• 
• 

Three Part Question

In [patients resuscitated from cardiac arrest], does [control of blood glucose levels compared with no control of blood sugar] [improve neurological outcome and survival?]

Clinical Scenario

A 50 year old man presents to the emergency department after resumption of spontaneous circulation (ROSC) following resuscitation from an out of hospital (OOH) ventricular fibrillation (VF) cardiac arrest. He is now stable, his trachea is intubated and his lungs ventilated. He is in a coma. He is awaiting an ICU bed. His blood glucose is 13.5 mmol/L. You wonder whether you should correct his blood sugar.

Search Strategy

Ovid interface on world wide web: Journals@Ovid Full Text and PPV journals (between 1996 - November 2008)
Medline on world wide web (between 1950 - November 2008)
3 handpicked papers by van den Berghe et al.
1 handpicked paper by NICE SUGAR study investigators

[cardiac arrest OR cardiopulmonary resuscitation] AND [blood glucose OR normoglycaemia OR glycaemic control] AND [neurological outcome OR neurologic dysfunction OR survival OR mortality]

Search Outcome

110 papers of which 98 were irrelevant.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Steingrub JS & Mundt DJ 1996 USA	85 patients with ROSC and ICU admission following CPR for in-hospital or out-of-hospital cardiac arrest (VF, PEA, asystole). Groups: Glucose <11.1 mmol/l [N=3] vs Glucose >11.1 mmol/l [N=13].	Retrospective case series chart review	Good/moderate neurologic recovery (Glasgow Pittsburgh Brain Stem Score 1-2) at 24 hours	67% (N=3) vs 0% (N=13) (p=0.03)	Small study group [N=16] with many confounding factors -diabetics (21%) not excluded, potentially perfusing rhythms included (PEA), in & out-of-hospital arrests, findings only applicable when CPR >5 minutes in duration. Cannot comment on survival outcomes. Retrospective study design cannot comment on causality.
			Good/moderate neurologic recovery (Glasgow Pittsburgh Brain Stem Score 1-2) at 48 hours	67% (N=3) vs 0% (N=11) (p=0.03)
Longstreth WT Jr et al 1986 USA	83 patients suffering OOH VF or asystolic cardiac arrest. Groups: Died (no ROSC) (N=43) vs Awoke (N=17) vs Never awoke (N=12)	Prospective cohort study	Blood glucose during CPR and just after ROSC	No association with neurologic recovery	Small study size. Study cannot comment on post-resuscitation glycaemic control.
			Rise of blood glucose during CPR	0.33 mmol/l/min vs 0.22mmol/l/min vs 0.24mmol/l/min (p<0.001)
Longstreth WT Jr & Inui TS 1984 USA	459 patients suffering OOH VF or asystolic cardiac arrest with ROSC. Groups: Never awakening (N=154) vs Awakening with persistent deficits (N=90) vs Awakening with no deficits (N=186)	Retrospective cohort study	Mean glucose level on admission	18.9 mmol/l (p<0.005) vs 15.9 mmol/l (p<0.02) vs 13.9 mmol/l (p<0.0005)	Long time period (10 years) using 2 glucose determination methods. Only markedly high blood glucose compared. Timing and quantity of glucose administered inconsistent. History of diabetes from notes only (underestimated as not formerly tested with HbA1C). Data collector not blinded to hypothesis tested. Retrospective study design cannot comment on causality.
Langhelle A et al 2003 Norway	263 patients suffering OOH cardiac arrest of cardiac origin. Grouped by serum glucose 1-24 hours after admission: <10.6 mmol/L vs >10.6 mmol/L	A retrospective multicentre cohort study	Survival to discharge	OR 3.33 (95% CI 2.01-5.53) (p<0.001, N=263)	Study did not analyse neurological outcome. Multicentre comparison showing survival benefits from many variabilities (pre-hospital care, post-arrest care and pre-morbid state of patients), not specifically glucose control; retrospective study design cannot comment on causality.
			Survival at 1 year from discharge	OR 2.50 (95% CI 1.11 - 5.65) (p<0.05, N=166)
Müllner M et al 1997 Austria	145 non-diabetics suffering (in-hospital or OOH) witnessed VF cardiac arrest with ROSC. Groups: Good neurological recovery (Cerebral performance category (CPC) 1-2) [N=85] vs Bad neurological recovery (CPC 3-5) [N=60]	Retrospective observational study	Glucose levels (mmol/l) at admission	11.8 vs 16.7 (p<0.0001)	Retrospective study cannot demonstrate benefit of avoiding hyperglycaemia; study shows glucose over first 24 hours prognostic of neurological recovery and cannot comment on management of glycaemia post-resus.
			Glucose levels (mmol/l) at 6 hours	8.3 vs 9.7 (p=0.14)
			Glucose levels (mmol/l) at 12 hours	7.3 vs 7.7 (p=30.1)
			Glucose levels (mmol/l) at 24 hours	7.1 vs 8.9 (p=0.006)
			Median glucose levels (mmol/l) over 24 hours	8.1 vs 10.2 (p=0.001)
			Mortality within 6 months	12% vs 83% (p<0.0001)
Losert H et al 2008 Austria	234 patients with witnessed VF or pulseless VT cardiac arrest with ROSC undergoing 24 hours therapeutic cooling. Aged 18-75 years and non-diabetic. Grouped by 12 hour glucose: 3.7-6.4 mmol/l [N=58] vs 6.4-7.9 mmol/l [N=59] vs 8.0-10.7 mmol/l [N=48] vs 11.5-16.2 mmol/l [N=39].	Retrospective analysis of prospective multi-centre trial on mild therapeutic hypothermia.	Good neurological recovery (Pittsburgh cerebral performance category 1-2) at 6 months	64% vs 75% vs 29% vs 18% (p<0.001)	Potential inter-laboratory glucose imprecision as multi-centre trial. Study cannot comment on hypoglycaemic incidence as one-off 12 hour blood glucose measurements used. Study can only comment on non-diabetic patients. Underestimate of diabetics as diagnoses assumed from medical notes, not verified with HbA1C. Retrospective analysis of previous multi-centre trial unable to determinine optimal glucose level or comment on causality.
			Odds ratio of good neurological recovery at 6 months	8.05 (95% CI 3.03-21.40) vs 13.41 (95% CI 4.90-36.67) vs 1.88 ((95% CI 0.67-5.26) vs 1 (ref)
			Survival at 6 months	78% vs 76% vs 42% vs 31% (p<0.001)
Longstreth WT et al 1993 USA	748 patients suffering OOH cardiac arrest. At arrest groups received: 5% glucose in water (N=371) vs 0.45% saline (N=377)	Double blinded randomised controlled trial	Good neurologic recovery (awakening)	No significant association. 16.7% vs 14.6%	Potentially perfusable rhythms included in study (PEA). Types of IV fluid administered post-hospital admission inconstistent. Study cannot comment on post-admission management of glycaemia. Inconsistent and small volumes of 5% glucose administered at arrest.
			Survival to hospital	No significant association. 38.0% vs 39.8%
			Survival to hospital discharge	No significant association. 15.1% vs 13.3%
Van den Berghe G et al 2001 Belgium	All adults [N=1548] admitted to the (predominantly surgical) ICU unit Feb 2000- Jan 2001. Groups: Intensive insulin therapy [N=765] (glucose maintained 4.4-6.1 mmol/l) vs Conventional insulin therapy (glucose maintained 10-11.1 mmol/l)[N=783]	Prospective randomised controlled trial	Mortality during intensive care	4.6% (N=35) vs 8% (N=63) (95% CI 22-62%, p<0.04)	Study not blinded to treatment due to risks of hypoglycaemia. Study on surgical patients only (not post-arrest patients). Study design cannot distinguish between benefits of infused insulin or prevention of hyperglycaemia.
Sunde K et al 2007 Norway	137 patients suffering OOH cardiac arrest with ROSC. Groups: Patients treated according to a standardised treatment protocol (including therapeutic hypothermia, percutaneous coronary intervention (PCI), control of haemodynamics, seizures, ventilation and glucose control 5-8mmol/l) (N=58) vs Patients not treated with standarised treatment protocol (N=61)	Observational study comparing prospective case period [September 2003-May 2005] with retrospective control period [Febuary 1996- Febuary 1998]	Glucose at 12 hours (mmol/l)	9.7 vs 8.0 (p=0.033)	Favourable outcomes cannot be ascribed to glucose control but to interventions such as PCI and therapeutic hypothermia. Study design cannot comment on causality or randomise groups.
			Glucose at 24 hours (mmol/l)	7.5 vs 6.4 (p=0.028)
			Favourable neurological outcome at discharge (cerebral performance category ( CPC) 1-2)	56% (N=34) vs 26% (N=15) (OR 3.61, 95% CI 1.66-7.84, p=0.001)
			Survival to discharge	56% (N=34) vs 31% (N=18) (OR 2.8 (95% CI 1.32-7.84, p=0.007)
			Survival at 1 year	56% (N=34) vs 26% (N=15) (OR 3.61, 95% CI 1.66-7.84, p=0.001)
Calle PA et al 1989 Belgium	417 patients suffering OOH VF or asystolic cardiac arrest. Grouped day 14: Good neurologic response vs Cerebral failure vs Other organ failure	Retrospective (1983-1984)& prosective (1985 onwards) cohort study	Mean glucose level on admission (mmol/l)	16.8 (SEM ±0.61) vs 19.4 (SEM ±0.56) vs 18.8 (SEM ±0.67)(p<0.001)	Not controlled for diabetics. Quantity of glucose administered during CPR inconsistent. Inconsistent time between ROSC and blood glucose sampling. Cannot comment on post-resus glycaemic control. Retrospective study cannot show causality - cannot exclude other resuscitation parameters contributing brain damage, and cannot comment on management of post-resus glycaemia.
Oksanen T et al 2007 Finland	99 patients suffering OHH VF cardiac arrest with ROSC treated with therapeutic hypothermia. Groups: strict glucose control (blood glucose 4-6 mmol/l) (N=39) vs moderate glucose control (6-8 mmol/L) (N=51)	Randomised controlled trial	Mortality at 30 days	No significant association 33% (N=13) vs 35% (N=18) (95% CI -18% to +22%, p=0.846)	Similar glucose levels between groups. Ethical issues prevent groups being randomised to blood glucose levels >8 mmol/l. Groups not blinded. Study ceased prematurely due to small mortality differences between groups. Small sized study groups with only ~10% power to detect ~8% mortality difference(19).
			Moderate hypoglycaemia (<3mmol/l)	18% (N=7) vs 2% (N=1) (p<0.01)
			Severe hypoglycaemia (<2.2mmol/l)	0% vs 0%
Van den Berghe et al 2003 Belgium	All adults in the Van Den Berghe 2001 study (N=1548) (admitted to the (predominantly surgical) ICU unit Feb 2000- Jan 2001). Multivariate logistic regression analysis of mortality on: Mean blood glucose level (per 1.11mmol/l added) and Amount of infused insulin (per 10 units added/day)	Prospective randomised controlled trial	Mortality	OR=1.30 and OR=1.06 (p=<0.001 & p=0.005 respectively)	Study not blinded due to adverse effect of hypoglycaemia. Study on surgical patients (not post-arrest patients). Study shows glucose level and not quantity of insulin infused is related to mortality in 2001 Van den Berghe study.
Van Den Berghe et al 2006 Belgium	1200 adult patients admitted to medical ICU for ≥3 days. Groups: Strict glucose control (SGC) (4.4-6.1mmol/l) [N=595] vs Conventional therapy (insulin administered when glucose >11.9mmol/l) [N=605]	Prospective randomised controlled trial	In-hospital mortality	No significant association 37.3% (N=222) vs 40% (N=242) (p=0.33)	Not possible to blind treatment due to adverse effects of hypoglycaemia. Findings not specific to post-resuscitation patients.
			ICU mortality	No significant association 24.2% (N=144) vs 26.8% (N=162) (p=0.31)
			ICU mortality in patients admitted ≥3 days	43.0% (N=166) vs 52.5% (N=200) (p=0.009)
Skrifvars MB et al 2003 Finland	98 patients suffering OOH witnessed VF cardiac arrest. Grouped on mean blood glucose (mmol/l) during first 72 hours: 5.8-6.8 [N=22] vs 6.9-7.9 [N=22] vs 7.9-8.9 [N=22] vs 9.1-27.9 [N=22]	Retrospective cohort study	Survival at 6 months	9% vs 23% vs 50% vs 64% (p<0.002)	Diabetic status not accounted for. Neurological recovery not included in analysis as missing data. Potentially confounding factors not accounted - 11% treated with mild hypothermia, diabetic status (13%) not adjusted, different methods of blood glucose measurement. Retrospective study cannot comment on causality or post-resus management of glycaemia.
MacKenzie CF 1975 West Indies	100 consecutive cases of cardiac arrest including 10 with both ROSC and blood glucose levels. Grouped by blood glucose level: <2.2 mmol/l [N=1] vs 2.2-6.6 mmol/l [N=4] vs 6.7 – 22.2 mmol/l [N=4] vs >22.2 mmol/l [N=1]	Retrospective case series	Survival to discharge from hospital	0% vs 25% [N=1] vs 25% [N=1] vs 0%	Case series with small numbers of patients and no statistically significant findings.
NICE-SUGAR Study investigators, 2009, Australia, New Zealand & Canada	6022 medical and surgical patients admitted to intensive care and randomised to: Intensive insulin control (4.5-6 mmol/l) [N=3010] vs conventional insulin control (≤10.0 mmol/l) [N=3012]	International multi centre randomised controlled trial	Mortality at 90 days	27.5% [N=829] vs 24.9% [N=751] OR 1.14 (CI 1.02 – 1.28) p value 0.02	Study cannot comment on neurological outcome of patients. No subgroup of post cardiac arrest patients. Not possible to blind treatment due to adverse effects of hypoglycaemia 12.5 – 13 % of patients given steroids

Comment(s)

Studies have shown high glucose levels post cardiac arrest to be prognostic for both poor neurological outcome and poor survival. The 2001 and 2003 studies in a surgical ICU by Van den Berghe et al showed a survival benefit of tight glucose control (4.4-6.1 mmol/l) which protected the peripheral and central nervous systems. However this survival benefit was not shown in the 2006 study in a medical ICU by Van den Berghe except when a patient was admitted for to ICU ≥3 days. The recent randomised controlled trial by Oksanen involving OOH VF cardiac arrest patients showed no difference in survival between strict glucose control (4-6mmol/l) and moderate glucose control (6-8 mmol/l). It did show more frequent (18%) episodes of moderate hypoglycaemia (<3.0 mmol/l) in the strict glucose control group; although no episodes of severe hypoglycaemia (<2.2mmol/l) occurred in either group. The multicentre trial by Losert et al supported this finding and showed no survival or neurological benefit of glucose levels <6.4 mmol/l compared with 6.4-8mmol/l. The NICE SUGAR study showed a survival benefit from conventional glucose control (<10.0 mmol/l) when compared with strict glucose control (4.5 – 6.0 mmol/l). The strict glucose control group suffered more episodes of severe hypoglycaemia (<2.2 mmol/l) than the conventional glucose control group (6.8% compared with 0.5% respectively). The Intensive Care Society and European Resuscitation Council states that "patients admitted to a critical care environment after cardiac arrest should have their blood glucose monitored frequently and hyperglycaemia treated with an insulin infusion. The blood glucose concentration that triggers insulin therapy, and the target range of blood glucose concentrations, should be determined by local policy"(15)(18). A recent consensus statement on post cardiac arrest syndrome by Neumar et al states that "regardless of the chosen glucose target range, blood glucose must be measured frequently, especially when insulin is started and during the cooling and rewarming periods"(16). Current evidence would suggest glucose levels between 4.5 and 10 mmol/l as stricter and lower glucose levels could expose patients to harmful hypoglycaemia. Further large prospective studies specifically on OOH cardiac arrest patients are necessary to determine the optimal glucose levels.

Clinical Bottom Line

It seems reasonable to follow Intensive Care Society Guidance "patients admitted to a critical care environment after cardiac arrest should have their blood glucose monitored frequently and hyperglycaemia treated with an insulin infusion. The blood glucose concentration that triggers insulin therapy, and the target range of blood glucose concentrations, should be determined by local policy" (18). In patients suffering cardiac arrest, current evidence would suggest keeping glucose levels between 4.5 and 10 mmol/l as stricter and lower glucose levels could expose patients to harmful hypoglycaemia. Regular glucose monitoring and avoiding severe hypoglycaemia (<2.2mmol/l) are more important than strict glucose control.

References

1. Steingrub JS & Mundt DJ Blood glucose and neurologic outcome with global brain ischemia Critical Care Medicine 1996; May; 24:802-806
2. Longstreth WT Jr, Diehr P, Cobb LA, Hanson RW, Blair AD Neurologic outcome and blood glucose levels during out-of-hospital cardiopulmonary resuscitation Neurology 1986;36:1186-91
3. Longstreth WT Jr & Inui TS High Blood Glucose Level on Hospital Admission and Poor Neurological Recovery After Cardiac Arrest Annals of neurology 1984 Jan;15(1):59-63.
4. Langhelle A, Tyvold SS, Lexow K, Hapnes SA, Sunde K, Steen PA In-hospital factors associated with improved outcome after out-of-hospital cardiac arrest. A comparison between four regions in Norway Resuscitation 2003; 56; 247-263
5. Müllner M, Sterz F, Binder M, Schreiber W, Deimel A, Laggner AN Blood Glucose Concentration After Cardiopulmonary Resuscitation Influences Functional Neurological Recovery In Human Cardiac Arrest Survivors Journal of Cerebral Blood Flow & Metabolism 1997; 17; 430-436
6. Losert H, Sterz F, Roine RO, Holzer M, Martens P, Cerchiani E, Tiainen M, Mullner M, Laggner AN, Herkner H, Bischof MG Strict normoglycaemic blood glucose levels in the therapeutic management of patients within 12h after cardiac arrest might not be necessary. Resuscitation 2008 Feb;76(2):214-20. Epub 2007 Sep 17.
7. Longstreth WT, Copass MK, Dennis LK, Rauch-Matthews ME, Stark MS, Cobb LA Intravenous glucose after out-of-hospital cardiopulmonary arrest: A community-based randomized trial Neurology 1993; 43, 2534-2541
8. Van den Berge G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, Vlasselaers D, Ferdinande P, Lauwers P, Bouillon R Intensive insulin therapy in critically ill patients New England Journal of Medicine 2001; vol 345; 19; 1359-1367
9. Sunde K, Pytte M, Jacobsen D, Mangschau A, Jense LP, Smedsrud C, Draegni T, Steen PA Implementation of a standarised treatment protocol for post resuscitation care after out-of-hospital cardiac arrest Resuscitation 2007; 73, 29-39
10. Calle PA, Buylaert WA, VanHaute OA, The Cerebral Resuscitation Study Group Glycemia in the Post-Resuscitation Period Resuscitation 1989; 17, suppl S181-S188
11. Oksanen T, Skrifvars MB, Varpula T, Kuitunen A, Pettila V, Nurmi J, Castren M Strict versus moderate glucose control after resuscitation from ventricular fibrillation Intensive Care Medicine 2007; 33, 2093-2100
12. Van den Berge G, Wouters P, Bouillon R, Weekers F, Verwaest C, Schetz M, Vlasselaers D, Ferdinande P, Lauwers P Outcome benefit of intensive insulin therapy in the critically ill: Insulin dose versus glycaemic control Critical Care Medicine 2003; 31;2; 359-366
13. Van den Berghe G, Wilmer A, Hermans G, Meersseman W, Wouters PJ, Milants I, Van Wijngaerden E, Bobbaers H, Bouillon R Intensive Insulin Therapy in the Medical ICU The New England Journal of Medicine 2006; 354 (5); 449-461
14. Skrifvars MB, Pettila V, Rosenberg PH, Castren M A multiple logistic regression analysis of in-hospital factors related to survival at six months in patients resuscitated from out-of-hospital ventricular fibrillation Resuscitation 2003; 59, 319-328
15. Nolan JP, Deakin CD, Soar J, Bottiger BW, Smith G European Resuscitation Council Guidelines for Resuscitation 2005 Section 4. Adult advanced life support. Resuscitation 2005; Dec 67 Suppl 1:S39-86
16. Neumar RW, Nolan JP, Adrie C, Aibiki M, Berg RA, Böttiger BW, Callaway C, Clark RS, Geocadin RG, Jauch EC, Kern KB, Laurent I, Longstreth WT Jr, Merchant RM, Morley P, Morrison LJ, Nadkarni V, Peberdy Post-cardiac arrest syndrome: epidemiology, pathophysiology, treatment, and prognostication. A consensus statement from the International Liaison Committee on Resuscitation Circulation 2008 Dec 2;118(23):2452-83. Epub 2008 Oct 23.
17. Mackenzie CF A review of 100 cases of cardiac arrest and the relation of potassium, glucose and haemoglobin levels to survival West Indian Journal of Medicine 1975; 24: 39-45
18. The Intensive Care Society Standards for the Management of Patients After Cardiac Arrest 2008 The Intensive Care Society 2008
19. Padkin A Comment on "strict versus moderate glucose control after resuscitation from ventricular fibrillation" by Oksanen et al. Intensive Care Medicine 2008 May;34(5):969; author reply 970. Epub 2008 Mar 21.
20. The George Institute Normoglycaemia in Intensive Care Evaluation and Survival Using Glucose Algorithm Regulation - NICE-SUGAR
21. NICE-SUGAR Study Investigators. Intensive versus conventional glucose control in critically ill patients. New England Journal of Medicine 2009 Mar 26;360(13):1283-97.