Myburgh on Fluids ICN NSW 2015

Fluid management: primum non nocere

UNSW

John Myburgh

Sydney

Rudyard Kipling 1865-1936

I keep six honest serving-men:�(They taught me all I knew)�Their names are What and Where and When�And How and Why and Who

Leith Infirmary 1831

Thomas Aitchinson Latta c1790-1833

“The most wonderful and satisfactory effect is the immediate consequence of the injection.” “The solution that was used consisted of two drachms of muriate, and two scruples of carbonate of soda to sixty ounces of water. It was at the temperature of 108 or 110o”

“The quantity necessary to be injected will probably be found to depend upon the quantity of serum lost..”

Lewins: London Medical Gazette 1832

“Verily sir, this is an astonishing method of medication, and I predict will lead to wonderful changes and improvements in the practice of medicine ”

Lewins: London Medical Gazette 1832

Sydney Ringer 1834-1910

Alexis Hartmann 1898-1964

Ernest Starling 1866-1927

Thomas Graham 1805-1869

Crystalloids “substances such as salt, sugar and urea that could be crystallised with ease” Colloids (from Κθλλη, glue) “these included substances such as gelatin or glue, gum, egg-albumin, starch and dextrin” Colloid properties “non- crystallisable, form gummy masses when evaporated to dryness, diffuse with extreme slowness and would not pass through animal membranes”

Edwin Cohn 1892-1953

“Das Blut ist ein ganz besonder Saft” Faust (Goethe)

Arthur Guyton 1919-2003

Who?

Mean arterial pressure

Cardiac output Right atrial pressure

Mean systemic pressure

Perfusion pressure

Unstressed volume

Stressed volume

Venous return

Arteriolar tone

15mmHg 15mmHg

5mmHg

60mmHg

60mmHg

Guyton 1955

Vincent: New Engl J Med 2013

0-24h 24-72h 72-96h >96h

“I don’t care if you use dog’s piss, as long as you

use it carefully.”

Malcolm Fisher AO

What?

Roberts: BMJ 1998

RRD 1.68 (1.25 – 2.23)

Overall excess mortality of 6%

(95% C.I. 3 - 9%)

24/30 studies n=1104/1419

Favours albumin

Favours control

Hypovolaemia

Hypoalbuminaemia

Burns

TOTAL

SAFE Study Investigators: NEJM 2004

2001-2003 Multicentred blinded RCT

Albumin vs saline ICU patients

n=6997

Primary outcome: Mortality at 28d

Fluid volumes

Ratio of albumin to saline for first four days = 1:1.4

1 2 3 40

500

1000

1500

2000 AlbuminSaline

p<0.001

p<0.001

p=0.026

Day

Volu

me

adm

inis

tere

d (m

L)


Should you change practice?


Mortality at 28 days Mortality at 2 years

Albumin and intracranial pressure

Cooper: J Neurotrauma 2013

P=0.059 (Test for common relative risk)

Sepsis

SAFE Study Investigators: Int Care Med 2011

MVLR adjusting for baseline covariates in patients with complete data: 919/1218 (75.5%)

0.71 (0.52 – 0.97) p=0.03.

Caironi: New Eng J Med 2014

Multicentred open-label RCT 20% albumin (>30g/L) vs crystalloid: severe sepsis

n=1818 Primary outcome: Mortality at 28d

2008-2012

What about synthetic colloids?

Choice of Colloid: Severe sepsis

0

50

100

150

200

250

300

350

400

450

OCEANIA AMERICAS ASIA NORTHERNEUROPE

SOUTHERNEUROPE

WESTERNEUROPE

All

mL

per p

erso

n

Albumin Starch Gelatin Dextran

Choice of Colloid: Severe sepsis

SAFE TRIPS Investigators: Crit Care 2010

“Hydrops lysosomalis generalisatus”

Renal Dickenmann: AJKD 2005

Hepatic Schmidt-Hieber: Eur J Haem 2006

Skin Sirtl: BJA 1999

Tissue accumulation and HES

Renal replacement therapy: 31.0 v 18.8% p=0.001

Brunkhorst: New Engl J Med 2008

P=0.48 P=0.09

Multicentred 2x2; open label RCT 10% HES 200/0.5 vs in Ringer’s lactate

n=537/600 (adaptive); severe sepsis Primary outcome: Mortality at 28 days

2003-2005

“New” generation HES

Concentration 6% Molecular weight ~ 130 kD Molar substitution ratio ~ 0.4 Carrier 0.9% saline

Ringer’s acetate

Perner: New Engl J Med 2012

Multicentred blind RCT 6% HES 130/0.42 in Ringer’s acetate vs Ringer’s acetate

n=798; severe sepsis Primary outcome: Mortality or RRT at 90d

2009-2011

Myburgh: New Engl J Med 2012

Multicentred blind RCT 6% HES 130/0.4 in saline vs 0.9% saline

n=7000; ICU patients Primary outcome: Mortality at 90d

2009-2012

6S 2012

P=0.09

SepNet (VISEP) 2008

P=0.07

70mL/kg (33 to 114.2)

44Lm/kg (24 to 75)

17mL/kg (9 to 31)

Regulatory responses 14 June 2013 (Revised 11 October 2013). Restriction of HES in high-risk patients 24 June 2013. “Boxed” warning against use of HES in high-risk patients 27 June 2013. Withdrawal of registration of HES and recall of unused stock 8 April 2014 Restriction of HES in high-risk patients.

EU EMA’s PRAC recommend that marketing authorisations for HES products be suspended

UK MHRA suspends use of HES infusions, recommends crystalloids for fluid resuscitation; supports with position statement by Faculty of Intensive Care Medicine, Intensive Care Society, and Royal College of Anesthetists

USA FDA issues Safety Letter recommending boxed warning for HES solutions on increased mortality, severe renal injury, and risk of bleeding

Germany BfArM recommends to stop using HES products

Italy HES products suspended and recalled

Ireland Irish Board of Medicine recommends stop use and distribution of HES

Poland Polish competent Authorities decide to stop use and distribution of HES immediately

Switzerland, France, Spain, Czech Republic Recommend not to use HES in specific indications

Canada Health Canada issues advisory with contraindications and warnings for the use of HES in patients with sepsis, renal impairment or severe liver disease.

Australia Australian TGA initiated risk/benefit review, added contraindications for sepsis and liver disease, strengthened warning on risk of severe renal impairment and bleeding disorders

Regulatory responses

The Empire strikes back ……….

Brit J Anaes / Acta Anes Scan: 2013

D Angus USA M Antonelli ITA A Artigas ESP M Bauer GER R Bellomo AUS G Bernard USA J Bion UK L Brochard FRA C Brun BuissonFRA F Brunkhorst GER V BumbasirevicSER H Burchardi GER P Caironi ITA J Carlet FRA J Chalmers AUS J Chastre FRA G Citerio ITA D Cook CAN J Cooper AUS P Dellinger USA T Evans UK S Finfer AUS H Flaaten NOR

R Freebairn NZ C French AUS D Gattas AUS L Gattinoni ITA H Gerlach GER E G-BourboullisGRE C Hartog GER C Hinds UK U Kaisers GER M Levy USA J Lipman AUS S MacMahon AUS D McAuley UK S McGuiness NZ L McIntyre CAN M Maggorini SUI J Mancebo ESP J Marshall CAN R Moreno POR J Morgan AUS J Myburgh AUS C Natanson USA R Norton AUS

D Payen FRA A Perner DEN V Perkovic AUS A Pesenti ITA V Pettilla FIN C Putensen GER M Quintel GER M Ranieri ITA K Reinhardt GER A Rhodes UK C Richard FRA N Riedermann GER I Roberts UK G Rubenfeld CAN F Schortgen FRA G Sigurdsson ICE C Sprung ISR N Stochetti ITA P Suter SUI J Takala SUI T Thompson USA A Turner AUS T Walsh UK

S Webb AUS N Webster UK T Welte GER M White AUS C WiedermannGER D Young UK R Zarychanski CAN

Brit J Anaes: On line 12 December 2013

Is the PRAC decision in the best interest of patients? Increased relative risk of death ~ 6%

Increased relative risk of RRT ~ 27% Could there be a place for HES in the future?

Only through robust, unbiased clinical trial network Is there new evidence of safety for HES?

CRISTAL RAFTING BaSES

Bion: Int Care Med 2013

Annane: JAMA 2013

Multicentred open-label RCT Colloids vs crystalloids

Hypotensive, hypovolaemic patients No pre-randomisation fluids

n=2857/3010 Primary outcome: Mortality at 28d

2003-2012

Study or Subgroup

1.1.1 Low Risk of Bias

Yates 2014

Perner 2012

Myburgh 2012Subtotal (95% CI)

Total events

Heterogeneity: Tau² = 0.00; Chi² = 2.07, df = 2 (P = 0.35); I² = 4%

Test for overall effect: Z = 2.12 (P = 0.03)

1.1.2 Intemediate Risk of Bias

Alavi 2012

Skhirtladze 2014

Nagpal 2012

Feldheiser 2013

Hamaji 2013

James 2011

Gondos 2010

Siegemund 2012

Guidet 2012Subtotal (95% CI)

Total events



1.1.3 High Risk of Bias

Du 2011

Dubin 2010

Yang 2011

Hung 2012

Lu 2012

Zhao 2013

Zhu 2011Subtotal (95% CI)

Total events

Heterogeneity: Not applicable

Test for overall effect: Not applicable

Total (95% CI)

Total events



Test for subgroup differences: Chi² = 0.13, df = 1 (P = 0.72), I² = 0%

Events

5

201

597

803

0

1

1

1

1

12

15

33

40

104

2

1

0

0

7

5

2

0

907

Total

104

398

33153817

32

81

35

26

24

56

50

117

100521

21

9

26

41

22

80

450

4338

Events

2

172

566

740

0

0

0

0

0

6

14

36

32

88

2

5

0

0

12

5

4

0

828

Total

98

400

33363834

28

79

35

28

24

53

50

124

96517

22

11

25

39

20

40

450

4351

Weight

0.2%

29.1%

59.2%88.5%

0.1%

0.1%

0.1%

0.1%

0.8%

1.7%

4.1%

4.7%11.5%

100.0%

M-H, Random, 95% CI

2.36 [0.47, 11.86]

1.17 [1.01, 1.36]

1.06 [0.96, 1.18]1.10 [1.01, 1.20]

Not estimable

2.93 [0.12, 70.79]

3.00 [0.13, 71.22]

3.22 [0.14, 75.75]

3.00 [0.13, 70.16]

1.89 [0.77, 4.68]

1.07 [0.58, 1.98]

0.97 [0.65, 1.45]

1.20 [0.83, 1.74]1.15 [0.91, 1.46]

Not estimable

Not estimable

Not estimable

Not estimable

Not estimable

Not estimable

Not estimableNot estimable

1.11 [1.02, 1.20]

HES 130/0.38-0.45 Crystalloid Risk Ratio Risk Ratio

M-H, Random, 95% CI

0.01 0.1 1 10 100Favours HES130/0.38-0.45 Favours crystalloid

HES and mortality

Increased mortality: RR 1.11 (1.02-1.20)

NNH 53

French: unpublished (with permission)

HES and RRT Study or Subgroup

1.2.1 Low RIsk of Bias

Yates 2014

Perner 2012

Myburgh 2012Subtotal (95% CI)

Total events



1.2.2 Intermediate Risk of Bias

Skhirtladze 2014

Lee 2011

Nagpal 2012

James 2011

Guidet 2012

Siegemund 2012Subtotal (95% CI)

Total events



Total (95% CI)

Total events



Test for subgroup differences: Chi² = 0.45, df = 1 (P = 0.50), I² = 0%

Events

4

87

235

326

1

1

1

2

21

28

54

380

Total

104

398

33523854

81

53

35

56

100

117442

4296

Events

0

65

196

261

0

0

0

3

11

23

37

298

Total

98

400

33753873

79

53

35

53

96

124440

4313

Weight

0.2%

24.4%

61.0%85.7%

0.2%

0.2%

0.2%

0.7%

4.5%

8.5%14.3%

100.0%

M-H, Random, 95% CI

8.49 [0.46, 155.59]

1.35 [1.01, 1.80]

1.21 [1.00, 1.45]1.25 [1.07, 1.47]

2.93 [0.12, 70.79]

3.00 [0.12, 72.02]

3.00 [0.13, 71.22]

0.63 [0.11, 3.63]

1.83 [0.93, 3.59]

1.29 [0.79, 2.11]1.44 [0.99, 2.11]

1.28 [1.11, 1.47]

HES 130/0.38-0.45 Crystalloid Risk Ratio Risk Ratio

M-H, Random, 95% CI

0.01 0.1 1 10 100Favours [experimental] Favours [control]

Increased use of RRT RR 1.28 (1.11-1.47)

NNH 52

French: unpublished (with permission)

HES- Sales Analysis- EU5

Note: Total HES Sales irrespective of bottle size and % Source: IMS Colloids Q2 2014 Data

6S trial CHEST

1Reg Action- Jun’13

HES- Sales Analysis- China

Source: IMS Colloids Q2 2014 Data

Note: Total HES Sales irrespective of bottle size and %

David Suzuki 1936-

The whole sector of public dialogue has been badly contaminated, deliberately, by the corporate sector. The whole purpose is to sow confusion and doubt. And it has worked.

Tissue accumulation and gelatin

Skinssnes: Surg Gyne Obs 1947

Bayer: Critical Care Medicine 2011

Single centre, sequential observational trial 2005 – 2009: HES à Gelatin à Crystalloid

Colloid Trials n RR 95%CI

Albumin 56 9920 1.01 0.93 to 1.10

HES 25 9147 1.10 1.02 to 1.19 Gelatin 11 506 0.91 0.49 to 1.72 Dextran 9 834 1.24 0.94 to 1.65

Colloids vs crystalloids

Perel: Cochrane Library 2013

What about crystalloids?

Hartog Jacob Hamburger 1859-1924

‘Normal’ saline

0.9% saline is the most commonly used resuscitation fluid worldwide.

Titrational volumetric determination of osmotic pressure in red blood cells 0.9% concentration of salt in human blood: “Normal“ saline Actual normality = 0.6% saine

Morgan: Crit Care 2008

Why does saline cause an acidosis?

n=851 Cl- 154

Na+ 154

Na+ 154 Cl-

130

24 mEq/l

A “balanced” crystalloid will reduce extracellular SID at a rate that precisely counteracts a dilutional alkalosis induced by weak acids.

Venkatesh: Anaes Int Care 2006

Why does saline cause an acidosis?

n=851

-10

-8

-6

-4

-2

0

2

4

6

8

10

0 50 100 150 200

[Hb] (g/L

SBE

(mEq

/L) Need to give lots of saline

Need to give it fast

Model Result Rat endotoxin infusion + fluid resuscitation with NS, CSL, HES

Saline: ↓survival + worse SID and BE than HES and CSL

Kellum: CCMed 2002

Rat CLP + HCl infusion ↑[Cl-]: hypotension Kellum: Chest 2004

Rat CLP + HCl infusion ↓BE: ↑TNF, IL-6 and IL-10 Kellum: Chest 2006

Cell culture LPS stimulation with HCl or lactic acid

HCl: pro-inflammatory LA: anti-inflammatory (NO, IL-6:IL-10 and NFKB binding)

Kellum: AJPRICP 2004

Animal data: saline / sepsis

Model Result Greyhound denervated kidney + renal arterial hypertonic infusion

NH4Cl > NaCl: ↓RBF + GFR Wilcox: JCI 1983

Rat isolated kidney ↑[Cl-]: ↑ pressor response to A-II

Quilley: BJP 1983

Rat HCl infusion ↑markers of intestinal injury Pedoto: JLCM 2001

Animal data: kidney and gut

Model Result Volunteers: 50ml/kg CSL or NS over 1hr

CSL: ↓ osmolality,↑ diuresis NS: ↑ time to micturition

Wiliams: A&A 1999 Reid Clin Sci 2003

Renal TP recipients: CSL v NS

NS: ↑ Acidosis ↑K+ with markers of intestinal injury

O’Malley: A&A 2005

Elderly surgical patients: CSL v NS

NS: ↑acidosis and CO2 gap Wilkes: A&A 2001

Humans: kidney and gut

Gunnerson: Crit Care 2006

Acidosis and mortality

n=851

Adrenaline-induced lactate

Wutrich: Shock 2010

Observational study n=100

Adrenaline-induced Δ lactate 0-4 hours

Yunos: JAMA 2012 Grade 2 or Grade 3 AKI Use of RRT in ICU

Log rank p=0.001 Log rank p=0.004

Single centre, sequential pilot observational trial 6m saline, gelatin, 4% albumin / 6m RL, PL-148, 20% albumin

n=760

OR 0.52

Yunos: Int Care Med 2015 HR 1.32

Raghunathan: Crit Care Med 2014

Registry, propensity-score based observational trial

Saline vs buffered salt solutions

Shaw: Int Care Med 2014 Hospital mortality 3.0 vs 31.1%

McCluskey: Anesth Analg 2013

Shaw: Ann Surg 2012

Favours Plasmaltye n=926

Favours Saline n=30994

Registry, propensity-score based observational trial Saline vs buffered salt solutions

Saline and acid-base over time Nested, cohort study within the SAFE study

n=691, 3 general ICUs

Bellomo: Crit Care Med 2006

Volume of fluid is predictor of acid base change Changes are minor – alkalosis predominates Influenced by disease severity and time

Plasma-‐Lyte 148® vs saline n=2281 pa6ents

Primary outcome

AKI Secondary outcome:

RRT, Hospital mortality

The Plasma-‐Lyte 148® vs saline study (PLUS)

n=8800 pa6ents

John Maynard Keynes 1883-1946

The difficulty lies, not in new ideas, but in escaping old ones, which ramify, for those brought up with them, as most of us have been, into every corner of our minds.

How?

The physiological fallacy Fluid bolus therapy is self-evidently beneficial

Based on a common phenotype and surrogate variables An inference, that cannot be measured, is made that the patient has inadequate organ blood flow.

The second inference, that also cannot be measured, is that a fluid bolus will restore the complexity of altered haemodynamics in a predictable and safe fashion for the duration of the illness.

FBT and haemodynamics

Bellomo: Crit Care 2014

FBT and renal function

Saotome: Int Care Med 2010

Wan: Anesth Anal 2007

Risk factors for AKI

Non-modifiable Old age Male sex Pre-existing CKD Hypertension Diabetes mellitus Chronic liver disease Chronic heart disease Malignancy

Modifiable Anemia Cardiac surgery Fluid overload Radiocontrast media Nephrotoxic drugs Synthetic colloids (HES, gelatin) Chloride-rich solutions (0.9% saline)

Rewa: Nephrol 2014

Myburgh: NEJM 2013

The endothelial glycocalyx model

Woodcock: BJA 2012

Crystalloid to colloid volume ratios

SAFE (4% albumin) 1.4:1 VISEP (HES 200/0.5) 1.4:1 CHEST (HES 130/0.4) 1.4:1 6-S (HES 130/0.42) 1.0:1 CHRYSMAS (HES 130/0.4) 1.2:1 FIRST (HES 130/0.4) 1.4:1

Mortality at 4 hours Mortality at 4 weeks

Maitland: New Eng J Med 2011

Multicentred open-label RCT Albumin vs saline bolus vs no bolus in febrile hypotensive children

n=3141/3600 Primary outcome: Mortality at 48h

2009-2011

4.6 v 2.6%: HR:1.79 (1.17 to 2.74); p=0.008

Maitland: BMC Med 2013

Multicentred open-label RCT Protocolised liberal v conservative fluid strategy x 7d: ALI

n=1001 Primary outcome: Mortality at 60d

2000-2005

NHLBI ARDSnet: New Engl J Med 2006

Fluid volumes and outcomes

Boyd: Crit Care Med 2011

VASST study: fluid balance / CVP at 12h and 4d n=778

Fluid volumes and outcomes

Bouchard: Kidney Int 2009

Dialysed Non-dialysed

Program to Improve Outcome in Acute Kidney Disease n=618

5 US centres

Restrictive fluid strategies

Brandstrup: Ann Surg 2003

Multicentre, single blinded RCT Restrictive vs standard fluids, surgical patients

n=172 Perioperative complications

Permissive hypovolaemia in burns

Parkland

Permissive

Arlati: Resuscitation 2007

Multicentre, retrospective 2-cohort study n=24

05

10152025

Perce

ntCa

lculat

e_inp

ut_1

05

10152025

Perce

ntCa

lculat

e_inp

ut_2

05

10152025

Perce

ntCa

lculat

e_inp

ut_3

-4800 3200 11200 19200 27200 35200 43200 51200 59200 6720005

10152025

Perce

ntCa

lculat

e_inp

ut_4

fluid_input1

NAME

OF F

ORME

R VA

RIAB

LE

Baseline

Pre-ICU: Add 25 mls/kg bolus

Reduce additional fluid by 25%

Reduce additional Fluid by 50%

22 L 25 L 17 L 12 L

%

Saxena: CTG Noosa 2014

Restrictive fluid strategies Cumulative Fluid Input at Day 7

Mikkelsen: Am J RespCrit Care Med 2012

OR 4.03 (1.53–10.59), p= 0.004

FACCT follow-up: 122 of 213/406 survivors Neuropsychological assessment at 12 months

Fluids are unvalidated, lethal drugs Drug: A term of varied usage. In medicine, it refers to any substance with the potential to prevent or cure disease or enhance physical or mental welfare. (www.WHO.int)

How and why?


Fluids should be administered with the same caution that is used with any intravenous drug Consider the type, dose, indications, contraindications, potential for toxicity and cost.

Resuscitation fluids should only be used in patients with symptomatic hypovolaemia.

How?


Fluid resuscitation is a component of a complex physiological process Identify the fluid that is most likely to be lost and replace the fluid lost in equivalent volumes

Consider serum osmolality and the acid-base status when selecting a resuscitation fluid

Consider cumulative fluid balance and actual body weight when selecting the dose of resuscitation fluid

Consider the early use of catecholamines as concomitant treatment of shock

When, how and why?


Fluid requirements change over time in critically ill patients. The cumulative dose of resuscitation and maintenance fluids is associated with pathological oedema that is associated with adverse outcomes

Oliguria is a normal response to hypovolaemia and should not be used solely as a trigger or end-point for fluid resuscitation, particularly in the post-resuscitation period.

When and how?


Fluid requirements change over time in critically ill patients.

The use of a fluid challenge in the post-resuscitation period (>24 hours) is questionable

The use of hypotonic maintenance fluids is questionable once dehydration has been corrected.

Who and what?


Specific considerations apply to different categories of patients. Bleeding patients require control of haemorrhage and transfusion

Isotonic, buffered salt solutions are pragmatic initial resuscitation fluids for the majority of acutely ill patients.

Consider saline in patients with hypovolaemia and alkalosis

Consider albumin during early resuscitation of patients with sepsis

Who and what?


Specific considerations apply to different categories of patients. Saline or isotonic crystalloids are indicated in traumatic brain injury

Albumin is contraindicated in traumatic brain injury

Hydroxyethyl starch should not be used in any patient population

The safety of other semi-synthetic colloids has not been established The safety of hypertonic saline has not been established

Paracelus 1495

“The dose makes the poison”

Myburgh on Fluids ICN NSW 2015

Health & Medicine

Transcript of Myburgh on Fluids ICN NSW 2015