to 3 mL lidocaine 1% 8 to 12 mL bupivacaine 0.5% the morning of the operation (for T1‐T8 sensory block) Bupivacaine 0.5% at 4 to 8 mL/h until ICU admission Bupivacaine 0.2% and sufentanil 1 mcg/mL at 3 to 7 mL/h from arrival to ICU until the day after the operation	IV ketobemidone
Usui 1990	TEA (T6‐T7) inserted 4 cm past needle tip 24 hours before surgery and kept for 1 or 2 days after extubation Morphine 3 mg given after surgery and repeated as required	Morphine 10 mg IV as required Additional co‐analgesia as required
Volk 2003	TEA (C7‐T3) inserted the day before surgery Lidocaine 2% for T1‐T6 sensory block Bupivacaine 0.5% 6 to 10 mL hourly during surgery Bupivacaine 0.25% at 6 to 12 mL/h for at least 24 hours	IV patient‐controlled analgesia with piritramide
Yang 1996	TEA (T4‐T5) inserted 3 cm cephalad in the right lateral decubitus position Lidocaine 2% 3 mL Bupivacaine 0.375% and fentanyl 5 mcg/mL 0.06 mL/cm of body length Bupivacaine 0.25% with fentanyl 5 mcg/mL 0.03 mL/cm of body length every hour	Not reported
Yilmaz 2007	TEA (T3‐T6) inserted cranially 3 to 4 cm 16 to 24 hours before systemic heparinization (Perifix 18G, Braun) Loading with morphine 5 mcg/kg and 6 mL bupivacaine 0.25% at least 45 minutes before surgical incision 6 mL bupivacaine 0.12% with fentanyl 2.5 mcg/kg every 6 hours for 48 hours, after which catheters were withdrawn	IV fentanyl 0.7 mcg/kg/h
Yung 1997	TEA or upper lumbar epidural inserted 24 hours before surgery Lidocaine 1.5% 25 to 30 mL with ketamine 15 mg, morphine 1 mg/10 kg for surgery Morhine 1 mg in 10 mL normal saline every 12 hours for 5 days for postoperative analgesia	IV meperidine HCl
Zawar 2015	TEA (C7‐T2) catheters inserted 4 to 5 cm cranially using hanging drop technique If a “bloody tap” was to occur, the operation was postponed for 24 hours and participant was excluded from the study Bolus of 6 to 14 mL ropivacaine 0.75% for T1‐T10 sensory block (sensory loss to cold pack and needle prick) Infusion of 5 to 15 mL/h ropivacaine 0.2% for 72 hours after surgery	IV tramadol hydrochloride 100 mg 8 hourly
Zhou 2010	TEA (T4‐T6) inserted in lateral decubitus position the day before surgery Bolus 8 to 20 mL lidocaine 1% PCEA with ropivacaine 0.125% and fentanyl 2 mcg/mL at 4 mL/h plus 2 mL bolus (lockout time 20 minutes)	IV PCA with fentanyl

Study	Criteria
Aguero‐Martinez 2012	New pathological Q wave (duration ≥ 0.04 second and depth ≥ 25% of the R wave or QRS complex) in more than 1 derivation. Non‐specific changes that included elevation of the ST segment > 1.5 mm from the isoelectric line in 2 or more leads of the same region, ST depression > 2 mm in the precordial leads, or reversal of the T wave for longer than 48 hours; absence of R wave in the precordial leads. Ventricular or atrioventricular conduction defects Enzymatic criteria: 5 times normal values: troponin > 1 mcg/mL, CK > 250 U/L, CK‐MB > 133 U/L, LDH > 800 U/L, LDH 1/LDH 2 > 1 in blood samples collected between postoperative days 2 and 3, and GOT > 90 U/L Echocardiographic criteria: new segmental motility disorders Anatomopathological criteria: in dead patients
Bakhtiary 2007	Unspecified
Barrington 2005	Transmural infarction defined as new Q waves
Bektas 2015	1. Cardiac biomarkers (with troponins preferred) rise > 10 times 99% upper reference limit (URL) from normal preoperative level 2. New pathological Q waves or new left bundle branch block (LBBB) and/or imaging or angiographic evidence of new occlusion of native vessels or grafts, new regional wall motion abnormality, or loss of viable myocardium
Caputo 2011	New Q waves of 0.04 ms and/or reduction in R waves > 25% in at least 2 contiguous leads on ECG
Celik 2015	ECG monitored (ST analysis); CK‐MB and troponin I levels measured at fourth and 24th hours
de Vries 2002	Myocardial infarction defined as a new Q wave on ECG and CK 180 U/L with CK‐MB 10% of total
Dohle 2001	Myocardial infarction assessed by ECG changes and CK‐MB values
Fillinger 2002	New Q waves of at least 0.04 second duration or postoperative elevation of serum creatine phosphokinase confirmed by creatine phosphokinase isoenzyme pattern
Hansdottir 2006	New Q waves or CK‐MB isoenzyme concentration ≥ 50
Heijmans 2007	Myocardial infarction not mentioned in the report
Jakobsen 2012	Perioperative myocardial infarction, defined as new Q waves of 0.04 ms and/or reduction in R waves > 25% in at least 2 contiguous leads on ECG
Kendall 2004	ECG changes (new Q wave, or loss of R wave progression, or new permanent left bundle branch block) and increase in creatinine kinase myocardial fraction (CK‐MB) to > 120 units per litre
Kilickan 2006	Unspecified
Liem 1992	CK‐MB values ≥ 80 IU/L and evidence of new Q waves or bundle branch block on postoperative ECG
Loick 1999	Unclear
Lyons 1998	Unclear
Mehta 1998	Incidence of perioperative myocardial infarction also analysed by an independent cardiologist, as per the appearance of new Q waves in the ECG and increase in creatine phosphokinase‐myocardial band isoenzyme (CPK‐MB) levels to > 70 ng/mL in the first 12 hours postoperatively
Mehta 2010	2‐lead ECG and CPK, CPK‐MB levels
Neskovic 2013	New ECG changes with positive enzymes (CK‐MB and troponin)
Obersztyn 2018	ECG and elevated serum enzymes
Onan 2011	Unspecified
Onan 2013	Unspecified
Palomero 2008	Myocardial infarction defined by analysis of the ECG (new Q waves or increases in ST segment > 3 mm)
Priestley 2002	New Q waves (assessed by the blinded cardiologist) on a 12‐lead ECG on days 0, 1, 2, and 4 and assessment of venous blood levels of troponin  T and creatine kinase‐MB fraction on arrival in the ICU, and again at 4, 12, and 24 hours and on postoperative day 2
Scott 2001	Q waves, ST segment increase of 3 mm, and a myocardial specific serum creatinine kinase level ≥ 60 ng/mL
Stenseth 1994	Unspecified
Stenseth 1996	Unspecified
Svircevic 2011	Creatine kinase muscle–brain isoenzymes > 75 units per litre (5 times upper limit of normal level) and peak creatine kinase muscle–brain isoenzyme/creatine kinase ratio > 10% or new Q wave infarction
Zawar 2015	Myocardial infarction defined as developing ECG changes, new Q waves on postoperative ECG ≥ 0.03 second in duration in 2 or more adjacent leads lasting until discharge, rise in creatine phosphokinase‑MB and troponin I, and new regional wall motion abnormalities

Study	Neurological complication
Aguero‐Martinez 2012	Neurological complication: any new‐onset psychiatric or neurological disorder with altered consciousness with or without focalization
Barrington 2005	Stroke
Bektas 2015	Stroke: All participants were postoperatively managed in the cardiac surgery intensive care unit. Postoperative stroke was suspected when a patient showed focal neurological deficits or delayed recovery of mental status after surgery. Such patients were referred to stroke neurologists and were evaluated by computed tomography. Post coronary artery bypass grafting, stroke was diagnosed as: 1) newly developed neurological deficits within 14 days of coronary artery bypass grafting; and 2) Low‐density lesions on postoperative computed tomography that were not observed preoperatively. Strokes that occurred within 24 hours after coronary artery bypass grafting were defined as immediate, whereas all others were considered delayed
Caputo 2011	Stroke/transient ischaemic attack: diagnosis of stroke was made if evidence showed new neurological deficit with morphological substrate confirmed by computed tomography or nuclear magnetic resonance imaging
Celik 2015	Stroke: neurological findings of participants (hemiparesis, hemiplegia, etc.) were followed
Fillinger 2002	Neurological event: new sensorimotor neurological events
El‐Shora 2018	Stroke
Hansdottir 2006	Stroke: defined as a new central neurological deficit
Heijmans 2007	Stroke
Jakobsen 2012	Transitory Ischaemic attack lasting less than 24 hours
Neskovic 2013	Stroke: new motor or sensory deficit after surgery
Onan 2013	Cerebrovascular accident
Palomero 2008	Focal neurological dysfunction defined as a sensory or motor deficit affecting 1 or more limbs appearing 5 days after surgery
Royse 2003	Stroke
Scott 2001	Cerebrovascular accident defined as a new motor or sensory deficit affecting 1 or more limbs and present on awakening from anaesthesia or occurring within the next 5 days
Stenseth 1996	Hemiparesis
Svircevic 2011	Stroke: a new motor or sensory deficit of central origin, persisting longer than 24 hours, preferably confirmed by computed tomography, resulting in a drop of 2 points on the Rankin scale
Tenenbein 2008	Stroke or transient ischaemic attack
Zawar 2015	Stroke was documented if diagnosed on computed tomography scan or magnetic resonance imaging

Epidural analgesia compared with systemic analgesia for cardiac surgery with or without cardiopulmonary bypass in adults
Patient or population: adults undergoing cardiac surgery with or without cardiopulmonary bypass  Settings: trials were conducted in university hospitals (n = 60) or at a tertiary care centre (n = 3). Trials were conducted in Australia (n = 3); Bangladesh (n = 1); Canada (n = 1); China (n = 2); Cuba (n = 1); Czech Republic (n = 2); Denmark (n = 5); Egypt (n = 1); Germany (n = 5); India (n = 6); Italy and UK (n = 1); Japan (n = 2); Korea (n = 1); Lithuania (n = 1); Macedonia (n = 1); Norway (n = 3); Poland (n = 1); Russia (n = 1); Serbia (n = 1); Spain (n = 1); Sweden (n = 3); Taiwan (n = 1); Turkey (n = 8); The Netherlands (n = 4); UK (n = 5); and USA (n = 3)  Intervention: epidural analgesia  Comparison: systemic analgesia
Outcomes	Illustrative comparative risks (95% CI)*		Risk difference or relative effect  (95% CI)	No. of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Assumed risk	Corresponding risk
	Systemic analgesia	Epidural analgesia
Mortality  (0 to 30 days)	Study population		RD 0.00   (‐0.01 to 0.01)	3418  (38 studies)	⊕⊕⊝⊝  low^a
	6 per 1000	7 per 1000  (4 to 13)
Myocardial infarction (0 to 30 days)	Study population		RD ‐0.01   (‐0.02 to 0.00)	2713  (26 studies)	⊕⊕⊝⊝  low^a
	40 per 1000	28 per 1000 (21 to 39)
Pulmonary complications (0 to 30 days)	Respiratory depression		RD ‐0.03   (‐0.05 to ‐0.01)	1736  (21 studies)	⊕⊕⊝⊝  low^b	NNTB 32 (95% CI 22 to 102)
	Study population
	70 per 1000	42 per 1000 (30 to 57)
	Pneumonia		RD ‐0.03   (‐0.07 to 0.01)	1107  (10 studies)	⊕⊕⊕⊝  moderate^c
	Study population
	148 per 1000	79 per 1000 (59 to 105)
Atrial fibrillation or atrial flutter (0 to 2 weeks)	Study population		RD ‐0.06   (‐0.10 to ‐0.01)	2431  (18 studies)	⊕⊕⊕⊝  moderate^c	NNTB 14 (95% CI 8 to 90)
	327 per 1000	258 per 1000 (234 to 283)
Risk of neurological complications (0 to 30 days)	Cerebrovascular accident		RD ‐0.00   (‐0.01 to 0.01)	2232  (18 studies)	⊕⊝⊝⊝  very low^d
	Study population
	12 per 1000	11 per 1000 (6 to 18)
	Epidural haematoma		RD 0.00 (‐0.01 to 0.01)	3982 (53 studies)	⊕⊝⊝⊝  low^a
	Study population
	0 per 1000	0 per 1000 (0 to 2)
Duration of tracheal intubation	Mean duration of tracheal intubation was 0.78 SMD lower (‐1.01 to ‐0.55)			3353  (40 studies)	⊕⊕⊕⊝  moderate^c	The difference was equivalent to 2.4 hours^e and was more evident in older trials (see text)
Pain at rest at 6 to 8 hours after surgery	Mean pain scores were 1.35 SMD lower (‐1.98 to ‐0.72)			502 (10 studies)	⊕⊕⊕⊝  moderate^f	The difference was equivalent to 1 on a score from 0 to 10^e
Haemodynamic support (in hospital)	Hypotension or need for vasopressor boluses		RD 0.21 (0.09 to 0.33)	870 (17 studies)	⊕⊝⊝⊝  low^g	The number needed to harm is 4 (95% CI 3 to 12)
	Study population
	451 per 1000	284 per 1000 (243 to 330)
	Inotropic or vasopressor infusions		RD 0.00 (‐0.06 to 0.07)	1821 (23 studies)	⊕⊝⊝⊝  low^h
	Study population
	344 per 1000	338 per 1000 (302 to 376)
The corresponding risk* (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Confidence intervals were calculated using VassarStats (http://www.vassarstats.net/).  CI: confidence interval; NNTB: number needed to treat for an additional beneficial outcome; RD: risk difference; SMD: standardized mean difference.
GRADE Working Group grades of evidence.  High quality: further research is very unlikely to change our confidence in the estimate of effect.  Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low quality: we are very uncertain about the estimate.

Epidural analgesia compared with peripheral nerve blocks for cardiac surgery without cardiopulmonary bypass in adults
Patient or population: adults undergoing cardiac surgery without cardiopulmonary bypass Settings: trials were conducted in university hospitals in Egypt (n = 1) or India (n = 3) Intervention: epidural analgesia Comparison: peripheral nerve blocks (erector spinae plane block (n = 1) or paravertebral blockade (n = 3))
Outcomes	Illustrative comparative risks (95% CI)*		Risk difference or relative effect  (95% CI)	No. of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Assumed risk	Corresponding risk
	Peripheral nerve block	Epidural analgesia
Mortallity  (0 to 30 days	Study population		RD ‐0.03 (‐0.08 to 0.02)	145 (1 study)	⊕⊝⊝⊝ very low^a
	43 per 1000	13 per 1000 (2 to 72)
Myocardial infarction (0 to 30 days)	Study population		RD 0.00 (‐0.07 to 0.07)	76 (2 studies)	⊕⊝⊝⊝ very low^a
	0 per 1000	0 per 1000 (0 to 90)
Pulmonary complications (0 to 30 days)	We found no data for this outcome (respiratory depression or pneumonia)
Atrial fibrillation or atrial flutter (0 to 2 weeks)	We found no data for this outcome
Risk of neurological complications (0 to 30 days)	Cerebrovascular accident
	Study population		RD 0.00 (‐0.03 to 0.03)	145 (1 study)	⊕⊝⊝⊝ very low^a
	0 per 1000	0 per 1000 (0 to 49)
	Epidural haematoma
	Study population		RD 0.00 (‐0.03 to 0.03)	271 (4 studies)	⊕⊕⊝⊝ low^b
	0 per 1000	0 per 1000 (0 to 27)
Duration of tracheal intubation	Study population		MD ‐0.08 hour (‐0.54 to 0.38 hour)	271 (4 studies)	⊕⊝⊝⊝ very low^a
	6.82 ± 2.14 hours (mean ± SD)	6.67 ± 2.31 hours (mean ± SD)
Pain at rest at 6 to 8  hours after surgery (score from 0 to 10)	Study population		MD 0.12 (‐0.42 to 0.66)	90 (2 studies)	⊕⊝⊝⊝ very low^a
	2.20 ± 0.79 (mean ± SD)	1.80 ± 0.22 (mean ± SD)
Haemodynamic support (in hospital)	Hypotension or need for vasopressor boluses		RD 0.05 (‐0.08 to 0.18)	40 (1 study)	⊕⊕⊝⊝ low^b
	Study population
	50 per 1000	0 per 1000 (0 to 161)
	Inotropic or vasopressor infusions
	We found no data for this outcome
The corresponding risk* (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Confidence intervals were calculated using VassarStats (http://www.vassarstats.net/) with no continuity correction.  CI: confidence interval; MD: mean difference; RD: risk difference; SD: standard deviation.
GRADE Working Group grades of evidence.  High quality: further research is very unlikely to change our confidence in the estimate of effect.  Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low quality: we are very uncertain about the estimate.

Epidural analgesia compared with intrapleural analgesia for cardiac surgery in adults
Patient or population: adults undergoing cardiac surgery without cardiopulmonary bypass Settings: university hospital in India Intervention: epidural analgesia Comparison: intrapleural analgesia
Outcomes	*Illustrative comparative risks (95% CI)**		Risk difference or relative effect  (95% CI)	No. of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Assumed risk	Corresponding risk
	Intrapleural analgesia	Epidural analgesia
Mortality (0 to 30 days)	We found no data for this outcome
Myocardial infarction (0 to 30 days)	Study population		RD 0.00 (‐0.07 to 0.07)	50 (1 study)	⊕⊕⊝⊝ low^a
	0 per 1000	0 per 1000 (0 to 71)
Pulmonary complications (0 to 30 days)	We found no data for this outcome (respiratory depression or pneumonia)
Atrial fibrillation or atrial flutter)  (0 to 2 weeks)	We found no data for this outcome
Risk of neurological  complications (0 to 30 days)	Cerebrovascular accident
	We found no data for this outcome
	Epidural haematoma
	Study population		RD 0.00 (‐0.07 to 0.07)	50 (1 study)	⊕⊕⊝⊝ low^a
	0 per 1000	0 per 1000
Duration of tracheal intubation	Study population		MD ‐0.30 (‐1.20 to 0.60 hour)	15 (1 study)	⊕⊕⊝⊝ very low^b	17 participants in the epidural analgesia group and 14 in the intrapleural analgesia group were extubated in the operating room Means and SDs given by study authors are those for the rest of the participants
	4.1 ± 0.59 hours (mean ± SD)	3.8 ± 1.13 hours (mean ± SD)
Pain at rest at 6 to 8 hours (score from 0 to 10)	Study population		MD 0.84 (0.31 to 1.37)	50 (1 study)	⊕⊕⊝⊝ low^a
	4.52 ± 1.08	3.68 ± 0.82
Haemodynamic support (in hospital)	We found no data for this outcome
The corresponding risk* (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Confidence intervals were calculated using VassarStats (http://www.vassarstats.net/).  CI: confidence interval; MD: mean difference; RD: risk difference; SD: standard deviation.
GRADE Working Group grades of evidence.  High quality: further research is very unlikely to change our confidence in the estimate of effect.  Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low quality: we are very uncertain about the estimate.

Epidural analgesia compared with wound infiltration for adults undergoing cardiac surgery without cardiopulmonary bypass
Patient or population: adults undergoing cardiac surgery without cardiopulmonary bypass Settings: university hospital in Taiwan Intervention: epidural analgesia Comparison: wound infiltration
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Assumed risk	Corresponding risk
	Wound infiltration	Epidural analgesia
Mortallity  (0 to 30 days)	We found no data for this outcome
Myocardial infarction  (0 to 30 days)	We found no data for this outcome
Pulmonary complications (0 to 30 days)	We found no data for this outcome (respiratory depression or pneumonia)
Atrial fibrillation or atrial flutter  (0 to 2 weeks)	We found no data for this outcome
Risk of neurological  complications  (0 to 30 days)	We found no data for this outcome (cerebrovascular accident or epidural haematoma)
Duration of tracheal intubation	One trial with 37 participants published as a conference abstract reported no difference in time to tracheal extubation between epidural analgesia and intravenous patient‐controlled analgesia plus wound infusion (numbers and P value not provided) (very low quality)^a
Pain at rest at 6 to 8 hours (score from 0 to 10)	One trial with 37 participants published as a conference abstract reported lower pain scores with epidural analgesia (numbers and P value not provided) (very low quality)^a
Haemodynamic support (in hospital)	We found no data for this outcome
The corresponding risk* (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval.
GRADE Working Group grades of evidence.  High quality: further research is very unlikely to change our confidence in the estimate of effect.  Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low quality: we are very uncertain about the estimate.

Study	Definition
Aguero‐Martinez 2012	Change > 20% of basal value after local anaesthetic injection
Bach 2002	Mean arterial blood pressure < 55 mmHg
Barrington 2005	Mean arterial blood pressure < 65 mmHg
Berendes 2003	Unspecified
Celik 2015	Intraoperative hypotension
de Vries 2002	Mean arterial blood pressure < 60 mmHg
Fawcett 1997	Mean arterial blood pressure < 60 mmHg
Dohle 2001	Unspecified
Fillinger 2002	Unspecified
Greisen 2012	Unspecified
Gurses 2013	Mean arterial blood pressure < 50 mmHg
Jakobsen 2012	Mean arterial blood pressure < 60 mmHg
Kendall 2004	Variation > 20% from baseline
Kilickan 2006	Systolic arterial blood pressure < 80 mmHg
Kirno 1994	Unspecified
Kirov 2011	Mean arterial blood pressure < 60 mmHg
Kundu 2007	Unspecified
Liem 1992	Change in mean arterial blood pressure ≥ 20% of baseline value
Moore 1995	Mean blood arterial pressure < 50 mmHg
Palomero 2008	Mean blood arterial pressure < 50 mmHg
Royse 2003	Systolic arterial blood pressure ≤ 100 mmHg
Scott 2001	Mean arterial blood pressure ≤ 40 mmHg
Stenseth 1994	Mean arterial blood pressure < 65 mmHg
Stenseth 1996	Mean arterial blood pressure ≤ 65 mmHg
Tenling 1999	Mean arterial blood pressure decreased > 30% from baseline
Tenenbein 2008	Mean arterial blood pressure < 55 mmHg
Volk 2003	Unspecified
Yilmaz 2007	Mean arterial blood pressure < 50 mmHg

Date	Event	Description
19 November 2018	New citation required and conclusions have changed	Conclusions unchanged for outcomes included in the previous version New conclusions provided for new outcomes, along with new comparisons
19 November 2018	New search has been performed	Review updated in 2018 by new authors: Joanne Guay and Sandra Kopp Methodology updated 38 new trials included; 3 new comparisons and 3 outcomes (duration of tracheal intubation, pain, haemodynamic support requirements) added

Date	Event	Description
26 September 2017	New search has been performed	Review undertaken by 2 new review authors
22 September 2017	Amended	Change made to review authors Previous review authors replaced by Joanne Guay and Sandra Kopp
1 July 2013	Amended	Contact details for Geert J. van der Heijden amended

CINAHL, EBSCOhost S1	(MM "Heart Surgery+") OR (MM "Cardiopulmonary Bypass+") OR CABG OR ((coronary or heart or cardiac* OR cardio* or valve) N3 (surg* or graft* OR bypass OR shunt or plasty or replacement)))	63,212
S2	(MM "Analgesia, Epidural+") OR (MM "Anesthesia, Epidural+") OR (MM "Epidural Analgesia Administration (Iowa NIC)") OR (epidural* OR peridural* OR extradural* OR subarachnoid* OR neuraxial*)	18,984
S3	(MH "Placebos") OR ( (MM "Randomized Controlled Trials") OR (MM "Random Assignment") OR (MH "Prospective Studies") OR (MH "Multicenter Studies") OR (MH "Clinical Trials") OR (MH "Single‐Blind Studies") OR (MH "Triple‐Blind Studies") OR (MH "Double‐Blind Studies") ) OR ( placebo* or multicenter or prospective or ((random* or control) and trial) )	755,253
S4	S1 AND S2 AND S3	103
S5	S1 AND S2 AND S3 Published: 2012‐2018	42

# 5	101	#4 Timespan=2012‐2018
# 4	331	#3 AND #2 AND #1
# 3	4,539,758	TS= clinical trial* OR TS=research design OR TS=comparative stud* OR TS=evaluation stud* OR TS=controlled trial* OR TS=follow‐up stud* OR TS=prospective stud* OR TS=random* OR TS=placebo* OR TS=(single blind) OR TS=(double blind)
# 2	19,362	TS=((epidural* OR peridural* OR extradural* OR subarachnoid* OR neuraxial) AND (an?esth or analg*))
# 1	191,475	TS=(((coronary or heart or cardio* or cardiac* or valve) NEAR/5 (surg* or graft* or bypass or plasty or replacement)) or cabg)

Study	Epidural analgesia			Systemic analgesia
Study	Mean	SD	N	Mean	SD	N
El‐Baz 1987	9.00	3.00	30	18.000	5.000	30
Liem 1992	7.72	6.58	25	19.000	4.830	25
Konishi 1995 (1)	6.60	3.70	31	9.200	5.400	18
Konishi 1995 (2)	5.80	3.10	31	9.200	5.400	18
Stenseth 1996	5.40	2.04	26	10.800	3.569	26
Fawcett 1997	5.80	1.00	8	9.200	2.400	8
Loick 1999	9.98	2.65	25	14.630	9.150	21
Tenling 1999	3.62	0.47	14	7.970	3.070	14
Kunstyr 2001 (4)	6.07	2.93	7	7.000	2.990	20
Kunstyr 2001 (5)	6.07	2.93	7	5.810	2.350	21
Kunstyr 2001 (6)	6.07	2.93	7	6.260	3.440	20
Fillinger 2002	10.70	1.40	30	9.500	0.800	30
Berendes 2003	3.40	1.90	36	9.200	4.300	37
Royse 2003	2.60	2.50	37	5.400	3.100	39
Huh 2004	4.61	4.75	27	13.430	7.010	29
Hansdottir 2006	2.30	1.10	53	7.300	19.200	55
Kilickan 2006 (7)	7.57	6.05	20	14.620	4.200	20
Kilickan 2006 (8)	10.00	5.32	20	8.520	4.720	20
Petrovski 2006	3.50	0.80	56	6.800	0.700	54
Yilmaz 2007	7.44	1.36	17	9.370	1.980	17
Tenenbein 2008	0.26	0.63	25	0.170	0.210	25
Kilickan 2008 (10)	5.00	3.20	15	6.600	4.000	15
Kilickan 2008 (11)	5.50	2.60	15	7.000	3.300	15
Palomero 2008	11.70	7.52	10	12.500	2.400	12
Lenkutis 2009	6.04	0.56	30	11.060	1.640	30
El‐Morsy 2012	7.30	6.40	25	10.700	8.200	25
Onan 2013	2.90	1.10	20	4.700	1.200	20
Neskovic 2013 (12)	6.67	4.66	18	8.830	5.270	27
Gurses 2013	4.10	1.70	32	6.800	2.000	32
Celik 2015	7.20	1.82	20	11.700	2.020	20
de Vries 2002 (13)	0.15	0.08	28	0.220	0.120	29
Kendall 2004 (14)	5.30	4.10	5	6.900	2.800	10
Kendall 2004 (15)	5.30	4.10	5	6.600	3.100	10
Bakhtiary 2007	6.00	2.30	66	7.000	4.200	66
Sharma 2010	9.33	2.24	30	11.670	3.020	30
Mehta 2010	10.80	3.19	31	13.500	2.880	31
Aguero‐Martinez 2012	4.52	2.84	29	7.830	5.240	30
Neskovic 2013 (16)	4.38	4.31	17	5.810	3.500	19

	Epidural analgesia	Systemic analgesia
Mean (hours)	6.1	9.1
Std. deviation	3.0	4.0

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Mortality at 0 to 30 days	38	3418	Risk Difference (IV, Fixed, 95% CI)	0.00 [‐0.01, 0.01]
1.1 With cardiopulmonary bypass	28	1844	Risk Difference (IV, Fixed, 95% CI)	‐0.01 [‐0.02, 0.01]
1.2 Off‐pump surgery	8	729	Risk Difference (IV, Fixed, 95% CI)	0.00 [‐0.01, 0.02]
1.3 With and without cardiopulmonary bypass	3	845	Risk Difference (IV, Fixed, 95% CI)	0.00 [‐0.01, 0.01]
2 Mortality at 6 months	7	407	Risk Difference (IV, Fixed, 95% CI)	‐0.00 [‐0.03, 0.03]
2.1 With cardiopulmonary bypass	7	407	Risk Difference (IV, Fixed, 95% CI)	‐0.00 [‐0.03, 0.03]
3 Mortality at 1 year	5	849	Risk Difference (M‐H, Fixed, 95% CI)	‐0.01 [‐0.03, 0.00]
3.1 With cardiopulmonary bypass	4	197	Risk Difference (M‐H, Fixed, 95% CI)	‐0.02 [‐0.07, 0.03]
3.2 With and without cardiopulmonary bypass	1	652	Risk Difference (M‐H, Fixed, 95% CI)	‐0.01 [‐0.03, 0.01]
4 Myocardial infarction (0 to 30 days)	26	2713	Risk Difference (IV, Fixed, 95% CI)	‐0.01 [‐0.02, 0.00]
4.1 With cardiopulmonary bypass	16	1153	Risk Difference (IV, Fixed, 95% CI)	‐0.01 [‐0.03, 0.01]
4.2 Off‐pump surgery	8	713	Risk Difference (IV, Fixed, 95% CI)	‐0.01 [‐0.03, 0.01]
4.3 With and without cardiopulmonary bypass	3	847	Risk Difference (IV, Fixed, 95% CI)	‐0.00 [‐0.03, 0.02]
5 Respiratory complications: respiratory depression (0 to 30 days)	21	1736	Risk Difference (M‐H, Fixed, 95% CI)	‐0.03 [‐0.05, ‐0.01]
5.1 With cardiopulmonary bypass	15	1246	Risk Difference (M‐H, Fixed, 95% CI)	‐0.04 [‐0.07, ‐0.01]
5.2 Off‐pump surgery	5	299	Risk Difference (M‐H, Fixed, 95% CI)	‐0.01 [‐0.05, 0.02]
5.3 With and without cardiopulmonary bypass	2	191	Risk Difference (M‐H, Fixed, 95% CI)	0.01 [‐0.06, 0.08]
6 Respiratory complications: pneumonia (0 to 30 days)	10	1107	Risk Difference (M‐H, Random, 95% CI)	‐0.03 [‐0.07, 0.01]
6.1 With cardiopulmonary bypass	7	677	Risk Difference (M‐H, Random, 95% CI)	‐0.03 [‐0.09, 0.04]
6.2 Off‐pump surgery	3	322	Risk Difference (M‐H, Random, 95% CI)	‐0.02 [‐0.07, 0.04]
6.3 With and without cardiopulmonary bypass	1	108	Risk Difference (M‐H, Random, 95% CI)	‐0.05 [‐0.12, 0.01]
7 Atrial fibrillation or flutter within 2 weeks	18	2431	Risk Difference (M‐H, Random, 95% CI)	‐0.06 [‐0.10, ‐0.01]
7.1 With cardiopulmonary bypass	11	1118	Risk Difference (M‐H, Random, 95% CI)	‐0.04 [‐0.08, 0.00]
7.2 Off‐pump surgery	6	551	Risk Difference (M‐H, Random, 95% CI)	‐0.09 [‐0.22, 0.03]
7.3 With and without cardiopulmonary bypass	2	762	Risk Difference (M‐H, Random, 95% CI)	‐0.04 [‐0.11, 0.03]
8 Neurological complications: cerebrovascular accident (0 to 30 days)	18	2232	Risk Difference (M‐H, Fixed, 95% CI)	‐0.00 [‐0.01, 0.01]
8.1 With cardiopulmonary bypass	13	1067	Risk Difference (M‐H, Fixed, 95% CI)	‐0.01 [‐0.02, 0.01]
8.2 Off‐pump surgery	4	403	Risk Difference (M‐H, Fixed, 95% CI)	‐0.00 [‐0.03, 0.03]
8.3 With and without cardiopulmonary bypass	2	762	Risk Difference (M‐H, Fixed, 95% CI)	0.00 [‐0.01, 0.01]
9 Neurological complications: epidural haematoma (0 to 30 days)	53	3982	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.01, 0.01]
9.1 With cardiopulmonary bypass	39	2231	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.01, 0.01]
9.2 Off‐pump surgery	10	841	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.02, 0.02]
9.3 With and without cardiopulmonary bypass	4	910	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.01, 0.01]
10 Duration of tracheal intubation	40	3353	Std. Mean Difference (Random, 95% CI)	‐0.78 [‐1.01, ‐0.55]
10.1 Cardiopulmonary bypass	27	1570	Std. Mean Difference (Random, 95% CI)	‐0.75 [‐1.03, ‐0.47]
10.2 Off‐pump surgery	11	943	Std. Mean Difference (Random, 95% CI)	‐0.90 [‐1.38, ‐0.41]
10.3 With and without cardiopulmonary bypass	3	840	Std. Mean Difference (Random, 95% CI)	‐0.60 [‐1.42, 0.23]
11 Duration of tracheal intubation in hours (for studies for which means and standard deviations could be extracted)	33	2062	Mean Difference (IV, Random, 95% CI)	‐2.91 [‐3.61, ‐2.21]
11.1 With cardiopulmonary bypass	23	1249	Mean Difference (IV, Random, 95% CI)	‐3.23 [‐4.30, ‐2.17]
11.2 Off‐pump surgery	9	627	Mean Difference (IV, Random, 95% CI)	‐1.87 [‐3.36, ‐0.37]

	Epidural analgesia	Systemic analgesia
Mean	1.92	4.17
SD	1.80	1.70

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Mortality at 0 to 30 days	1	145	Risk Difference (M‐H, Fixed, 95% CI)	‐0.03 [‐0.08, 0.02]
1.1 With cardiopulmonary bypass	1	145	Risk Difference (M‐H, Fixed, 95% CI)	‐0.03 [‐0.08, 0.02]
2 Myocardial infarction (0 to 30 days)	2	76	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.07, 0.07]
2.1 Off‐pump surgery	2	76	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.07, 0.07]
3 Neurological complications: cerebrovascular accident (0 to 30 days)	1	145	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.03, 0.03]
3.1 With cardiopulmonary bypass	1	145	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.03, 0.03]
4 Neurological complications: epidural haematoma (0 to 30 days)	4	271	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.03, 0.03]
4.1 With cardiopulmonary bypass	2	195	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.03, 0.03]
4.2 Off‐pump surgery	2	76	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.07, 0.07]
5 Duration of tracheal intubation (hours)	4	271	Mean Difference (IV, Fixed, 95% CI)	‐0.08 [‐0.54, 0.38]
5.1 With cardiopulmonary bypass	2	195	Mean Difference (IV, Fixed, 95% CI)	‐0.04 [‐0.53, 0.44]
5.2 Off‐pump surgery	2	76	Mean Difference (IV, Fixed, 95% CI)	‐0.35 [‐1.70, 0.99]
6 Pain scores at rest at 6 to 8 hours	2	90	Mean Difference (IV, Fixed, 95% CI)	0.12 [‐0.42, 0.66]
6.1 With cardiopulmonary bypass	1	50	Mean Difference (IV, Fixed, 95% CI)	0.0 [‐0.59, 0.59]
6.2 Off‐pump surgery	1	40	Mean Difference (IV, Fixed, 95% CI)	0.8 [‐0.61, 2.21]
7 Pain scores on movement/coughing at 6 to 8 hours	2	90	Mean Difference (IV, Fixed, 95% CI)	‐0.15 [‐0.69, 0.39]
7.1 With cardiopulmonary bypass	1	50	Mean Difference (IV, Fixed, 95% CI)	‐0.08 [‐0.69, 0.53]
7.2 Off‐pump surgery	1	40	Mean Difference (IV, Fixed, 95% CI)	‐0.40 [‐1.57, 0.77]
8 Pain at rest at 24 hours	3	231	Mean Difference (IV, Random, 95% CI)	0.11 [‐0.41, 0.63]
8.1 With cardiopulmonary bypass	2	195	Mean Difference (IV, Random, 95% CI)	0.28 [‐0.34, 0.91]
8.2 Off‐pump surgery	1	36	Mean Difference (IV, Random, 95% CI)	‐0.54 [‐1.38, 0.30]
9 Pain on movement/coughing at 24 hours	2	86	Mean Difference (IV, Random, 95% CI)	0.31 [‐0.62, 1.24]
9.1 With cardiopulmonary bypass	1	50	Mean Difference (IV, Random, 95% CI)	0.72 [0.22, 1.22]
9.2 Off‐pump surgery	1	36	Mean Difference (IV, Random, 95% CI)	‐0.24 [‐1.11, 0.63]
10 Pain at rest at 48 hours	2	195	Mean Difference (IV, Random, 95% CI)	0.51 [‐0.77, 1.80]
10.1 With cardiopulmonary bypass	2	195	Mean Difference (IV, Random, 95% CI)	0.51 [‐0.77, 1.80]
11 Pain at rest on movement/coughing at 48 hours	1	50	Mean Difference (IV, Fixed, 95% CI)	1.36 [0.76, 1.96]
12 Hypotension or need for vasopressor	1	40	Risk Difference (M‐H, Random, 95% CI)	0.05 [‐0.08, 0.18]
12.1 Off‐pump surgery	1	40	Risk Difference (M‐H, Random, 95% CI)	0.05 [‐0.08, 0.18]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Myocardial infarction (0 to 30 days)	1	50	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.07, 0.07]
1.1 Off‐pump surgery	1	50	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.07, 0.07]
2 Neurological complications: epidural haematoma (0 to 30 days)	1	50	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.07, 0.07]
2.1 Off‐pump surgery	1	50	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.07, 0.07]
3 Duration of tracheal intubation (hours)	1	15	Mean Difference (IV, Fixed, 95% CI)	‐0.30 [‐1.20, 0.60]
3.1 Off‐pump surgery	1	15	Mean Difference (IV, Fixed, 95% CI)	‐0.30 [‐1.20, 0.60]
4 Pain scores at rest at 6 hours	1	50	Mean Difference (IV, Fixed, 95% CI)	0.84 [0.31, 1.37]
4.1 Off‐pump surgery	1	50	Mean Difference (IV, Fixed, 95% CI)	0.84 [0.31, 1.37]

Methods	Parallel RCT Ethics committee: approved by the institutional ethics committee Informed consents: written informed consents obtained Site: Surgical‐Clinic Hermanos Ameijeiras Hospital, Cuba, Havana Setting: university hospital Dates of data collection: between September 2008 and March 2010 Funding: departmental Registration: RPCEC00000131 2012
Participants	Adults undergoing off‐pump CABG; mean age: 60.2; sex distribution: 11 females and 49 males Inclusion criteria No previous cardiac surgery LVEF > 45% No intra‐aortic balloon pump support Dysrhythmias or neurological disease Urine output > 0.5 mL/kg and creatinine < 132 mmol/L Normal chest x‐ray Exclusion criteria Not consenting Absolute contraindication to regional anaesthesia Myocardial infarction within the last 30 days Inotropic drug Heart failure Pulmonary hypertension or chronic obstructive lung disease
Interventions	Intervention Single injection epidural analgesia (N = 30) Comparator Systemic analgesia (N = 30) Premedication: IV midazolam 0.05 mg/kg Induction: midazolam, fentanyl, lidocaine, and atracurium Maintenance: propofol, isoflurane, and atracurium Surgery: off‐pump CABG
Outcomes	Relevant to this review Risk of mortality Risk of myocardial infarction Risk of in‐hospital pulmonary complications (respiratory insufficiency) Risk of atrial fibrillation or atrial flutter during surgery Pain scores Haemodynamic variables Others Complications related to regional anaesthesia ICU length of stay Hospital length of stay Costs
Notes	Correspondence: information received from study authors Conflict of interest: no conflict of interest DOI: n/a The trial also contains a third group with intrathecal analgesia not retained in the review
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized with a computer‐generated table
Allocation concealment (selection bias)	Low risk	Sealed opaque envelopes
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Quote: "patients were not blinded"
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "blinded"
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No loss to follow‐up
Selective reporting (reporting bias)	Low risk	No failed epidural reported
Other bias	Unclear risk	The group given systemic analgesia included more aged participants

Methods	Parallel randomized controlled trial Ethics committee: approved by the ethics committee Informed consents: written informed consents obtained Site: St. Vincent’s Hospital, Melbourne, Australia Setting: university hospital Dates of data collection: from December 1999 to March 2002 Funding: grants from the Australian Society of Anaesthetists and the Australian and New Zealand College of Anaesthetists Registration: unspecified
Participants	120 participants scheduled for elective coronary artery bypass grafting surgery; mean age 62.5 years; sex distribution: 16 females and 104 males Inclusion criteria Patients scheduled for elective CABG surgery (using cardiopulmonary bypass (CPB)) were eligible Exclusion criteria Emergency or repeat CABG surgery Combined valve and CABG surgery Aspirin ingestion within 6 days of surgery Platelet count 150 × 10⁹/L International normalized ratio 1.1 Active neurological disease Cutaneous disorders at the epidural insertion site
Interventions	Intervention Epidural analgesia (N = 60) Comparator Systemic analgesia (N = 60) Premedication: temazepam, ranitidine, and morphine Induction: midazolam, fentanyl, propofol, and rocuronium Maintenance: propofol Surgery: CABG with CPB using a membrane oxygenator
Outcomes	Relevant to this review Risk of mortality Risk of myocardial infarction Risk of pulmonary complications Risk of atrial fibrillation or atrial flutter Risk of neurological complications (cerebrovascular accident) Tracheal extubation Pain scores Haemodynamic variables Others Arterial blood PaO₂ and PaCO₂
Notes	Correspondence: email sent 16 March 2018; no reply Conflict of interest: none reported DOI: 10.1213/01.ANE.0000146437.88485.47
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Patients were randomized the day before surgery to 2 groups. The random allocation sequence was computer‐generated in permuted blocks of 4 and was enclosed in sequentially numbered opaque sealed envelopes
Allocation concealment (selection bias)	Low risk	Opaque sealed envelopes
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Not reported
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Twelve‐lead ECGs were recorded before surgery and on postoperative days 1 and 5 and were assessed by 2 observers blinded to group allocation and postoperative clinical course. No mention of blinding for any other outcome
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No loss to follow‐up. Two participants with failed epidural were kept in the analysis
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported
Other bias	Unclear risk	All participants were included in the intention‐to‐treat analysis. Prevalence of cerebrovascular and peripheral vascular disease was more frequent in the epidural group

Methods	Parallel RCT Ethics committee: not reported Informed consents: not reported Site: Far Eastern Memorial Hospital, New Taipei City, Taiwan Setting: university hospital Dates of data collection: unspecified Funding: unspecified Registration: unspecified
Participants	37 participants; mean age: not reported; sex distribution: not reported Inclusion criteria Undergoing minimally invasive cardiac surgery Exclusion criteria Not reported
Interventions	Intervention Epidural analgesia (N = 18) Comparator Wound local anaesthetic infusion plus IV PCA (N = 19) Induction and maintenance: not reported Surgery: off‐pump CABG or valve surgery
Outcomes	Relevant to this review Tracheal extubation Pain scores Others ICU length of stay Hospital length of stay
Notes	Conflict of interest: not reported Correspondence: email sent 16 March 2018; no reply DOI: n/a Conference abstract, limited information
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "randomly allocated"; no details provided
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Not reported
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Conference abstract; limited details provided
Selective reporting (reporting bias)	Low risk	Conference abstract; limited details provided
Other bias	Unclear risk	Conference abstract; limited details provided

Methods	Parallel RCT Ethics committee: approved Informed consents: obtained Site: New Delhi, India Setting: university hospital Dates of data collection: not reported Funding: unspecified Registration: unspecified
Participants	41 participants, for participants included in the analysis: mean age: 56 years; sex distribution: 7 females and 33 males Inclusion criteria Consenting patients undergoing minimally invasive direct coronary artery bypass surgery Exclusion criteria Ejection fraction 35%, with an anomaly of the vertebral column Receiving heparin Receiving antiplatelet medications within the last week With significant respiratory disease Requiring inotropic support or intra‐aortic balloon counterpulsation
Interventions	Intervention Thoracic epidural analgesia (N = 21) Comparator Paravertebral blockade (N = 20) Premedication: lorazepam and morphine Induction and maintenance: midazolam, fentanyl (total dose 5 mcg/kg), nitrous oxide, isoflurane, and vecuronium Surgery: off‐pump CABG
Outcomes	Relevant to this review Risk of myocardial infarction Tracheal extubation Risk of serious neurological complications from epidural analgesia Pain scores Haemodynamic variables Others Respiratory function
Notes	Correspondence: email sent 16 March 2018; no reply Conflict of interest: not reported DOI: 10.1053/jcan.2001.23271
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "randomized study"; no details provided
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Not reported
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "an independent observer who was blinded to the analgesia technique recorded"
Incomplete outcome data (attrition bias)   All outcomes	Low risk	One failed epidural
Selective reporting (reporting bias)	Low risk	All results reported
Other bias	Low risk	Not in intention‐to‐treat Groups had similar demographic data

Methods	Parallel RCT Ethics committee: the Human Ethics Committee of the Sahlgrenska Academy, Goteborg University, Goteborg, Sweden, approved the study protocol Informed consents: all participants gave written informed consent Site: Goteberg, Sweden Setting: university hospital Dates of data collection: from 1 April 2002 to 31 December 2003 Funding: support was provided solely from institutional and/or departmental sources Registration: unspecified
Participants	113 participants; mean age: 66.5 years; sex distribution: 37 females and 76 males Inclusion criteria Patients undergoing elective cardiac surgery (CABG, cardiac valve procedures, combined CABG and valve procedures, or the Maze procedure, with or without CABG) Exclusion criteria Contraindications to epidural anaesthesia Abnormal coagulation tests (i.e. partial thromboplastin time > 45 s or prothrombin time (international normalized ratio) > 1.5 or platelet count < 80,000) Recent (1 week) treatment with thrombolytic or potent antiplatelet drugs (streptokinase, alteplase, clopidogrel, abciximab, tirofiban, integrelin) Aspirin treatment was not considered a contraindication to placement of a thoracic epidural catheter
Interventions	Intervention Epidural analgesia (N = 55) Comparator Systemic analgesia (N = 55) Premedication: midazolam Induction and maintenance: propofol, remifentanil, and atracurium Surgery: CABG, valve procedures, or both with CPB using a membrane oxygenator
Outcomes	Relevant to this review Myocardial infarction Risk of atrial fibrillation or atrial flutter Risk of neurological complications: cerebrovascular accident Pain scores Others Sedation scores Lung function Quality recovery score Length of hospital stay
Notes	Correspondence: email sent 16 March 2018; no reply Conflict of interest: not reported DOI: n/a
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were randomly assigned the day before surgery to 1 of 2 regimens
Allocation concealment (selection bias)	Low risk	Sealed envelopes
Blinding of participants and personnel (performance bias)   All outcomes	High risk	The infusion bag (250 mL) of the patient‐controlled analgesia pump was changed only once during the postoperative treatment period (72 h) by the nursing team, which was neither blinded to treatment nor involved in assessment of patients The decision to allow hospital discharge was made by the surgical team not blinded to treatment
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Atelectasis was defined as new area(s) of lobar or sublobar atelectatic consolidation with an air bronchogram by a radiologist blinded to treatment Evaluation of quality of recovery score, level of mobilization, pain, degree of sedation, lung function, and eligibility for hospital discharge was performed between 1:00 and 3:00 PM each day by either of two investigators. These investigators were blinded to the assigned treatment The blinded investigators were not involved in nursing of the participants Less than 5% of epidural participants revealed by mistake to the blinded observer the presence of the epidural catheter
Incomplete outcome data (attrition bias)   All outcomes	Low risk	113 participants were randomized, 110 participants received allocated treatment, and 97 participants were eventually analysed
Selective reporting (reporting bias)	Low risk	All results were reported
Other bias	Unclear risk	Three participants were excluded because of inability to place the epidural catheter. In 1 of these participants, the catheter was positioned intradurally, and another participant did not co‐operate. A malfunctioning epidural catheter was considered in 7 participants after extubation. Three of these participants had the epidural catheter replaced in the ICU, and 4 were treated with intravenous patient‐controlled analgesia with morphine These 7 participants were analysed on an intention‐to‐treat basis Groups had similar demographic data, except for a higher incidence of off‐pump CABG in the epidural group and a longer cardiopulmonary bypass time in the systemic analgesia group

Methods	Parallel RCT Ethics committee: the study was approved by the authors’ hospital’s Medical Ethics Committee Informed consents: written informed consents were obtained Site: Maastricht, The Netherlands Setting: university hospital Dates of data collection: unspecified Funding: unspecified Registration: unspecified
Participants	60 participants; mean age: 60 years; sex distribution: not reported Inclusion criteria Undergoing elective cardiac surgery Exclusion criteria Left ventricular ejection fraction < 25% Hypothermic circulatory arrest Recent myocardial infarction Preoperative inotropic or intra‐aortic balloon pump metabolic Neurological diseases
Interventions	Intervention Epidural analgesia (N = 15) Comparator Systemic analgesia (N = 45) Premedication: midazolam Induction and maintenance: propofol, remifentanil, or alfentanil and pancuronium Surgery; CABG with CPB using a hollow‐fibre membrane oxygenator
Outcomes	Relevant to this review Risk of mortality Risk of atrial fibrillation or atrial flutter Risk of neurological complications (cerebrovascular accident) Others Inflammation markers
Notes	Correspondence: letter sent 16 March 2018; no reply Conflict of interest: not reported DOI: 10.1053/j.jvca.2007.02.008
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "60 patients scheduled to undergo coronary artery bypass surgery were randomized"; no details
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Quote: "the study was blinded for the opioid infusion, except in the thoracic epidural group"
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Quote: "the study was blinded for the opioid infusion, except in the thoracic epidural group"
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No loss to follow‐up mentioned
Selective reporting (reporting bias)	Low risk	All results were reported
Other bias	Low risk	No failed epidural mentioned Groups had similar demographic data

Methods	Parallel RCT Ethics committee approval: approved by the Central Denmark Region Committee on Biomedical Research Ethics and the Danish Medicine Agency Informed consents: written informed consents were obtained from all participants Site: Aarhus University Hospital, Skejby, Aarhus, Denmark Setting: university hospital Dates of data collection: unspecified Funding: unspecified Registration: Eudra CT 2005‐000617‐35
Participants	60 participants; mean age: 71.3 years; sex distribution: 21 females and 39 males Inclusion criteria Low‐ to moderate‐risk participants between the ages of 65 and 80 years scheduled for CABG with or without AVR Exclusion criteria Ejection fraction < 0.3 Myocardial infarction within the last 4 weeks Diabetes Severe pulmonary or arterial hypertension Contraindication for TEA No preoperative optimal echocardiographic imaging
Interventions	Intervention Epidural analgesia (N = 30) Comparator Systemic analgesia (N = 30) Induction and maintenance: propofol or sevoflurane, sufentanil, and rocuronium Surgery: CABG, valve procedure, or both with CPB, using a hollow‐fibre membrane oxygenator
Outcomes	Relevant to this review Risk of mortality Risk of myocardial infarction Risk of neurological complication: cerebrovascular accident Time to tracheal extubation Haemodynamic variables Others Rescue analgesics Acute kidney injury: renal function expressed as changes in s‐creatinine: fewer TEA patients (13.3% vs 36.7%; P = 0.074, Chi² test) developed acute kidney injury ICU length of stay (hours from arrival in the ICU to discharge to the surgical ward) Time to eligible to ICU discharge (predefined scoring system evaluated at regular intervals) Hospital length of stay
Notes	Conflict of interest: study authors declare that they have no competing interests Correspondence: email sent 16 March 2018; no reply DOI: 10.1053/j.jvca.2012.05.007 and 10.1053/j.jvca.2012.05.008
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomization by the standard envelope method with blocks of 20 participants was performed immediately before insertion of the epidural catheter the day before surgery
Allocation concealment (selection bias)	Low risk	Randomization by the standard envelope method with blocks of 20 participants was performed immediately before insertion of the epidural catheter the day before surgery
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Not reported
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Sixty‐three patients were approached; 2 declined, and 1 was excluded because surgery was changed to off‐pump CABG surgery
Selective reporting (reporting bias)	Unclear risk	Quote: "there were no significant differences in blood loss, urine output, administration of crystalloids and adverse events (not shown)"
Other bias	Low risk	Groups well balanced for preoperative characteristics One participant in the TEA group did not receive a functional epidural, but because the protocol was based on an intention‐to‐treat principle, this participant was analysed in the TEA group

PERMALINK

Epidural analgesia for adults undergoing cardiac surgery with or without cardiopulmonary bypass

Joanne Guay

Sandra Kopp

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

1. Postoperative analgesia.

Types of outcome measures

Primary outcomes

Secondary outcomes

2. Diagnostic criteria for myocardial infarction.

3. Diagnostic criteria for pulmonary complications.

4. Diagnostic criteria for neurological complications: cerebrovascular accident.

5. Criteria for tracheal extubation.

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

'Summary of findings’ table and GRADE

Results

Description of studies

Results of the search

1.

Included studies

Source of funding

Setting

Participants

Interventions

Comparators

Excluded studies

Studies awaiting classification

Ongoing trials

Risk of bias in included studies

2.

3.

Allocation

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

Risk of bias for each study

Effects of interventions

Summary of findings for the main comparison. Epidural analgesia compared with systemic analgesia for cardiac surgery with or without cardiopulmonary bypass in adults.

Summary of findings 2. Epidural analgesia compared with peripheral nerve blocks for cardiac surgery in adult.

Summary of findings 3. Epidural analgesia compared with intrapleural analgesia for cardiac surgery in adults.

Summary of findings 4. Epidural analgesia compared with wound infiltration for cardiac surgery in adults.

Comparison 1: epidural analgesia compared with systemic analgesia

Primary outcomes

1. Risk of mortality

1a. Mortality at 0 to 30 days

Methods	Parallel RCT  Ethics committee: approved by the institutional review committee  Informed consents: patient consents were obtained  Site: Kocaeli University School of Medicine, Kocaeli, Turkey  Setting: university hospital  Dates of data collection: unspecified  Funding: unspecified  Registration: unspecified
Participants	80 participants; mean age: 59.9 years; sex distribution: 16 females and 64 males Inclusion criteria Undergoing elective CABG with CPB Exclusion criteria Compromised coagulation (thromboplastin time < 80%, prothrombin time > 40 seconds, or platelets < 100 nL)
Interventions	Intervention Epidural analgesia with LVEF ≤ 0.4 (N = 20) or > 0.4 (N = 20) Comparator Systemic analgesia with LVEF ≤ 0.4 (N = 20) or > 0.4 (N = 20) Premedication: midazolam  Induction: midazolam, fentanyl, and vecuronium  Maintenance: nitrous oxide, propofol, and fentanyl  Suregry: CABG with CPB
Outcomes	Relevant to this review Risk of mortality Risk of myocardial infarction Risk of atrial fibrillation or atrial flutter Tracheal extubation Pain scores Haemodynamic variables Other Right atrial biopsies
Notes	Correspondence: email sent 16 March 2018; no reply  Conflict of interest: not reported  DOI: n/a
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomly distributed sealed envelopes
Allocation concealment (selection bias)	Low risk	Sealed envelopes
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Study was not blinded
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Study was not blinded
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No loss to follow‐up
Selective reporting (reporting bias)	Low risk	All results reported
Other bias	Low risk	No failed epidural  Groups well balanced

Methods	Parallel RCT Ethics committee: study protocol and informed consent form were approved by the Ethics Committee of Northern State Medical University, Arkhangelsk, Russian Federation Informed consents: written informed consent was obtained from every patient Site: Northern State Medical University, Troitsky Avenue 51, Arkhangelsk, 163000, Russian Federation; University of Tromsø, MH‐Breivika, Tromsø, 9038, Norway; and University Hospital of North Norway, Sykehusveien 38, Tromsø, 9038, Norway Setting: university hospital Dates of data collection: from January 2008 to September 2009 Funding: supported by a grant from the Government of Arkhangelsk region, “Young Pomor scientists”, and departmental funds Registration: NCT01384175
Participants	93 participants; for the participants included in the analysis, mean age: 55.6 years; sex distribution: 42 females and 48 males Inclusion criteria ASA III adult patients with coronary artery disease ASA III and scheduled for elective off‐pump coronary artery bypass grafting Exclusion criteria Age < 18 years Severe valve dysfunction Peripheral vascular disease Simultaneous interventions (carotid endarterectomy, aneurysm repair, etc.) Transfer to CPB during surgery
Interventions	Intervention Epidural analgesia as a continuous infusion (N = 31) or as PCEA (N = 31) Comparator Systemic analgesia (N = 31) Premedication: diazepam Induction: fentanyl, propofol, and pipecuronium Maintenance: propofol, fentanyl, and pipecuronium Surgery: off‐pump CABG
Outcomes	Relevant to this review Risk of mortality at 28 days Risk of myocardial infarction (troponin‐T) Risk of atrial fibrillation or atrial flutter (no difference between groups) Tracheal extubation Pain scores Haemodynamic variables Others Sedation scores ICU length of stay Hospital length of stay
Notes	Correspondence: information received from study authors Conflict of interest: study authors declare that they have no competing interests DOI: 10.1186/1471‐2253‐11‐17
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were randomized to 3 groups, using the envelope method
Allocation concealment (selection bias)	Low risk	Participants were randomized to 3 groups, using the envelope method
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Not blinded
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Not blinded
Incomplete outcome data (attrition bias)   All outcomes	Low risk	One participant was withdrawn from analysis in each group due to protocol violation (transfer to CPB)
Selective reporting (reporting bias)	Low risk	All results reported
Other bias	Low risk	No failed epidural reported Groups well balanced

Methods	Parallel RCT Ethics committee: unspecified Informed consents: unspecified Site: New Tokyo Hospital, Matsudo, Japan Setting: university hospital Dates of data collection: October 1993 to March 1994 Funding: unspecified Registration: unspecified
Participants	97 participants: mean age 64 years: sex distribution: not reported Inclusion criteria Heart surgery patients with average age of 64 years Exclusion criteria Not available from the partial translation
Interventions	Intervention Epidural analgesia with butorphanol (N = 31) or morphine (N = 31) Comparator Systemic analgesia (N = 35) Induction and maintenance: low‐dose fentanyl, nitrous oxide, and isoflurane Surgery: mainly CABG with CPB
Outcomes	Relevant to this review Tracheal extubation Haemodynamic variables Others Rescue analgesia Blood gas ICU length of stay
Notes	Correspondence: email sent 16 March 2018; no reply Conflict of interest: not reported DOI: n/a
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote. "divided"; no details provided
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Not reported
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No loss to follow‐up mentioned
Selective reporting (reporting bias)	Low risk	All results reported
Other bias	Unclear risk	No failed epidural mentioned Some participants had laparotomy to take the gastroepiploic artery used for coronary grafting

Methods	Parallel RCT Ethics committee: obtained Informed consents: not reported Site: Clinic for Anesthesiology and Intensive Care, Military Medical Academy, Belgrade, Serbia; Faculty of Medicine of the Military Medical Academy, University of Defence, Belgrade, Serbia; and Dedinje Cardiovascular Institute, Belgrade, Serbia Setting: university hospital Dates of data collection: February 2002 to October 2005 Funding: departmental (academic trial; part of a PhD thesis) Registration: not registered
Participants	82 participants; mean age: 54.8 years; sex distribution: 13 females and 68 males and 1 unclear Inclusion criteria Scheduled for coronary artery bypass surgery (more than 1 graft) LVEF > 30% No contraindication to TEA Exclusion criteria Acute infection Immunological disease Myocardial infarction up to 1 month before surgery Diabetes mellitus type 1 Acute or chronic renal failure Chronic lung disease Stroke or transitory ischaemic attack Coagulation disorders
Interventions	Intervention Epidural analgesia with off‐pump CABG (N = 17) or CABG with CPB (N = 18) Comparator Systemic analgesia and off‐pump CABG (N = 19) or CABG with CPB (N = 27) Induction: midazolam, propofol, fentanyl, and pancuronium Maintenance: propofol, fentanyl, and pancuronium Surgery: CABG with or without CPB
Outcomes	Relevant to this review Risk of mortality Others Blood loss and transfusion requirements ICU length of stay Hospital length of stay
Notes	Correspondence: information received from study authors Conflict of interest: none DOI: 10.2298/VSP1305439N
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated randomization
Allocation concealment (selection bias)	Low risk	Allocation concealment was done by: (quote) "envelopes"
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Quote: "open label study"
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Quote: "open label study"
Incomplete outcome data (attrition bias)   All outcomes	Low risk	One of the participants had incomplete data and was excluded from further statistical analysis
Selective reporting (reporting bias)	Unclear risk	Initially, the study was designed for a larger number of participants, but for technical reasons, enrolment of patients was stopped earlier (82 participants) Otherwise, all results were reported
Other bias	Unclear risk	Not in intention‐to‐treat: there were 3 conversions from off‐pump to standard surgery and cardiopulmonary bypass; these participants were assigned to different groups according to the anaesthetic technique applied Groups well balanced except for LVEF

Methods	Parallel RCT Ethics committees: approved by the Scientific Ethics Committees for Copenhagen and Frederiksberg (KF 02‐124/98), the Danish Data Protection Agency, and the Danish Medicines Agency Informed consents: written informed consents obtained Site: Copenhagen, Denmark Setting: university hospital Dates of data collection: not reported Funding: supported by The Danish Heart Foundation (grant no. 99‐2‐3‐79‐22764 and no. 99‐1‐5‐92‐22709) and an unrestricted grant from AstraZeneca, Denmark Registration: unspecified
Participants	163 participants; mean age: 64.4 years; sex distribution: 18 females and 145 females Inclusion criteria Patients scheduled for elective CABG; sinus rhythm Exclusion criteria Off‐pump surgery Implanted pacemaker Use of amiodarone within 4 months of enrolment History of amiodarone toxicity Known thyroid disease Liver disease Uncontrolled heart failure Resting heart rate < 50 beats/min in the absence of medical therapy known to slow the heart rate Anticoagulant medication with warfarin Coagulopathy Pregnancy Use of antiarrhythmic drugs other than alpha₁‐receptor antagonists Calcium channel antagonists Digoxin.
Interventions	Intervention Epidural analgesia with (N = 35) or without amiodarone (N = 44) Comparator Systemic analgesia with (N = 36) or without amiodarone (N = 48) Induction: midazolam, fentanyl, and pancuronium Maintenance: isoflurane and fentanyl or epidural analgesia Surgery: CABG with CPB
Outcomes	Relevant to this review Risk of mortality Risk of atrial fibrillation or atrial flutter Haemodynamic variables Other Length of hospital stay
Notes	Correspondence: letter sent 16 March 2018; no reply Conflict of interest: none other than the grants received DOI: 10.1053/j.jvca.2004.08.006
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were randomly assigned to 4 groups; randomization was 1:1:1:1. Randomization list was generated from a computerized table of random numbers
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Quote: "the study was conducted in an open manner"
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Quote: "the study was conducted in an open manner"
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Of the 196 patients included, 163 were evaluated: 18 patients had surgery cancelled, and 4 patients had a change in surgical procedure. One withdrew consent preoperatively, and 6 withdrew consent postoperatively. One patient had a stroke before surgery, and in 1 patient placement of the epidural catheter was unsuccessful. Two patients were excluded because of protocol violations
Selective reporting (reporting bias)	Low risk	Appears to be free of other sources of bias. Sample size calculation stated
Other bias	Unclear risk	Not in intention‐to‐treat Groups had similar demographic data

Methods	Parallel RCT Ethics committee: the hospital ethics committee approved the study Informed consents: all participants gave written informed consent Site: Glasgow, Scotland, UK Setting: university hospital Dates of data collection: unspecified Funding: departmental/institutional Registration: unspecified
Participants	408 participants; mean age: 59 years; sex distribution: 56 females and 352 males Inclusion criteria Patients undergoing elective CABG Normal coagulation screen LVEF > 35% Exclusion criteria Abnormal preoperative coagulation screen that included prothrombin time, international normalized ratio, fibrinogen, platelet count, and activated partial thromboplastin time
Interventions	Intervention Epidural analgesia (N = 206) Comparator Systemic analgesia (N = 202) Premedication: temazepam, ranitidine, and metoclopramide Induction and maintenance: propofol, alfentanil, and pancuronium Surgery: CABG with CPB
Outcomes	Relevant to this review Risk of mortality Risk of myocardial infarction Risk of pulmonary complications (respiratory depression or pneumonia) Risk of neurological complications (cerebrovascular accident) Others Acute confusion Significant bleeding Renal failure Incidence of major organ complications
Notes	Correspondence: letter sent 16 March 2018; no reply Conflict of interest: none reported DOI: n/a
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "patients were randomized to one of two regimens...by using cards drawn from a sealed envelope"
Allocation concealment (selection bias)	Low risk	Sealed envelope
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Study was conducted in an open manner; therefore, neither the anaesthesiologists nor the nurses taking measurements were blinded to participants’ treatment
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Study was conducted in an open manner; therefore, neither the anaesthesiologists nor the nurses taking measurements were blinded to participants’ treatment
Incomplete outcome data (attrition bias)   All outcomes	Low risk	12 participants had insufficient data
Selective reporting (reporting bias)	Low risk	All results reported
Other bias	Low risk	In intention‐to‐treat for remaining 408 participants Groups had similar demographic data

Methods	Parallel: 2‐centre RCT Ethics committee: local human research ethics committees of the 2 participating centres (METC Isala Clinics, Zwolle, The Netherlands; and METC MST, Enschede, The Netherlands) approved of the study Informed consents: written informed consent was obtained from all participants Site: Utrecht, The Netherlands Setting: university hospitals Dates of data collection: from March 2004 to September 2007 Funding: the health renewal project provided from institutional sources: Isala Clinics Hospital 02/19. Registration number: 100000461 Registration: ISRCTN50434243
Participants	654 participants; mean age: 64.5 years; sex distribution: 111 females and 543 males Patients scheduled for elective cardiac surgery, including off‐pump procedures
Interventions	Intervention Epidural analgesia N = 325 Comparator Systemic analgesia N = 329 Induction: remifentanil, etomidate, and pancuronium Maintenace: propofol or sevoflurane and remifentanil Surgery: CABG with or without CPB
Outcomes	Relevant to this review Risk of mortality Risk of myocardial infarction Risk of pulmonary complications Risk of atrial fibrillation or atrial flutter Risk of neurological complications (cerebrovascular accident) Other Length of hospital stay
Notes	Correspondence: email sent 16 March 2018; no reply Conflict of interest: none reported DOI: 10.1097/ALN.0b013e318201d2de
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The random allocation sequence was concealed and computer‐generated in permuted unequal blocks, accessible through an Internet site
Allocation concealment (selection bias)	Low risk	The random allocation sequence was concealed
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Quote: "It was not possible for either the patient or the care providers to be blinded for treatment allocation"
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "All components of the primary endpoint were evaluated by an independent event committee blinded for randomization, consisting of a cardiologist, cardiothoracic surgeon, nephrologist, pulmonologist, and a neurologist"
Incomplete outcome data (attrition bias)   All outcomes	Low risk	One participant was excluded because his surgery was cancelled, and one participant withdrew his consent after randomization
Selective reporting (reporting bias)	Low risk	All results were reported
Other bias	Low risk	Intention‐to‐treat Groups had similar demographic data

Study	Reason for exclusion
Amat‐Santos 2012	Different study design: not an RCT: "depending on the preference of the anaesthesiologist responsible for the case"
Anderson 2005	Different study design: not an RCT: "the lack of randomization is a limitation"
Casalino 2006	Different study design: not an RCT: case series of 144 patients
Chae 1998	Different study design: classified as "no adequate sequence generation" by original review authors
Chakravarthy 2005	Different study design: prospective audit of cases conducted over a 13‐year period
Crescenzi 2009	Different study design: not an RCT: case‐matched, non‐randomized study
Djaiani 2000	No original data
El‐Morsy 2012a	Different study population: children
Jideus 2001	Different study design: classified as not randomized by previous review authors
Joachimsson 1989	Different study design: not an RCT: "two groups of consecutive patients meeting the inclusion criteria were investigated"
Kaunienė 2016	No outcome of interest measured
Kessler 2002	Different study design: not an RCT and different intervention: "use of TEA alone was applied in awake patients with multi‐vessel coronary artery disease who underwent CABG via median sternotomy"
Kessler 2005	Different study design: classified as "no adequate sequence generation" by previous review authors
Kunstyr 2008	Different study population: pulmonary endarterectomy with cardiopulmonary bypass
Kurtoglu 2009	Different intervention: compares general vs epidural anaesthesia for minimally invasive direct coronary artery bypass
Lagunilla 2006	Different intervention: "In the post‐operative period, 0.2% ropivacaine with 5 mg/ml fentanyl was used for analgesia in all patients, employing a patient controlled system"
Liang 2012	Different intervention: comparison between epidural anaesthesia perioperatively and postoperatively
Liem 1998	Different study design: not an RCT: case report
Martinez 2012	Different intervention: general anaesthesia compared with epidural anaesthesia or intrathecal morphine for beating heart surgery
Novikov 2011	Different study population: aorto‐femoral bypass
Olivier 2005	Different intervention: comparison of 3 different epidural solutions
Orsolya 2015	Different study population: robot‐assisted laparoscopic urogenital surgery
Ortega 2011	Different intervention: all participants had epidural analgesia with bupivacaine alone or bupivacaine plus morphine
Ovezov 2011	Different intervention: all participants had epidural analgesia
Rao 2016	Different intervention: all participants had epidural anaesthesia
Salman 2012	Different study design: not an RCT: "retrospective study"
Salvi 2004	Different study design: not an RCT: retrospective review of prospectively collected data
Schmidt 2005	Different intervention: all participants had epidural analgesia
Stenger 2013	Different study design: not an RCT: retrospective cohort study of prospectively registered data using population‐based healthcare databases
Stenseth 1993	Different intervention: all participants had epidural analgesia and were randomized to light or deep general anaesthesia
Thorelius 1996	Different study design: not an RCT: classified as "no adequate sequence generation" by previous review authors
Thorelius 1997	Different study design: not an RCT
Toda 2013	Different study design: not an RCT: "in this prospective non‐randomized study"
Turfrey 1997	Different study design: not an RCT: "Using computerised patient medical records, we analysed the frequency of respiratory, neurological, renal, gastrointestinal, haematological and cardiovascular complications in these two groups"
Yashiki 2005	Different intervention: TEA vs general anaesthesia

Trial name or title	Non‐analgesic benefits of combined thoracic epidural analgesia in elderly off‐pump coronary artery bypass grafting patients
Methods	Randomized by sealed opaque envelope; blinded participants
Participants	Participants undergoing off‐pump coronary artery bypass grafting Inclusion criteria: comorbidities Exclusion criteria: patient refusal, signs of infection over the spine, coagulation disorders, on antiplatelet agent, low‐molecular‐weight heparin or heparin infusion, emergency cases, unstable angina, left main stem disease, dysrhythmias, on steroids, undergoing combined procedures, on intra‐aortic, balloon pulsation, on rosuvastatin
Interventions	Intervention: epidural analgesia (5 to 15 mL of ropivacaine 0.75% followed by 6 to 14 mL as an infusion) Comparator: unspecified
Outcomes	1. Stress response 2. Hypercoagulability
Starting date	Registered: 27 April 2012 Last refreshed: 14 January 2019 Status: opened to recruitment
Contact information	Dr. Bhanu Prakash Medanta ‐ The Medicity, Sec‐38, Haryana, Gurgaon, 122001, Sonipat, Hatyana, India Email: doctorbhanu@yahoo.in
Notes	Found 6 February 2019 Funded by Industry (AstraZeneca Pharma India Ltd,, Avishkar, PB No. 2483, Bellary Road, Hebbal, Bangalore 560024)

Trial name or title	A study of central and mixed venous oxygen saturation with outcomes in open heart surgery patients between two groups conventional general anaesthesia and combined with perioperative thoracic epidural or intravenous analgesia
Methods	Randomized (computer generated; open list of random numbers; participant, investigator and outcome assessor blinded)
Participants	80 adults with coronary artery disease aged from 40 to 70 years Inclusion criteria: requiring off‐pump open heart bypass surgery Exclusion criteria: abnormal coagulation profiles, requiring salvage coronary artery bypass grafting, cardiogenic shock, heart valve pathology, antiplatelet therapy continuing local infection; renal, metabolic, neurological, or psychiatric disorders
Interventions	Intervention: epidural analgesia (10 mL bupivacaine 025%) Comparator: intravenous analgesia
Outcomes	1. Central venous and mixed venous oxygen saturation during intraoperative period and time for extubation and length of postoperative intensive care unit stay  2. Stress response, inotropic/vasodilatory support, pain outcome, and analgesia requirements  3. New arrhythmia, postoperative blood loss, perioperative myocardial infarction, neurological events, infective complication, if present  4. Any other adverse effects during the study period
Starting date	Started: 26 March 2013 Completed: 25 November 2013 Trial registered retrospectively: 5 May 2018 Unpublished results
Contact information	Dr. Chaitali Sen Professor and Head Institute of Post Graduate Medical Education & Research, Kolkata Department of Cardiac Anesthesiology IPGMER and SSKM Hospital, 242 A J C Bose Road Kolkata, West Bengal, 700020, India Email: chaitali03@rediffmail.com
Notes	Found on 6 February 2019