Landmark trials

Question 1

GALA (Lancet 2008)

” GA vs LA for carotid endarterectomy”

Answer 1

Parallel, multicentre, randomised controlled trial

P - Sx and aSx carotid artery stenosis

I - LA

I - GA

O - CVA, MI, death within 30days

Results:

• No difference between GA/LA in 30 day CVA, AMI or death rates
• Other trials found LA cheaper and lead to better early post-op neurocognitive function
• Choice should be determined by surgeon, anaesthetist and patient

Question 2

ARISE (NEJM, 2014)

Early-goal directed therapy vs usual care in early sepsis

Answer 2

Multicentre, RCT

P - ED patients with early sepsis

I - Early goal-directed therapy (EGDT)

• – Rx provided by study team trained in EGDT for 6hr period
• **- Arterial line and CVL for Scvo2 monitoring
• **– Rx as per EGDT algorithm*
• *C -** usual therapy
• • Rx, care location, monitoring, Ix determined by clinical treating team. Nil Scvo2 measurements allowed during trial period (6hrs)*

O -

• • 1o - all cause mortality at 90days
• • 2o & 3o - mortality in ICU/mortality at 28days/hospital mortality at 60days, specific cause mortality at 90days, length of stay in ED/ICU/hospital, need/duration mechanical ventilation, vasopressor support, renal-replacement therapy, destination at time of discharge

Results:

• No significant difference in all-cause mortality at 90days
• No significant difference in 28-day or in-hospital mortality, during of support or length of hospital stay

Question 3

B-Aware (Lancet, 2004)

Risk of awareness - 0.1 - 0.2% general surgical population; increased during cardiac surgery, LSCS, trauma surgery

Answer 3

Multicentre, prospective, double-blind RCT

• *P** - high risk of awareness (n = 2503)
• (Emergency LSCS, high-rish cardiac surgery [LVEF >30%, CI <2.1L/min/m2, severe AS, pulm HTN, pn/off pump CABG), acute trauma with hypovolvaemia, rigid bronchoscopy, significant impairment of CV status and expected intraoperative hypotension requiring Rx, severe end-stage lung disease, PHx awareness, anticipated difficult intubation when awake intubation not planned, known/suspected heavy EtOH intake, chronic benzo or opioid use, current protease inhibitor therapy*

I - BIS-guided anaesthesia (BIS 40 - 60)

C - routine care (whatever anaesthetist chose to do)

O -

–1o - incidence of confirmed awareness

– 2o - possible awareness, recovery times, hypnotic drug administration, incidence of marked hypotension, anxiety/depression, 30day mortlity

(pt recollection of intraoperative events, by use of structured questionnaire - at 2-6hrs, 24 - 36hrs, 30days post-op)

• *Results:**
• 2 awareness cases BIS, 11 usual care
• 43% pt received TIVA
• BIS significantly reduced incidence of awarness cf routine care (82% risk reduction with relaxant GA)
• Reduction in awareness - BIS 0.91% vs usual 0.17%, ARR 0.74% in BIS gp
• However, NNT is 138, with cost of $2200 to reduce 1 case of Awareness

Question 4

B-UNAWARE (Lancet, 2008)

Answer 4

Single centre RCT, prospective, blinded

P - high risk of awareness, to hhave GA with sevo/des/iso/N2O
(n = 2000)
(Major criteria: long-term anticonvulsant/benzo/opiate/cocaine use, EF >40%, h/o awareness, h/o difficult intubation or anticipated difficult intubation, ASA 4 or 5, AS, end-stage lung-disease, marginal ex tolerance not resulting from MSK dysfunction, pulm HTN, planned oper-heart surgery, daily etoh)

I - BIS-guided anaesthesia (target range 40 - 60)

C - ETAG-guide anaesthesia (end-tidal anaesthetic gas, range 0.7 - 1.3 MAC)

** Alarms for target range used in study to highlight if outside of range**

O - Awareness at 24hrs, 24 - 74hrs, 30days

• *Results:**
• 2 cases of definite awareness in each group
• No difference between groups
• ETAG 0.7 - 1.3 MAC as good as BIS 40 - 60
• Anaesethesia awareness occurred even when BIS and ETAG values were within target ranges
• Findings do not support routine use of BIS as part of standard practice
• May be useful in TIVA cases
• Overall incidence of anaesthesia awareness = 0.2% (95% CI 0.08 - 0.53)

Question 5

BART (NEJM, 2008)

“Blood conservation using antifibrinolytics in a randomised trial”
(Comparison of aprotinin and lysine analogues in high-risk cardiac surgery)

Answer 5

Multicentre, blinded RCT

P - high-risk surgical pts (n = 2331) (i.e. not just CABG pts
(Surgery requiring CPB - rpt cardiac surgery, isolated MV replacement, combined valve/CABG surgery, multiple valve replacement or repair, ascending aortic/arch surgery)

• *I -** Aprotinin vs TXA vs ACA (aminocaproic acid)
• (Aprotinin - protease inhibitor, TXA and ACA - lysine analogues)*

C - nil

• *O -**
• • 1o - massive postoperative bleeding
• –* *Massive post-op bleeding (composite bleeding >1.5L during any 8-hr period or MTP [>10units PRBC within 24hrs])
• • Rpt surgery due to haemorrhage
• • Cardiac tamponade wihthin 24hrs of protamine administration
• Death from haemorrhage within 30days*

– 2o - all-cause mortality at 30days

Results:

• Early termination of trial due to higher rate of death in aprotinin group (55% increased relative risk of death w/in 30days with aprotinin cf other groups)
• Trend towards increased renal impairment, AMI with aprotinin
• ↑ 30 day death in aprotinin 6% vs. TXA 4% vs. ACA 3.9%. ↑ RR death 1.53
• ↓ massive bleeding in aprotinin 9.5% vs. TXA 12.1%vs. ACA 12.1%
• Despite modest reduction in massive bleeding, aprotinin associated with higher mortality
• Observational studies showa protinin associated ↑ renal failure, CVS and cerebrovascular complications

Question 6

TRICC (NEJM, 1999)

“Transfusion requirements in critical care”

Answer 6

Multicentre RCT
To determine whether restrictive PRBC trasnfsion vs liberal PRBC transfusion produced equivalent results in critically ill patients

P - ICU patients

I - Transfusion threshold 70g/L (aiming for Hb 70 - 90g/L)

C - Transfusion threshold 100g/L (aiming for Hb 100 - 120g/L)

O - All-cause mortality at 30days

Results:

• Similar 30-day mortality between groups
• However, ​↓ mortality in restrictive group with pts:
  
  • Less acutely ill (APACHE score _<_20)
  • <55yo
• Significant ↓hospital mortality, cardiac complications, organ failure rates in restrictive group

Recommendations:

• Restrictive transfusion to maintain Hb 70-90g/L in critically ill, except possibly in patients with severe ischaemic syndromes/ACS

Question 7

SAFE (NEJM, 2004)

“Saline vs albumin fluid evaluation”

Answer 7

Multicentre, double-blind RCT

P - ICU patients (n = 6997)

I - 4% albumin

C - NaCl 0.9%

• *O -** All-cause mortality 28days
• Sub-group analysis = RR death in patients with trauma, RR death in severe sepsis

Results:

• No significant difference in all-cause mortality (P = 0.87)
• No significant different in length of ICU stay, hospital stay, duration of mechanical ventilation, days of renal-replacement therapy
• Sub-group analysis:
  
  • Death in trauma pt = RR 1.36 of death in albumin vs NaCl
    
    • Significance attributed to presence of head injury (mo difference in mortality b/w groups in trauma patients without brain injury)
  • ​Sepsis = RR 0.87 death in albumin vs NaCl

Outcomes:

• Albumin group received less fluids
• Sub-group analysis - possible worse outcomes with albumin in trauma with brain injuy; better outcomes in severe sepsis

Question 8

NABISH (NEJM 2001)

“National Acute Brain Injury Study Hypothermia”

Answer 8

Multicentre RCT

P - closed head injury

I - Hypothermia (33o) for 48hrs

C - Normothermia

O - Functional status at 6months

Results:

• No difference in functional status - 57% poor outcome, 28% mortality
• Hypothermia group had ↑hospital stay but ↓ ICP

Question 9

IHAST (NEJM, 2005)

“Mild intraoperative hypothermia duinr surgery for intracranial surgery” or “Intraoperative hypothermia for aneurysm surgery”

Answer 9

Multicentre, RCT

P - WFNS score 1 - 3 with SAH no more than 14days before planned surgical clipping (n = 1001)

I - Intraoperative hypothermia (33C, surface cooling techniques)

C - Intraoperative normothermia (36.5C)

O - Outcome at 90days, GCS assessed at 90days

Results:

• No significant difference between groups for pt undergoing craniotomy for aneurysmal SAH
  (no difference stay in ICU/hospital, death rates at followup = 6% in both groups, discharge destination)
• Increased post-op bacteraemia in hypothermic group (P=0.05)

Limitations:

• “Good patients” with WFNS Grade 1 - 3

WFNS Score:

• ​1 = GCS 15, no motor deficit
• 2 = GCS 13 - 14, no motor deficit
• 3 = GCS 13 - 14, motor deficit
• 4 = GCS 7 - 12, + motor deficit
• 5 = GCS 3 - 6, + motor deficit

Question 10

HACA (NEJM, 2002)

“Hypothermia after cardiac arrest”

Answer 10

Multicentre, RCT, partially blinded

P - Out of hospital VF cardiac arrest

I - Hypothermia (32 - 34C) for 24hrs

C - Normothermia for 24hrs

O - Neurological function at 6mths, complication rates at 7days

Results:

• Hypothermia group:
  
  • Improved neurological outcomes (55% vs 39%, NNT = 6)
  • Reduced mortality (41% vs 55%, NNT = 7)
• No difference in complications between groups
• Trend towards ↑ infection rates in hypothermia (non-significant)

Question 11

CRASH (Lancet, 2004)

“Corticosteroid randomisation after significant head injury”

Answer 11

Multilcentre, international placebo RCT

P - Head injury pt with GCS _<_14 within 8hours of injury (n = 10,008)

I - IV corticosteroids (48hr methylprednisolone infusion)

C - Placebo

• *O -
• **Death within 2 weeks of injury
• Death or disability at 6 months

Results:

• Trial ceased prematurely due to results
• Higher all-cause mortality at 2-weeks in corticosteroid group [21.1% vs 17.9%, 95% CI 1.09 - 1.27, P = 0.0001]
• Cause of increased risk of death unclear
• *Background**:
• Corticosteroids used for ~30yrs in Rx head injury
• Systemic r/v 1997 - suggested 1 - 2% reduced risk of death with corticosteroids

Question 12

CRASH-2 (2010)

“Clinical randomisation of an antifibrinolytic in severe haemorrhage”

Answer 12

B/G:
Ax of early administration of short course of TXA on death, vascular occlusive dz and need for PRBC in trauma pt

International, multicentre RCT, placebo

P - trauma pts with or without significant risk of bleeding within 8hrs of injury (n = 20,211)

I - TXA (LD 1g over 10mins then infusion 1g over 8hrs)

C - Placebo

• *O -** All-cause mortality within 4 weeks of injury
• Sub-group analysis (ITT)
• Bleeding
• Vascular occlusion (MI, CVA, PE)
• Multi-organ failure
• Head injury
• Other

Results:

• Significantly reduced all-cause mortality in TXA [14.5% vs 16%, RR 0.91, 95% CI 0.85 - 0.97, p = 0.035]
• Risk of death due to bleeding significantly reduced [4.9% vs 5.7%, p = 0.0077]
• Same risk VTE with TXA
• Lower risk AMI with TXA

Other:

• Better quality study than MATTERS
• Pt with or at significant risk of bleeding - only 50% received PRBC
• Strong evidence TXA reduces bleeding in multiple different forms of surgery, but some uncertainty remains re: overall mortality + VTE, esp post aprotinin
  *

Question 13

POISE (Lancet, 2008)

“Perioperative ischaemic evaluation trial”

Answer 13

International, multicentre, RCT, placebo, ITT

P - Pt at with or at risk of atherosclerotic disease undergoing non-cardiac surgery (n = 8351)

• *I -** Metoprolol XR
• (Drug started 2-4hrs pre-op, continued for 30days)*

C - Placebo

O - Composite of cardiac death, non-fatal MI, non-fatal cardiac arrest

Results:

• Metoprolol group:
  
  • ​Significant reduction in MI (4.2% vs 57%, p = 0.0017)
  • Significant increase in deaths (3.1% vs 2.3%, p = 0.0317)
  • Significant increase in CVA (1% vs0.5%, p = 0.0053)
  • Significantly more bradycardia, hypoTN, sepsis

Question 14

POISE-2 (Lancet, 2013)

“Perioperative ischaemic evaulation”

Answer 14

Blinded, randomised 2x2 factorial design
Allowed separate evaluation of clonidine vs placebo, aspirin vs placebo

P - Pts with or at risk of atherosclerotic disease undergoing non-cardiac surgery (n = 10,010)

• *I -** Drugs given just before surgery and cotninued until 72hrs post-op
• • Clonidine
• • Aspirin

C - Placebo

O - Composite of death or non-fatal MI at 30days

Results:

• Clonidine:
  
  • Nil significant reduction in composite outcome of CV death or non-fatal MI
  • Non-significant reduction in MI​
  • Significantly more clinically important hypoTN [47.6% vs 37.1%, P <0.001)
• Aspirin
  
  • No significant effect on death/non-fatal bleeding
  • Increased risk of major bleeding (4.6% vs 3.8%, p = 0.04)

Question 15

ENIGMA (Anaesthesiology, 2007)

“Evaluation of nitrous oxide in gas mixture for anaesthesia”

Answer 15

Multicentre RCT, partially blinded

P - Pts having non-cardiac surgery for >2hrs (n=2050)

I - N2O (70:30 air)

C - N2O free (80% O2:20% N2)

• *O** -
• 1o - duration hospital stay
• 2o - time in ICU, complications, death at 30days

Results:

• Increased long-term risk of myocardial infarction with N2O, but not increase in death or CVA
• N2O - increased wound infection, PONV, atelectasis, fever, pneumonia
• FiO2 different between groups (30% vs 80%)

Issues:

_QUESTION:_ are the results due to N2O or reduced O2? To be answered in ENIGMA-II

Recruitment of relatively unselected pts with low 30-day and long-term event rates –> may have been underpowered to confirm a “true” increased risk of MI in pts receiving N2O

Question 16

ENIGMA-II (Anaesthesiology 2015)

“Evaluation of N2O in gas mixture in anaesthesia”

Answer 16

International, multicentre, parallel-group, patient and observer-blinded, randomised trial

P - Non-cardiac patients (>45yo) at risk of perioperative CV events
(n= 7112)

I - N2O 70% (30% O2)

C - N2 70% (30% O2)

• O -** Composite outcome of death and major CV events at 1yr after surgery (non-fatal MI, cardiac arrest, PE, CVA)*
• 2o - disability or death, fatal or non-fatal MI, non-fatal CVA

Results:

• N2O does not increase the risk of composite outcome of death and major CV events (disability or death, MI, CVA)
• Supports long-term safety of N2O administration in non-cardiac sugical patients with known or suspected CV disease

Question 17

MASTER (Lancet, 2002)

“Multicentre Australian Trial of Epidural Anaesthesia”

Answer 17

Comparison of adverse outcomes in high-risk patients managed for major surgery with epidural block or alternative analgesic regimens with GA

Multicentre RCT, ITT

P - High risk pts undergoing abdominal surgery (n = 915)
(Morbid obesity, DM, CRF, respiratory insufficiency, major hepatocellular dz, AMI, myocardial ischaemia, >75yo + 2 other criteria [see study])

I - Intraoperative epidural + post-operative epidural analgesia for 72hrs

C - Control

O - Death at 30days, major post-surgical morbidity

Results:

• No significant difference in mortality at 30 days (epidural 5.1% vs control 4.3%, p = 0.67)
• ↓ risk of post-op resp failure (23% vs 30%, p = 0.02)
• Improved post-op analgesia with EDB
• No major adverse events related to insertion

“Most adverse morbid outcomes in high risk patients undergoing major abdominal surgery are not reduced by use of combined epidural and GA and post-op epidural analgesia. However, the improvement in analgesia, reduction in respiratory failure, and the low risk of srious adverse consequences suggest that many high-risk patienst underoing major intra-abdominal surgery and epidural anaesthesia intraoperatively with continuing post-operative epidural analgesia”

Question 18

IMPACT (NEJM, 2004)

“International multicentre protocol to assess single and combined benefits of anti-emetic interventions in clinical control trial”

Answer 18

Multicentre RCT, double-blinded
Factorial design

P - high baseline risk of PONV (>40% risk) (n = 5199)

• *I -** randomised to 1 of 64 possible combinations of 6 prophylactic interventions:
• Ondansetron 4mg or no ondansetron
• Dexamethasone 4mg or no dexamethasone
• Droperidol 1.25mg or no droperidol
• Propofol or volatile anaesthetic
• Nitrogen or N2O
• Fentanyl or remifentanil

O - PONV at 24hrs

Results:

• Ondansetron, dexamethasone and droperidol ↓ PONV 26%
• Propofol ↓ 19%
• Nitrogen ↓12%
• Remifentanil vs. fentanyl no difference
• Best risk reduction = TIVA propofol + triple anti-emetic = 70% risk reduction

Question 19

PACMAN (Lancet, 2005)

“PAC in the management of intensive care patients”

Answer 19

Multicentre RCT

P - ICU patients (n = 1040)

I - PAC

C - No PAC (option to use other CO monitoring device)

O - Mortality, complications

Results:

• No difference in hospital mortality between groups
  
  • 46/486 complication rate with PACs, none fatal
• No clear evidence of benefit or harm in using PAC
• Require further efficacy studies

Question 20

MAGPIE (Lancet, 2002)

“Magnesium for prevention of eclampsia”

Answer 20

International, multicentre RCT, semi-blinded, ITT

P - Parturients with BP >140/90 + proteinuria >+1 (_>_30ml/dL), not yet delivered ot <24hrs post-partum
(n = 10,414)

I - Magnesium

C - Placebo

O - Eclampsia, fetal death (in women randomised before delivery)

Results:

• 58% reduced risk of eclampsia in MgSO4 group (0.8% vs 1.9%, 95% CI 40 - 70)
• Non-significant reductionin maternal mortality in MgSO4 (RR 0.55)
• No clear difference in risk of fetal death pre-partum
• 24% side effects MgSO4 cf placebo (5%)
• 5% increase risk of LSCS
• “MgSO4 halves the risk of eclampsia, probably reduces the risk of maternal death. There do not appear to be substantive harmful effects to mother or baby in short term”
• No tocolytic or hypotensive effect

Question 21

COMET (Lancet, 2001)

“Comparative obstetric mobile epidural trial”

Answer 21

Double centre RCT

P - Women requesting labour analgesia (n = 1050, primips)

I - Epidural bolus

I - low dose CSE

I - Low dose infusion epidural

O -

1o - Mode of delivery
2o - Labour progress, efficacy of procedure, neonatal complications

Results:

• NVD rates higher in low-dose infusion group
  (42. 9% vs 42.7% CSE vs 35.1% traditional bolus; p = 0.04)
  
  • Differences due to reduced intrumental vaginal delivery
• APGARS _<_7 and neonatal resuscitation higher in infusion epidurals

** Low-dose epidural infusions benefits delivery outcome

Question 22

NICE-SUGAR (NEJM, 2009)

“Normoglycaemia in intensive care evaluation survival using glucose algorithm regulation”

Answer 22

Multicentre RCT

P - Pt expected to require Rx in ICU for _>_3 days (n = 6104)

I - Intensive glucose control (target 4.5 - 6mmol/L)

C - Conventional glucose control (_<_10mmol/l)

O - All-cause mortlity within 90days

Results:

• ↑ death in tight control (OR 1.14, 95% CI 1.02 - 1.28; p = 0.02)
  Similar for operative and non-operative pts
• ↑hypoglycaemia in tight control (6.8% vs 0.5%, p < 0.001)
• No significant difference in length of ICU/hospital stay, length of mechanical ventilation, renal-replacement therapy

Question 23

TTM (NEJM, 2016)

“Targeted temperature management at 33oC vs 36oC after cardiac arrest”

Answer 23

International, multicentre RCT

P - Unconscious, OOHCA (n = 950)

I - TTM 33oC

C - TTM 36oC

O -

1o - all-cause mortality to end of trial
2o - composite poor neurological function or death at 180days
(Cerebral Performance Category [CPC] Scale and modified Rankin Scale)

Results:

• No significant difference in all cause mortality (50% 33C vs 48% 36C, hazard ratio 33oC 1.06, 95% CI 0.89 - 1.28, p = 0.51)
• No significant different in neurological outcome/death at 180days (54% 33C vs 52% 36C, risk ratio 1.02, 95% CI 0.88 - 1.16, p = 0.78)

Question 24

ARDSNet (NEJM, 2000)

“Ventilation with lower tidal volume as comparted with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome”

Answer 24

Traditional tidal volume 10 - 15ml/kg

Multicentre, RCT

P - Pt with acute lung injury and ARDS (n = 861)

I - VT 6ml/kg predicted body weight, airway pressure <30cmH2O measured after 0.5sec end-expiratory pause

C - VT 12ml/kg predicted body weight, airway pressure <50cmH2O measured after 0.5sec end-expiratory pause

O -
1o - death prior d/c home and death prior to breating without assistance
2o - No. days without ventilator use from days 1 - 28

Results:

• Trial prematurely ceased due to significant reductionin mortality in low VT group (31% vs 39.8%, p = 0.007)
• Reduced no. of ventilator days in low VT group (p = 0.007)

Question 25

CHEST (NEJM, 2012)

“Crystalloid vs hydroxyethyl starch trial”

“Hydroxyethyl starch or saline for fluid resuscitation in intensive care”

Answer 25

Multicentre, RCT
Ix effect of HES, esp in context of previously reported AEs on renal function

P - Critically ill pts admitted to ICU (n = 7000)

I - Voluven (6% HES; HES 130/0.4 ratio)

C - NaCl 0.9%

Fluids given for all fluid resuscitaU d/c, death or 90days after randomisation

O -

1o - Death at 90days
2o - AKI, renal failure, Rx with renal-replacement therapy

Results:

• No significant difference in mortality (HES 18% vs NaCl 17%, RR HES 1.06, 95% CI 0.96 - 1.18, p = 0.26)
• Significantly increased renal-replacement need in HES group (7% vs 5.8%, RR 1.21, 95% CI 1 - 1.45, p = 0.04)
• HES associated with significantly more adverse events (5.3% vs 2.8%, p < 0.001)

HES in critically ill patients was associated with a significantly higher incidence of AKI and requirement for RRT

Question 26

FOCUS (NEJM, 2011)

“Functional outcomes in cardiovascular patients undergoing surgery for hip fracture”

Answer 26

Ix threshold for PRBC transfusion in hip # surgery

P - >50yo with h/o or RF for CV disease with Hb >100g/L after hip fracture surgery (n = 2016)

I - Restrictive transfusion strategy (Sx anaemia, Hb >80g/L)

C -Liberal transfusion strategy (Hb transfused to 100g/L)

O - Death or inability to walk across a room without human assistance at 60-days

Results:

• There were no significant differences in mortality, functional performance or any outcome
• Liberal transfusion (trigger Hb<100) is no better than restrictive transfusion (trigger anemia symptoms or Hb < 80)

Question 27

MATTERS (Archives of Surgery, 2012)

“Military application of TXA in Trauma Emergency Resuscitation Study”

Answer 27

Retrospective study looking at a military hospital in Afghanistan

P - Trauma pts requiring transfusion (n = 896)

I - TXA bolus 1g (rpted at clinical judgement by clinician)

O - Mortality, VTE,

Results:

• TXA significantly reduces mortality in trauma (17% v 23%), esp. where massive transfusion is required
  
  • TXA group had significantly lower mortality, despite having more severe injuries
  • TXA was associated with an OR 7 for survival
• Associated with increased incidence of PEs and DVTs though (overall roughly from ~0.5% to 2.5%) ?due to pts being sicker though
• *

Question 28

DECRA (2011)

“Decompressive craniectomy trial”

Answer 28

Ix whether decompressive craniectomy improves the functional outcome in patients with severe traumatic brain injury and refractory raised intracranial pressure

P - Severe diffuse TBI and intracranial hypertension refractory to 1st tier therapies (n = 155) [December 2002 - April 2010] [able to be recruited up to 72hrs post admission to hospital]

I - Bifrontotemperoparietal decompressive craninectomy

• *C -** Standard care
• Based on Brain Trauma Foundation guidelines:
• 1st tier Rx - sedation, normal PaCO2, mannitol, hypertonic saline, NMBDs, EVD*

-2nd tier Rx - mild hypothermia 35oC), barbiturates

O - Extended Glasgow Outcome Scale score at 6months

(Original primary outcome = unfavourable outcome (composite of death, vegetative state, severe disabilityI at 6months post injury)

Results:

• Decompressive craniectomy:
  
  • ↓ICP (p<0.02)
  • ↓length of ICU stay (p<0.001)
  • ↓ duration mechanical ventilation
  • BUT significantly higher chance of being dead/vegetative/severely disabled (odd ratio 2.21, 95% CI 1.14 - 4.26, p = 0.02)
• Rates of death at 6months similar in both groups (19% craniectomy vs 18% usual care)

Question 29

EVAR (NEJM, 2010)

“Endovascular vs open repair of AAA”

Answer 29

Multicentre, RCT

P - Large AAA >5.5cm diameter) (n = 1252) [years 1999 - 2004]

I - EVAR

C - Open AAA repair

O - Rates of death, graft-related complications, re-nterventions

Results:

• Significant reduction in 30-day operative mortality in EVAR group (1.8% vs 4.3%, EVAR odds ratio 0.39, 05% CI 0.18 - 0.87, p= 0.02)
• EVAR group had early benefit re: aneurysm-related mortality, but benefit lost by end of the study (partially due to fatal endograft ruptures)
• No significant difference in rates of death from any cause by the end of the study (p = 0.72)
• Rates of graft-related complications and reinterventions were higher with EVAR and new complications occurred up to 8yrs post randomisation >> contributing to higher overall costs

Question 30

ISAT (Lancet, 2005)

“International subarachnoid aneurysm trial”

Answer 30

Multicentre RCT

Endovascular coil vs craniotomy + clipping for subarachnoid aneurysm

P - Ruptured intrcranial aneurysms across 42 neurosurgical centres (n = 2143)

I - Endovascular coiling

C - Neurosurgical clipping

** Note - not all aneurysms are amendable to coiling e.g. wide neck

O -

1o - Death or dependent at 1yr (modified Rankin scale 3 - 6)

2o - rebleeding from Rx aneurysm, risk of seizures

Results:

• In pts with ruptured intracranial aneurysms amenable to both clipping and coiling, coiling leads to a greater chance of survival free of disability at 1 year (23.5% vs 30.9%, p = 0.0001)
• Survival benefit with coiling continues for at least 7 years
• Risk of epilepsy lower in coiling
• BUT Rebleeding rare, but more common with coiling

Question 31

WOMAN (Lancet, 2017)

“World maternal antifibrinolytic trial”

Answer 31

International, multicentre
Randomised, double-blind, placebo controlled trial

P - PPH post NVD or LSCS (n = 20,000)

I - TXA 1g IV (2nd dose 1g TXA given if bleeding continued after 30mins or re-bleed within 24hrs)

C - Placebo + usual care

O - Risk of death from PPH

Results:

• Significantly reduced risk of death from PPH in TXA (1.5% vs 1.9%, RR 0.81, 95% CI 0.65 - 1, p = 0.045), especially in women Rx within 3hrs of delivery (1.2% vs 1.7%, RR 0.69, 95% CI 0.52 - 0.91, p = 0.008)
• Hysterectomy not reduced in TXA group (RR 1.02, 95% CI 0.8 - 1.07, p = 0.84)
• The composite primary endpoint of death from all causes or hysterectomy was not reduced with TXA ( 5·3% vs5·5%, RR 0·97, 95% CI 0·87-1·09; p=0·65).
• Adverse events (including thromboembolic events) did not differ significantly in the tranexamic acid versus placebo group.

Interpretation:

• Tranexamic acid reduces death due to bleeding in women with post-partum haemorrhage with no adverse effects.
• When used as a treatment for postpartum haemorrhage, tranexamic acid should be given as soon as possible after bleeding onset.

Question 32

RELIEF (NEJM, 2018)

“Restrictive versus Liberal Fluid Therapy for Major Abdominal Surgery”

Answer 32

International, multicentre, RCT

P - Pt undergoing major abdominal surgery and at increased risk of complications (n = 3000)

I - Restrictive IV regimen (designed for net zero fluid balance. 5ml/kg induction and intraop, no pre-op fluid. 0.8ml/kg post-op infusion)

C - Liberal IV regimen (crystalloid 10ml/kg during induction then 8ml/kg to end of surgery; post-op 1.5ml/kg infusion for at least 24hrs)

** Both arms of IV therapy for 24hrs post surgery

O -
1o - Disability-free survival at 1yr
2o - AKI at 30days, renal-replacement therapy at 90days, composite of septic complications, surgical-site infection and death

Results:

• No significant difference in 1yr disability-free survival (81.9% restrictive vs 82.3% liberal, p = 0.61)
• Significantly increased rate of AKI in restrictive fluid group (8% vs 5%, p <0.001)
• Significantly increased rate of surgical site infection (16.5% vs 13.6%, p = 0.02) and renal-replacement therapy (0.9% vs 0.3%, p = 0.048) in restrictive fluid group
• Non-significant increase in septic complications/death in restrictive group (21.8% vs 19.8%, p = 0.19)

CONCLUSIONS

“Among patients at increased risk for complications during major abdominal surgery, a r_estrictive fluid regimen was not associated with a higher rate of disability-free survival_ than a liberal fluid regimen and was associated with a _higher rate of acute kidney injury._

Question 33

DREAMS (BMJ, 2017)

“Dexamethasone versus standard treatment for postoperative nausea and vomiting in gastrointestinal surgery: randomised controlled trial”

Answer 33

Multicentre, 2-arm parallel group, randomised trial, blinded post-op care

P - Elective open or laparoscopic bowel surgery for benign or malignant pathology (n = 1350)

I - Dexamethasone 8mg at induction

C - Standard care (no antiemetics)

O -

1o - PONV (within 24hrs)
2o - Vomiting within 72 and 120hrs, use of antiemetics and PONV at 24, 72, 120hrs rated by pt, fatigue and QoL at 12hrs or discharge and at 30days, time to return to oral intake, length of hospital stay, adverse events

Results:

• Significantly less vomiting within 24hrs in dex group (25.5% vs 33%; NNT 13, 95% CI 5 - 22, p = 0.003)
• Significantly less rescue anti-emetics in dex group (39.3% vs 51.9%, NNT 8, 95% CI 5 - 11, p < 0.001); reduction in antiemetics remained up to 72hrs
• No increase in complications with dex use

Question 34

ATACAS-Aspirin (NEJM, 2016)

Answer 34

Multicentre, double blind, 2x2 factorial trial

P – pt undergoing CABG at risk of major complications (age >70yo, LV impairment, concomitant valvular or aortic surgery, repeat surgery and medical comorbidities – COPD/CKD/obesity/PVD/pulm HTN; eligible if not taking aspirin or if aspirin ceased for at least 4days pre-op)

I – aspirin given prior to surgery (100mg)

C – Placebo

O –

1o - composite of death or thrombotic events (non-fatal MI, CVA, PE, renal failure, bowel infarction) within 30-days post-op
2o - blood transfusion, re-operation, respiratory failure, length of stay

Results:

• No clinical or statistical differences found (19.3% vs 20.4% asp vs placebo), RR 0.94 (95% CI .08 – 1.12, P = 0.55), ARR 1.12% (95% CI 2.29 – 4.54%, P = 0.55)
• 2o outcomes: No statistically significant difference in any of the individual outcomes of composite 1ooutcome
• More plt transfusion on day 1 in aspirin group (15 vs 13%, P = 0.23)
• More PRBC transfused from day 2 to discharge in aspirin group (18.1% vs 16.1%, P = 0.08)
• No difference in ICU stay, hospital stay, duration of mechanical ventilation

Question 35

ATACAS-TXA (NEJM, 2016)

Answer 35

Multicentre, double blind, 2x2 factorial trial

P - High risk patients undering CABG (n = 4631)

I - TXA (50 - 100mg/kg)

C - Placebo

O -
1o - composite of death and thrombotic complications (non-fatal MI, CVA, PE, renal failure, bowel infarction) within 30days
2o - blood transfusion, re-operation, respiratory failure, length of stay

Results:

• No significant difference in composite of death/thrombotic complications (18.1% TXA vs 16.7% placebo, p = 0.22)
• Significantly reduced:
  
  • ​Major haemorrhage and cardiac tamponade requiring re-operation in TXA group (1.4% vs 2.9%, p =0.001)
  • PRBC within 24hrs in TXA group (31% vs 49%, p <0.001)
• Significantly increased perioperative seizures in TXA group (0.7% TXA vs 0.1%, p = 0.002)