(inglés) guía infusión insulina para manejo hiperglicemia pctes críticos [2012]

7/28/2019 (Ingls) Gua Infusin Insulina para manejo Hiperglicemia pctes Crticos [2012]

1/26

Crit Care Med 2012 Vol. 40, No. 12 3251

From the Indiana University Health MethodistHospital, Indianapolis, IN (JJ); University o PittsburghMedical Center, Pittsburgh, PA (NB); Stamord Hospital,Stamord, CT (JK); Intermediate Care Unit, ChildrensHospital Boston, Boston, MA (MA); Endocrine Consultsand Care, Evanston, IL (SSB); Critical Care and Surgery,

Perelman School o Medicine at the University oPennsylvania, Philadelphia, PA (CD); PCC and SleepMedicine, University o Tennessee Health ScienceCenter, Memphis, TN (AXF); University o MissouriKansas City, Kansas City, MO (DG); University oPennsylvania, Philadelphia, PA (BK); SICU, Tuts/NewEngland Medical Center, Boston, MA (SAN); Pediatric

Critical Care, Riley Hospital For Children, Indianapolis, IN(MR); Novant Health, Winston Salem, NC (KS); Barnes-Jewish Hosptial, St. Louis, MO (LS); Barnes JewishHospital, St. Louis, MO (BT); Emory University Schoolo Medicine, Atlanta, GA (GU); Washington UniversitySt. Louis, St. Louis, MO (JM); McMaster University,Hamilton, Ontario, Canada (HS).

Supplemental digital content is available or thisarticle. Direct URL citations appear in the printed textand are provided in the HTML and PDF versions o thisarticle on the journals Web site (http://journals.lww.com/ccmjournal).

Dr. Agus has consulted or the Diabetes Technology

Society. He also has a pending pat ent on an ECMO-based glucose, sensor, which is not connected to ideadiscussed in this article. Dr. Braithwaite has a U.S. pat-ent. Dr. Kohl has received grant support rom Amylin andEli Lilly. Dr. Krinsley has perormed consulting work orMedtronic Inc., Edwards Lie Sciences, Baxter, RocheDiagnostics, and Optiscan Biomedical and has received

speakers ees rom Edwards Lie Sciences, RocheDiagnostics and Sano-Aventis. Dr. Nasraway has con-sulted or Optiscan, Echo Therapeutics. Dr. Geehan hasreceived grant support rom the Department o DeenseResearch. Dr. Rigby has received consulting ees romMedtronic. Dr. Schallom has received honoraria/speak-ing ees rom Roche Laboratories Speakers Bureau on

Glycemic Control. The remaining authors have not dis-closed any potential conficts o interest.

Listen to the iCritical Care podcasts or anin-depth interview on this article. Visit www.sccm.org/iCriticalCare or search SCCM at iTunes.

For inormation regarding this article, E-mail:

[email protected] 2012 by the Society o Critical Care

Medicine and Lippincott Williams and WilkinsThe American College o Critical Care Medicine

(ACCM), which honors individuals or their achievementsand contributions to multidisciplinary critical care medi-cine, is the consultative body o the Society o CriticalCare Medicine (SCCM) that possesses recognized exper-tise in the practice o critical care. The College has devel-oped administrative guidelines and clinical practice pa-rameters or the critical care practitioner. New guidelinesand practice parameters are continually developed, andcurrent ones are systematically reviewed and revised.

DOI: 10.1097/CCM.0b013e3182653269

Objective: To evaluate the literature and identiy important

aspects o insulin therapy that acilitate sae and eective inusion

therapy or a defned glycemic end point.

Methods:Where available, the literature was evaluated using

Grades o Recommendation, Assessment, Development, and Evalu-

ation (GRADE) methodology to assess the impact o insulin inu-sions on outcome or general intensive care unit patients and

those in specifc subsets o neurologic injury, traumatic injury, and

cardiovascular surgery. Elements that contribute to sae and eec-

tive insulin inusion therapy were determined through literature

review and expert opinion. The majority o the literature supporting

the use o insulin inusion therapy or critically ill patients lacks

adequate strength to support more than weak recommendations,

termed suggestions, such that the dierence between desirable

and undesirable eect o a given intervention is not always clear.

Recommendations:The article is ocused on a suggested glyce-

mic control end point such that a blood glucose 150 mg/dL trig-

gers interventions to maintain blood glucose below that level and

absolutely


2/26

3252 Crit Care Med 2012 Vol. 40, No. 12

The notion o tight glycemiccontrol (GC) became moreprominent in the critical careliterature in 2001 when a land-

mark study by Van den Berghe and col-leagues (1) demonstrated a signicantmortality benet when maintaining bloodglucose (BG) between 80 and 110 mg/dL.Prior to that publication, GC was not a

high priority in most intensive care unit(ICU) patients. Data have conrmed theobservation that hyperglycemia is associ-ated with an increase in death and inec-tion, seemingly across the board amongmany case types in the ICU (2, 3). Manycenters have attempted to assess the ea-sibility o maintaining normoglycemiain critically ill patients and to urtherestablish the potential risk or benet othis approach in a variety o ICU patientsubsets. While there have been confictingresults rom numerous studies, the ques-tion is no longer whether GC is benecialor not, but rather what is the appropriatedegree o GC that can be accomplishedsaely and with justiable utilization oresources.

This Clinical Practice Guideline willevaluate the available literature andaddress aspects o implementation thatpermit sae and eective insulin inusiontherapy. Methodology and assessment willbe emphasized to help clinicians achievethe BG goal that is considered to havethe greatest benet and saety or theirpatient population while avoiding clini-

cally signicant hypoglycemia.

GUIDELINE LIMITATIONS

Guidelines are limited by the availableliterature and the expertise o the writingpanel and reviewers. The recommenda-tions are not absolute requirements, andtherapy should be tailored to individualpatients and the expertise and equipmentavailable in a particular ICU. The use oan insulin inusion requires an appro-priate protocol and point-o-care (POC)

monitoring equipment with requent BGmonitoring to avoid hypoglycemia. Rec-ommendations may not be applicable toall ICU populations, and limitations willbe discussed when applicable. Future lit-erature may alter the recommendationsand should be considered when applyingthe recommendations within this article.

Intravenous (IV) insulin will be the pri-mary therapy discussed, but subcutane-ous (SQ) administration may also have arole or GC in stable ICU patients. Otheragents and approaches, including oral

hypoglycemic drugs, and other antidia-betic agents may be continued or restartedin selected patients, but will not be dis-cussed in this article. Studies evaluatinginsulin as a component o other therapies(such as glucoseinsulinpotassium) werenot evaluated.

TARGET PATIENT POPULATION

FOR GUIDELINEThese guidelines are targeted to adult

medical and surgical ICU patients as agroup, but individual population dier-ences regarding therapy or monitoring

will be discussed. Data on the glycemicmanagement o pediatric ICU patientsare limited, but will be described whereavailable.

METHODOLOGY

The Guideline Task Force was com-

posed o volunteers rom the Society oCritical Care Medicine with a specicinterest in the topic and the guidelineprocess. The Task Force members devel-oped a list o clinical questions regard-ing the appropriate utilization o insulininusions to achieve GC, consideringpatient/populations, interventions, com-parisons, and outcomes. Applicable lit-erature was compiled using a variety osearch engines (PubMed, OVID, GoogleScholar, reerence lists rom other pub-lications, search o Clinicaltrials.gov,

and the expertise and experience o theauthors). Searches were perormed peri-odically until the end o 2010 using theollowing terms: acute stroke, BG, car-diac surgery, critical care, critical illness,critically ill patients, dextrose, glucose,glucose control, glucose metabolism,glucose meters, glucose toxicity, glyce-mic control, glycemic variability, hyper-glycemia, hypoglycemia, ICU, insulin,insulin inusion, insulin protocols, insu-lin resistance, insulin therapy, inten-sive care, intensive insulin therapy,

mortality, myocardial inarction, neu-rocognitive unction, neuroprotection,outcomes, pediatric, pediatric intensivecare, point-o-care, point-o-care testing,sepsis, sternal wound inection, stresshyperglycemia, stress, stress hormones,stroke, subarachnoid hemorrhage, sur-gery, tight glycemic control protocols,and traumatic brain injury (TBI).

Published clinical trials were used asthe primary support or guideline state-ments, with each study evaluated andgiven a level o evidence. Abstracts and

unpublished studies or data were notincluded in the analysis. The Grading oRecommendations Assessment, Develop-ment and Evaluation (GRADE) system

was used to rate the quality o evidenceand strength o the recommendationor each clinical practice question (4).

A member o the GRADE group was avail-able to provide input and answer meth-

odologic questions.Meta-analyses using RevMan andGRADEPro sotware were applied to orga-nize evidence tables, create orest andunnel plots, and draw conclusions aboutthe overall treatment eects or specicoutcomes applicable to a particular rec-ommendation (5, 6).

Recommendations are classied asstrong (Grade 1) or weak (Grade 2) andare ocused on specic populations

where possible. Strong recommenda-tions are listed as recommendationsand weak recommendations as sugges-

tions. Throughout the development othe guidelines, there was an emphasis onpatient saety and whether the benet oadherence to the recommendation wouldoutweigh the potential risk, the burdenon sta, and when possible, the cost. Ithe risk associated with an interventionlimited the potential or benet, or i theliterature was not strong, the statement

was weakened to a suggestion. Individualpatient or ICU circumstances may infu-ence the applicability o a recommenda-tion. It is important to recognize that

strong recommendations do not neces-sarily represent standards o care.

Numerous discussions among theauthors led to consensus regarding therecommendations. Individual membersor subgroups drated the recommenda-tions and justications. Subsequently,each recommendation was reviewed bythe Task Force members who were pro-

vided the opportunity to comment, pro-pose changes, and approve or disapproveeach statement. Once compiled, eachmember was again asked to review the

article and provide input. Consensus wassought or recommendation statements,and controversial statements were repeat-edly edited and eedback provided throughsecret ballots until there was consensus.

Actual or potential conficts o interestwere disclosed annually, and transpar-ency o discussion was essential. Externalpeer review was provided through theCritical Care Medicine editorial process,and approval was obtained by the govern-ing board o the Society o Critical CareMedicine.


3/26


RECOMMENDATIONS

While the initial goal was to suggestglycemic targets or critically ill patients,the limited available literature has nar-rowed the scope o this article and theability to make recommendations or spe-cic populations. An overriding ocus ison the sae use o insulin inusions. Theglycemic goal range o 100150 mg/dL

is a consensus goal, and while it diersslightly rom the more stringent goal o110140 mg/dL or selected populations,recently published by the American Dia-betes Association, and the overall glucosegoal o 140180 mg/dL, this dierence isnot likely to be clinically signicant (7).

1. In adult critically ill patients, doesachievement o a BG < 150 mg/dL withan insulin inusion reduce mortality,compared with the use o an insulininusion targeting higher BG ranges?

We suggest that a BG 150 mg/dLshould trigger initiation o insulin ther-apy, titrated to keep BG < 150 mg/dL ormost adult ICU patients and to maintainBG values absolutely 150 mg/dL (12). Insulin inu-sion therapy is recommended or mostcritically ill patients, although selectedpatients may be managed on SQ therapyas discussed later in the article.

Several large randomized controlledtrials (RCTs) have addressed the impacto GC on mortality with variable results,although the ability to compare results

is hampered by diering populations,methodology, and end points (Table 1) (1,1316). Small randomized trials, denedas 180 mg/dL, which is undesirable withrespect to the immunosuppressive eectsand potential to exceed the renal thresh-old or glucosuria. Our recommendationis similar to the American Diabetes Asso-ciation guidelines or initiation o insulinor a glucose threshold no higher than180 mg/dL, and that a more stringentgoal o 110140 mg/dL may be used ithere is a documented low rate o severehypoglycemia (7).

In contrast, there are at least three pub-lished meta-analysis reviews published inthe peer-review literature that have sug-gested no signicant mortality benetsrom insulin inusion therapy to maintaintight GC (BG


4/26


Table 1. Summary o key clinical trials used to evaluate the impact o glycemic controla

Author

Study Quality

Intensive Care Unit

Population Design/End Point Design Assessment

Large randomized controlled trials

Van den Bergheet al (1)

Surgical, mechanicalventilation

Randomized80110 mg/dL vs. 180200 mg/dL

Research RN titrated insulin per protocol

Single center, Evaluated mean morningglucose

Van den Berghe

et al (14)

Medical, expected intensive

care unit stay >72 hrs

Randomized

Bedside RN titrated per paper protocol

Single center

Evaluated mean morning glucosePreiser et al (15) Medical and surgical Randomized

80110 mg/dL vs. 140180 mg/dL

Bedside RN titrated per protocol

Multicenter, Evaluated all glucose

values, also median morning value

The NICE-SUGAR

Investigators (16)

Medical and surgical Randomized

80110 mg/dL vs. 140180 mg/dL

Adjusted via computerized algorithm

Multicenter Evaluated mean time-

weighted glucose

Outcome based on 90-day mortalitySmall randomized controlled trialsBrunkhorst

et al (17)

Sepsis Randomized

80110 mg/dL vs. 180200 mg/dL

Bedside RN titrated per Van den Berghe protocol

Multicenter, Evaluated mean morning

glucose

De La Rosa

et al (18)

Medical and surgical Randomized 80110 vs. 180200 mg/dL

Bedside RN titrated per protocol

Single center

Evaluated mean morning glucose, also

daily minimum and maximum valuesArabi et al (19) Medical and surgical Randomized 80110 mg/dL vs. 180200 mg/dL

Bedside RN titrated per paper protocol

Single center, Evaluated daily average

glucoseFarah et al (20) Medical with >3- day

length o stay

Randomized 110140 mg/dL vs. 140200 mg/dL Single center, reported overall average

glucoseGrey and

Perdrizet (21)

Surgical, excluded patients

with diabetes

Randomized 80120 mg/dL vs. 180220 mg/dL Single center, reported daily average

and overall average glucose

Mackenzie et al (22) Medical and surgical Randomized 72 108 mg/dL vs. 180198 mg/dL Two centers, multiple glucose end

points reportedLarge cohortKrinsley (23) Medical and surgical Observational cohort


5/26


Findings

Commentc

No. o Patients

Actual Glycemic End Points

Mean sd (mg/dL)Hospital Mortality

Odds Ratio [95%

Condence Interval]bControl Glycemic Control Control Glycemic Control

783 765 Daily 153 33 Daily 103 19 0.64 [0.45, 0.91] Control glucose elevated with IV dextrose,stopped early or benet

605 595 Daily 153 31 Daily 111 29 0.89 [0.71, 1.13] Control glucose elevated with IV dextrose

542 536 144 (IQR 128162)

medianall values

117 (IQR 108130)

medianall values

1.27 [0.94, 1.7] Stopped early or hypoglycemia, many

protocol violations

3050 3054 144 23 115 18 28-day

1.09 [0.96, 1.23]

90-day1.14 [1.02, 1.28]

Glycemic control group did not achieve

target

290 247 Median daily 138

(IQR 111184)

Median daily 130

(IQR 108167)

28-day mortality

0.94 [0.63, 1.38]

Control target articially elevated with IV

dextrose. Study stopped early or hypo-

glycemia risk. Inadequate size to detect

dierence in mortality250 254 Medianall values

149 (IQR 124180)

Medianall values

120 (IQR 110134)

28-day mortality

1.08 [0.75, 1.54]

Did not achieve target glucose, in-

adequate sample to detect mortality

dierence257 266 171 34 115 18 0.78 [0.53, 1.13] Small trial, inadequate size to show

mortality impact48 41 174 20 142 14 1.37 [0.59, 3.16] Baseline imbalance in diabetes incidence

and admission glucose27 34 179 61 125 36 mg/dL,

Daily mean value lower

on each day

0.47 [0.12, 1.86] Lower nosocomial inection incidence

with glycemic control

119 121 8.4 2.4 7.0 2.4 0.73 [0.43, 1.24] Inadequate size to show mortality impact

800 800 152 93 131 55 0.45 [0.34, 0.60] High protocol adherence, subcutaneous

and IV insulin

942 2612 214 41 177 30 0.27 [0.19, 0.39]

p < .001

Remote historical controls on SQ insulin

only

2366


6/26


A. We suggest that there is no consis-tently demonstrated dierence inseveral morbidity measures (renalailure, transusion, bacteremia,polyneuropathy, and ICU length ostay [LOS]) when evaluated in thegeneral adult ICU population.

[Quality o evidence: very low]The ollowing were considered asmorbidity outcomes or evaluation,acute renal replacement therapy, inci-dence o transusion, bacteremia, criti-cal illness polyneuropathy, and ICU LOS.To analyze ICU LOS in those studies in

which data were reported nonpara-metrically, the median value was usedand interquartile range (IQR, 1.35) wasused as an estimate o sample standarddeviation (sd). Duration o mechanical

ventilation was not analyzed as there

was consensus that too many conound-ing variables existed or this outcome. Areduction in critical illness polyneurop-athy was not analyzed as this potentialbenet was reported in only one study.Our analysis suggests that no evidenceo benet was ound in ICU LOS with OR0.05, 95% CI [0.14, 0.05]; prevention

o bacteremia OR 0.81, 95% CI [0.58,1.11]; need or transusion OR 1.06, 95%CI [0.90, 1.26]; or need or renal replace-ment therapy OR 0.90, 95% CI [0.7,1.16], but variable study design, popula-tions, and end points limit the analysis.The orest and unnel plots are avail-able in Supplemental Digital Content 1(http://links.lww.com/CCM/A589 ).

B. We suggest implementation o mod-erate GC (BG < 150 mg/dL) in thepostoperative period ollowing cardiac

surgery to achieve a reduced risk odeep sternal wound inection andmortality.

[Quality o evidence: very low]The only large-scale RCT to date evalu-

ating the impact o tight GC on morbid-ity and mortality in a population weighted

with postoperative cardiac surgicalpatients was published in 2001 (1). Almosttwo thirds o this study population under-

went cardiac surgery. Patients in the GCgroup (80110 mg/dL) had lower ICU andhospital mortality rates compared withconventional therapy (BG 180200 mg/dL). Morbidity benets or the GC groupincluded a reduced need or renal replace-ment therapy, less chance o hyperbiliru-binemia, earlier cumulative likelihood o

weaning rom mechanical ventilation, andICU and hospital discharge. A ollow-up

Figure 1. Forest plots o (A) hospital or 28-day mortality and (B) intensive care unit mortality (1, 1421, 2529). CI, condence interval;MH, Mantel-Haenszel.
http://links.lww.com/CCM/A589http://links.lww.com/CCM/A589


7/26


preplanned subanalysis o the 970 high-risk cardiac surgery patients rom the orig-inal study conrmed a survival benet dueto GC up to 2 yrs ater hospital dischargeand longer or the subset treated or atleast 3 days (34). Additionally, a series oreports rom a clinical database o diabeticcardiac surgery patients suggested thatmaintenance o BG < 150 mg/dL is asso-

ciated with a reduction o sternal woundinection and an incremental decrease inhospital mortality compared with remotehistorical control patients treated withsliding-scale insulin (24, 3537). Anotherretrospective review o patients treated

with a combination o IV and SQ insulin inthe postoperative period showed a strongassociation between GC and reduction inmorbidity and mortality (38).

C. In the population o critically ill injured(trauma) ICU patients, we suggest thatBG 150 mg/dL should trigger initia-

tion o insulin therapy, titrated to keepBG < 150 mg/dL or most adult traumapatients and to maintain BG valuesabsolutely < 180 mg/dL, using a pro-tocol that achieves a low rate o hypo-glycemia (BG 70 mg/dL) to achievelower rates o inection and shorterICU stays in trauma patients.

[Quality o evidence: very low]A hypermetabolic stress response

resulting in hyperglycemia is commonin the trauma population (39). Hypergly-

cemia on admission or within the rst 2ICU days may be predictive o poor out-come (longer LOS, more inection) andhigher mortality (4042). Additionally,persistence o hyperglycemia is associated

with poor outcome (4345). A pre-traumadiagnosis o insulin-dependent diabetes

was not associated with higher mortalityor hospital LOS (46).

The benet o insulin therapy onimproving trauma patient outcome hasnot been clearly demonstrated (Table 2)(16, 27, 47, 48). In the Normoglycemia in

Intensive Care EvaluationSurvival UsingGlucose Algorithm Regulation (NICE-SUGAR) multicenter trial o 6,104 patients,trauma patients represented 15.5% o theconventional therapy group (BG goal 140180 mg/dL) and 14% o the GC group (goal80110 mg/dL) (16). Subset analysis indi-cated a trend toward lower mortality inthe GC group (OR 0.77, 95% CI [0.5, 1.18];p = .07). Although these data are hypoth-esis-generating and that trauma patientsmay benet more rom GC than the otherICU patients, additional prospective trials

are needed to conrm this nding. Thus,at this time we recommend that traumaICU patients should be managed in thesame ashion as other ICU patients.

D. We suggest that a BG 150 mg/dLtriggers initiation o insulin therapyor most patients admitted to an ICU

with the diagnoses o ischemic stroke,

intraparenchymal hemorrhage, aneu-rysmal subarachnoid hemorrhage, orTBI, titrated to achieve BG values abso-lutely < 180 mg/dL with minimal BGexcursions


8/26


Table 2. Summary o clinical trials evaluating impact o insulin therapy on patient outcome ater trauma

Author

Study Quality No. o Patients Findings

Comments

Intensive

Care Unit

Population

Design/Glucose

End Point

Design

Assessment Control

Glycemic

Control

Actual Glycemic End Points

Mean sd (mg/dL)

Hospital

Mortality

Odds Ratio

[95%

Condence

Interval]Control

Glycemic

Control

Scaleaet al (27)

Traumaintensive

care unit

Prospective datacollection, two

patient series

beore and

ater protocol

100150 mg/dL

Single center,reported

patterns

o glucose

control in

week 1.

Over 51%

had glucose

>150 mg/dL

in rst week

(poor protocol

eect)

1021 1108 NA NA Adjustedoutcomes:

Pre vs. post;

mortality

1.4 [1.1, 10]

Reducedvent days

and length

o stay with

improved

pattern o

glucose

control

Reed

et al (47)

Surgical and

trauma ICU,

n = 7261

Retrospective

query o

prospective

database,

pre- and

post-protocol

implementation

Single center,

uncontrolled

protocol

compliance.

Measured

glucose

control by

year. End point

was estimated

mortality ratio

measured

as actual/

estimated

mortality

Not

reported

by group

Not

reported

by group

Reported by

study year

2003: 141

2004: 134

Reported by

study year

2005: 129

2006: 125

p < 0.01

Estimated

mortality

ratio

measured

as actual/

estimated

mortality

unchanged

Lower mean

glucose not

correlated

with

estimated

mortality

risk

reduction.

Other actors

changed

during

observational

period (key

personnel,

population,

quality

emphasis)

Collieret al (48)

Trauma, onmechanical

ventilation

Prospectivepostprotocol

(80110 mg/dL)

vs. historical

control

Single center,Reported

mean glucose

Pre- vs. post-

mortality not

reported

383 435 130 11 124 13 1 Glucosedays above

150 mg/dL:

2.16 [1.0,

4.6]

p = .049

Preprotocol,historical

control, no

signicant

reduction

in mean

glucose pre

to post, did

not achieve

glucose

targetThe NICE-

SUGAR

Investiga-

tors (16)

Trauma

subset

Randomized,

Tight 81108 mg/

dL vs. control

< 180 mg/dL

Multicenter,

Reported mean

glucose and

time-weighted

mean overall

465 421 Not reported

or trauma

subset

Not reported

or trauma

subset

0.77 [0.5,

1.18]

p = .07 or

heterogene-

ity 90-daymortality

Hypothesis-

generating

subset

analysis

NA, not applicable, not available.

The impact o insulin-induced hypo-glycemia has varied among populations,and in some reports, hypoglycemia wasthought to be a marker or more seriousunderlying illness (79, 80). Risk actorsor SH include renal ailure, interruptiono caloric intake without adjustments inthe insulin inusion, sepsis with the use

o vasoactive inusions, insulin therapy,and the use o continuous renal replace-ment therapy with a bicarbonate-basedreplacement fuid (81). Some authors alsoound that diabetes, mechanical ventila-tion, emale sex, greater severity o illness,and longer ICU stays are associated withincreased risk o SH (80, 82). Additionally,

liver disease, immune compromise, andmedical or nonelective admissions arenoted as potential risk actors or theoccurrence o low BG (79). Physiologicchanges increase the eect o insulin asrenal ailure prolongs the hal-lie o insu-lin, leading to insulin accumulation, whilealso attenuating renal gluconeogenesis.


9/26


Figure 2. Forest plot o neurological mortality (hospital or 28-day) (1, 14, 16, 19, 23, 60, 65, 68, 7175). CI, condence interval;MH, Mantel-Haenszel.

Hepatic ailure can also lead to reducedhepatic gluconeogenesis. The reliability othe insulin inusion therapy protocol andrequency o BG monitoring also appear toinfuence the requency o hypoglycemia.

Multivariate regression models dem-onstrate that even a single episode o SHis independently associated with higherrisk o mortality (8085). The OR ormortality associated with one or moreepisodes was 2.28, 95% CI [1.41, 3.70];(p = .0008) among a cohort o 5,365patients admitted to a single mixed medi-

calsurgical ICU (82). Most other reportssimilarly indicate a higher risk o mortal-ity with hypoglycemia o varying severity(Table 4). Early hypoglycemia has beenassociated with longer adjusted ICU LOSand greater hospital mortality, especially

with recurrent episodes (86). Further-more, patients with more severe degreeso hypoglycemia sustained higher ICUand hospital mortality (85, 86). A greaterrisk o mortality (RR 2.18, 95% CI [1.87,2.53]; p < .0001) was similarly reported

with mild to moderate hypoglycemia

(BG 5569 mg/dL) in a post hoc analy-sis o prospective data collected in arandomized trial and two large cohorts(87). These data conrmed the results oanother cohort study that demonstratedthat mildmoderate hypoglycemia, BG5463 mg/dL, was independently asso-ciated with increased risk o mortality(85). In each o these studies, the mor-tality risk was greater with more severehypoglycemia (85, 87). Finally, the Leu-

ven investigators have recently publisheddata pooling the two interventional adult

trials to analyze the independent eectso hypoglycemia and glycemic variability(GV) on the risk o mortality (88). Theoccurrence o one or more episodes oSH was independently associated with ahigher risk o mortality (OR 3.233, 95%CI [2.251, 4.644];p < .0001).

Morbidity impact o SH is dicult toquantitate on critically ill patients as con-current illness and sepsis may increasethe risk o cognitive impairment, and it isunknown how hypoglycemia may inter-act with other risk actors. Low BG levels

lead to nonspecic neurologic symptoms,although severe or prolonged glycopeniamay produce neurocognitive impairment,seizures, loss o consciousness, perma-nent brain damage, depression, and death(8991). A number o actors includingsedation, medication, or underlying dis-ease may mask symptoms o neuroglyco-penia. To urther complicate the analysis,hyperglycemia has also been associated

with adverse eects on the brain (92). Fur-ther, the risk or neurologic injury may becompounded by additional oxidative stress

associated with rapid correction o hypo-glycemia with IV dextrose (93).

4. How should insulin-inducedhypoglycemia be treated in adult ICUpatients?

We suggest that BG < 70 mg/dL( 70 mg/dL with a goal toavoid iatrogenic hyperglycemia.

[Quality o data: very low]Although prevention o hypoglycemia

is important during insulin therapy, epi-sodes o low BG may occur despite rea-sonable precautions, and steps should betaken to recognize and treat it promptly.

With severe hypoglycemia, interruption othe insulin inusion is a prudent rst step.

This interruption may be adequate or apatient receiving exogenous dextrose, buttreatment with additional IV dextrose istypical, although there is no adequate datato dictate the optimal dose. While the rstpriority is patient saety through restora-tion o normoglycemia, rebound hyper-glycemia due to excessive replacementshould also be avoided, especially becausethe resulting increase in GV may contrib-ute to adverse outcomes (82, 83, 88, 93).

An IV dextrose dose o 1520 g has beenrecommended by the American Diabetes

Association, with instructions to recheckBG in 515 mins and repeat as needed(7). A dose o 25-g IV dextrose adminis-tered to nondiabetic volunteers producedsignicant but variable BG increases o162 31 mg/dL and 63.5 38.8 mg/dL

when measured 5 and 15 mins postinjec-tion, respectively (94). BG returned tobaseline by 30 mins, but the duration maybe dierent in patients receiving exog-enous insulin.

A ormula to calculate a patient-specic dose o dextrose has been used


10/26


Table 3. Summary o clinical trials in neurological patients

Author

Study Quality No o Patients Findings

Comments

Neuro-intensive

Care Unit

Population Design/End Point

Design

Assessment Control

Tight

Glycemic Control

Hospital Mortalitya

Odds Ratio [95%

Condence Interval]

Subsets o RCTVan den

Berghe

et al (1)

Mixed,

surgical,

mechanicalventilation

Randomized

80110 mg/dL vs.

180200 mg/dLResearch RN titrated

insulin per protocol

Single center

Evaluated

mean morningglucose

30 33 0.73 [0.21, 2.48] Control glucose

elevated with IV

dextrose, stoppedearly or benet

Van den

Berghe

et al (14)

Mixed,

medical

Randomized

Bedside RN titrated per

paper protocol

Single center

Evaluated

mean morning

glucose

31 30 1.05 [0.35, 3.15] Control glucose

elevated with IV

dextrose

Small RCT or subset o small RCTb

Scott et al

(68)

CVA Randomized

Fixed dose glucose-

potassium-insulin vs.

saline inusion

or 24 hrs

Single center,

Evaluated

glucose

trajectory over

treatment

period

28 25 28-day mortality

0.97 [0.29, 3.22]

No dierence in

serum glucose at

any point studied

Walters

et al (71)

CVA Randomized

Target 90140 mg/dL vs.

standard management

Single center

Evaluated

glucosetime

curve AUC

12 13 3.00 [0.11, 80.95] AUC reduced

Bilotta

et al (60)

Aneurysmal

subarachnoid

hemorrhage

Randomized

Target 80120 mg/dL vs.

80220 mg/dL

Single center,

evaluated

percentage o

glucose values

in target range

38 40 Six-month

mortality

0.78 [0.24, 2.58]

83% o control

and 69% o

intensive therapy

in target range

Gray

et al(72)

CVA Randomized, glucose-

potassium-insulin

inused to target

72126 mg/dL vs.

saline control

Multicenter,

evaluated glu-

cose every 8 hrs

using repeated

measures analy-

sis o variance

469 464 90-day mortality

1.14 [0.86, 1.51]

Average dierence

in glucose 10 mg/

dL (p < .001)

Arabi et al

(19)

Traumatic

brain injury

Randomized

80110 mg/dL vs.

180200 mg/dL

Bedside RN titrated per

paper protocol

Single center,

Evaluated aver-

age glucose level

39 55 1.43 [0.13, 16.39]

Bilotta

et al (73)

Traumatic

brain injury

Randomized, Target

80120 mg/dL vs.

80220 mg/dL

Single center,

evaluated mean

glucose values

49 48 1.02 [0.24, 4.35] Mean glucose

values 97 vs.

147 mg/dL

(p < .0001)Bilotta

et al (74)

Mixed, neuro-

surgical

Randomized, Target

80110 mg/dL vs.


11/26


Figure 3. Forest plot o severe hypoglycemia (1, 1419, 25, 26, 29). CI, condence interval;MH, Mantel-Haenszel.

in several reports (50% dextrose dose ingrams = [100 BG] 0.2 g), and it typi-cally advises administration o 1020 go IV dextrose, an amount lower than thatin traditional dosing methods (95, 96).This approach corrected the BG into thetarget range in 98% within 30 mins orpatients who had received IV insulin inu-sions (95, 97). Similarly, titrated replace-ment has been advocated or treatment oadults in the prehospital setting. Adminis-tration o 5-g aliquots o dextrose repeatedevery minute, using either 10% (50 mL)or 50% (10 mL) dextrose, restored mentalstatus to normal in approximately 8 mins

with both agents (IQR 515 and 411,respectively), but the 50% dextrose group

received a larger median dose o dextrose,25 g (IQR 1525) vs. 10 g (IQR 1015), anddeveloped a higher median posttreatmentBG (169 mg/dL vs. 112 mg/dL [p = 003]),respectively (98). The authors recom-mended titrating 10% dextrose in 50-mLIV (5-g) aliquots to treat the symptoms ohypoglycemia and to avoid overcorrec-tion o BG. The rate o administration oconcentrated dextrose solutions may alsobe important, as a report o cardiac arrestand hyperkalemia was associated withrapid and repeated administration o 50%

dextrose (99).A prehospital study comparing an

intramuscular 1-mg injection o gluca-gon to a 25-g IV dose o dextrose demon-strated a rapid and potentially excessiveBG response with dextrose, achieving14170 mg/dL increase in BG in the rst10 mins (100). The glucagon response wasslower, achieving a nal BG concentrationo 167 mg/dL ater 140 mins. Because vir-tually all ICU patients have venous access,IV dextrose is preerred over glucagon,due to the delay in glucagon response,

although additional testing o this inter-vention appears warranted.

Oral dextrose replacement (15 g) isused in ambulatory patients with hypogly-cemia, but is not tested or ICU patients.Fiteen grams o oral carbohydrate pro-duced a BG increase o approximately38 g/dL within 20 mins and providedadequate symptom relie in 14 0.8 minsin hypoglycemic adult outpatients (101).I oral replacement is used, dextrose orsucrose tablets or solutions are preerredor a more rapid or consistent responsecompared with viscous gels or orange

juice due to variable carbohydrate contentin commercial juice (101). The impact oabnormal gastric emptying has not been

studied but may alter the response totherapy, especially in an ICU population.

5. How oten should BG be monitored inadult ICU patients?

We suggest that BG be monitoredevery 12 hrs or most patients receivingan insulin inusion.

[Quality o evidence: very low]This is a consensus recommendation

based on limited data, as this question hasnot been tested in a prospective ashion.

The optimal requency o BG testing hasnot been established. Published protocolsgenerally initiate insulin therapy withhourly BG testing, and then may liber-alize the testing to every 4 hrs based onthe stability o the BG values within thedesired range, as well as an assessment opatient clinical stability. The personneltime required or BG monitoring is theprimary barrier to more requent moni-toring. We suggest that unstable patients(e.g., titrating catecholamines, steroids,changing dextrose intake) should have

BG monitored at least every hour to allowrapid recognition o BG outside the goalrange. More requent reassessment isneeded ater treatment o hypoglycemia,every 15 mins until stable.

A retrospective evaluation o datarom 6,069 insulin inusion episodes in4,588 ICU patients suggested that delaysin measuring BG contributed to the risko severe hypoglycemia. When a hypo-glycemic episode occurred, the mediandelay past the next hourly measurement

was 21.8 mins (IQR 12.229 mins) (97).Modeling suggested SH was likely with aslittle as a 12-min delay in the majority opatients who developed hypoglycemia.

Glucose checks every 4 hrs have been

used in some protocols; however, thereis a risk o unrecognized hypoglycemia

with prolonged measurement intervals;so these intervals are not recommendedas a routine component o insulin inu-sion protocols. The rates o hypoglyce-mia are above 10% or many protocolsusing BG checks every 4 hrs (1, 14, 15,17). One exception was reported with acomputerized protocol that tested anaverage o approximately six BG valuesper day but produced SH in only 1%o patients (102). With the higher rate

o hypoglycemia reported with every4-hourly BG testing, this requency isnot suggested unless a low hypoglycemiarate is demonstrated with the insulinprotocol in use.

6. Are POC glucose meters accurate orBG testing during insulin inusiontherapy in adult ICU patients?

We suggest that most POC glucosemeters are acceptable but not optimal orroutine BG testing during insulin inusion


12/26


Table 4. Clinical trials reporting rate and impact o hypoglycemia on outcome o critically ill patients

Author/Reerence Design n Population Results

Vriesendorp

et al (81)

Retrospective

cohort

156 (245 events)

155 control

Glucose


13/26


therapy. Clinicians must be aware opotential limitations in accuracy o glu-cose meters or patients with concurrentanemia, hypoxia, and interering drugs.

[Quality o evidence: very low]The use o glucose meters has become

common in hospitals due to their easeo use, availability, and ability to pro-

vide rapid results. Unortunately, in the

limited testing that has been reported,many o these devices lack accuracywhen used in critically ill patients. How-ever, insulin inusion therapy would beimpossible without some type o POCtesting methodology. The initial studyby Van den Berghe et al (1) on intensiveinsulin therapy used a precise arterialblood gas instrument or BG testing.Later trials have used a variety o POCdevices. One possible explanation or thegenerally unavorable results in subse-quent trials may be due to inappropriateinsulin dosing in response to inaccurate

BG results.Studies examining the accuracy o

POC glucose meters compared with areerence laboratory methodology oplasma glucose measurement reportedsignicant variability and bias betweenthese testing methods (103). Cliniciansmust be aware o the limitations with thespecic device used. Comparing data onspecic meters may be conounded by alack o consensus on the limits o accept-able error between the Food and Drug

Administration (allows up to 20% error)

and the American Diabetes Association(up to 5% error) standards. The Clini-cal and Laboratory Standards Instituteand International Organization or Stan-dardization 15197 guidelines allow upto 15 mg/dL variance or BG < 75 mg/dLand up to 20% o the laboratory analyzer

value or BG 75 mg/dL (104). The Clini-cal and Laboratory Standards Institutesuggests that a correlation above .9751is indicative o equivalence to the labora-tory standard (105). Simulation has sug-gested that meter error exceeding 17%

may double the number o potentially sig-nicant errors in insulin administrationand result in a higher risk o hypoglyce-mia (106). While meters have generallybeen considered acceptable within theusual ranges o BG testing (80200 mg/dL), additional laboratory testing o bloodsamples at the extremes o BG concentra-tion is needed to detect potential errorsand avoid over- or under-treatment withinsulin. The logistics o obtaining timelycentral laboratory measurement andreporting can be overwhelmingleading

to delays that could add signicant risk toecient insulin titration.

The methodology used by a POC meter(glucose oxidase vs. glucose-1-dehydro-genase) will impact the accuracy and thepotential or intererence by patient phys-iology, other circulating substances, andsample source. These have been reviewedelsewhere, but some specic actors are

pertinent to the ICU (107). For example,high Po2

(>100 mm Hg) can alsely lowerBG readings on POC meters that use glu-cose oxidase methods (108, 109).

Hematocrit (Hct) is an importantvariable or POC glucose testing in criti-cally ill patients. Most POC meters areapproved or BG measurement withina Hct range o 25%55%, but low Hcthas repeatedly been shown to alter theaccuracy o BG results with a POC meter.Lower Hct values generally allow metersto overestimate BG values, potentiallymasking hypoglycemia (110114). There

are no real-time alerts on meters to directclinicians to use other methodologiesin the ace o low Hct, although newermeters minimize Hct intererence by cor-recting abnormal values (115, 116). A or-mula may be applied to correct a meterBG value with low Hct (117). Newer glu-cose meters appear to have addressed thelimitations o older meters (118).

Drugs such as acetaminophen, ascor-bic acid, dopamine, or mannitol, along

with endogenous substances such as uricacid or bilirubin, may interere with the

accuracy o POC meters, especially thosemeters using glucose oxidase methodol-ogy (119). The direction o intererence onBG values depends on the device and theinterering substance. Glucosedehydro-genase-based assays are sensitive to inter-erence and alse elevation o results i thepatient receives medications containingmaltose (e.g., immune globulins) or ico-dextrin (e.g., peritoneal dialysis solutions).

An alternative POC method with acartridge-based amperometric methodis available or whole blood testing and

has been tested in critically ill popula-tions (120). There are ew limitations tothese cartridge-type devices using glucoseoxidase technology with the exception oknown intererence rom hydroxyureaand thiocyanate (121). A checklist orevaluation o POC glucose devices hasbeen published to improve the quality odevice evaluation (122).

7. When should alternatives to nger-stick capillary sampling be used inadult ICU patients?

We suggest arterial or venous wholeblood sampling instead o nger-stickcapillary BG testing or patients in shock,on vasopressor therapy, or with severeperipheral edema, and or any patient ona prolonged insulin inusion.

[Quality o evidence: moderate]Finger-stick capillary BG measurement

is typical when using a meter, although as

discussed, meters may introduce errorand bias in the BG value. Studies (Table 5)have compared BG in simultaneous sam-ples drawn rom dierent sites in criticallyill patients (105, 123135). These are di-cult to compare due to the dierences inreporting, testing methodology, and com-parators. O importance to clinicians isthat meter perormance deviated rom lab-oratory control by >20% in some reports,regardless o the blood source (130).

Samples rom an arterial site are mostsimilar to laboratory plasma or bloodgas analyzer BG values in paired sam-

ples. Venous specimens are also gener-ally acceptable, as long as care is takento avoid contamination o the specimenrom IV fuid inusing through a multilu-men catheter.

Finger-stick capillary glucose levelsmay provide signicantly dierent resultscompared with arterial or venous speci-mens when patients have low perusion

with hypotension, edema, vasopressorinusion, or mottled appearance o theskin (105, 124, 127130, 132). Hypoper-usion may increase glucose extraction

and increase the dierence between cap-illary whole blood and venous or arterialplasma glucose. Unortunately, there isno consistent pattern to the variability, asnger-stick testing BG results might belower or higher than arterial or venoussamples. Each institution should evaluatethe perormance o their selected meter ina variety o patient groups.

A sampling site hierarchy that priori-tizes arterial or venous sampling shouldbe established or BG monitoring o criti-cally ill patients. Devices that minimize

blood waste with catheter sampling areimportant to minimize the risk o ane-mia induced by requent phlebotomy.Finger-stick testing is invasive and otenpainul or patients who need requent BGmeasurements, and thus it should be thesite o last resort or avoided completely ithe patient is on vasopressors or exhibitshypoperusion.

8. Can continuous glucose monitoringreplace POC methods or critically illpatients?


14/26


Table 5. Summary o clinical trials evaluating the use o glucose meters on blood rom multiple sites or comparison o accuracy in various patient populations

Author Device Methodology Population Arterial POC vs. Laboratory

Cook et al (124) SureStepFlexxa

Single channel

GO

vs. serum in laboratory

67 ICU patients

67 samples

Glucose 62218 mg/dL

Hct 22%46.2%

Peripheral edema rated

NA

Finkielman et al

(125)

SureStepFlexxa

Single channelGO

vs. plasma in laboratory

197 ICU patients

816 samplesRetrospective data analysis

Arterial and venous POC

Mean dierence 7.9 17.6 mg/dLLOA +43.1, 27.2

Lacara et al

(126)

SureStepProa

GO

vs. laboratory (plasma or whole blood not

specied)

49 ICU patients

49 samples

Glucose 58265

Hct 31.7 0.8 sem

Arterial and venous POC

Bias 0.6

Precision 11.0 (p = .69)

Atkin et al (127) Accu-Chek II

GD

vs. serum in laboratory

25 hypotensive patients

39 normotensive patients

Glucose 52485

NA

Desachy

et al (135)

Accu-Chek

GD

vs. laboratory assay (plasma or whole blood

not specied)

103 patients

273 samples

Glucose 56675 mg/dL

Arterial and venous POC7% dier-

ent rom laboratory by >20%

LOA 42.4, 39.5

Kulkarni

et al (128)

Accu-Chek Advantage

GD

vs. arterial whole blood gas analyzer

54 ICU patients

493 samples

Glucose 37.742.5 mg/dL

Capillary vs. arterial blood gas analyzer

NA

Karon

et al (129)

Accu-Cheka comort curve

GD

Result is actored to agree with plasma results

vs. plasma in laboratory

20 coronary artery bypass grats patients

14 on pressors, none with systolic blood

pressure


15/26


Venous POC vs. Laboratory Capillary POC vs. Laboratory

Venous POC vs.

Capillary POC Conounders

Bias 9.51 mg/dL

Precision 8.44 mg/dL

21% samples >20 mg/dL dierence

LOA +26.5, 10.3

R2 = .288,p < .001

Bias 9.54 mg/dL


15% samples >20 mg/dL dierence

LOA +31.5, 12.5

R2 = .280,p = .02

Bias 0.03 mg/dL

No signicant

dierence between

samples

LOA +24.1, 24.0

Venous vs. nger stick No

signicant dierence

Low Hct contributed to

dierence between POC and

laboratoryNA NA NA Overall agreement, but

potential error or individualsamples

NA Bias 2.1 mg/dL


p = .23

NA Low Hct and Pco2

contributed

to glucose over prediction

Control: mean value 95.8% 1.1% o

laboratory value

Hypotension: 99.2% 2.5%

p < .05 vs. laboratory value

Control: mean value 91.8% 1.6% o

laboratory

Hypotension: 67.5% 5.7%,p < .001 vs.

laboratory

32% incorrectly diagnosed as hypoglycemic

NA Mean value rom dierent

methods were dierent

(p < .05)

NA 15% dierent rom laboratory by >20%

LOA 58.3, 55.3

NA Perusion index rom Phillips

monitor identied patients

with poor correlation

NA Bias 2.15 mg/dL


LOA 29.8, 2.5

Hypoperfusion: subset 75 samples

Bias 4.0Precision 16.2 mg/dLLOA 36.9, 28.4

NA Adequate agreement unless

patient has systolic blood

pressure 160 mg/dL with all methods

(p < .001). No report on vaso-

pressor eectNA Overall: 56.8% agreement

Vasopressor: 61.1% agreement overall, 25%

with glucose < 80 mg/dL

Edema: 55.8% agreement, 23.8% with

glucose


16/26


In the absence o compelling data,no recommendation can be made or oragainst the use o continuous glucosesensors in critical care patients.

[Quality o evidence: very low]The saety and potentially the eective-

ness o insulin inusion therapy could beimproved with more requent or continu-ous glucose measurement. Ultimately, a

closed-loop system (articial pancreas)could be used to titrate insulin inusiontherapy and minimize glucose variability,as it has been demonstrated to be easible(136). Continuous glucose sensors havebeen developed to measure interstitialand intravascular glucose concentrations,and this technology has been reviewed(137139). However, intravascular devicesremain in preclinical and limited clinicaltesting (136).

Interstitial measurement devicesmay be subject to the same limitationsas nger-stick BG testing, related to

variable tissue perusion, temperature,and local humoral actors in addition todelays related to glucose equilibration,and need or calibration. Initial reportso continuous interstitial glucose sensorshave demonstrated acceptable accuracyin select patients (133, 140143). Con-current norepinephrine inusion did notalter the accuracy o continuous SQ glu-cose monitoring (140). Additional evalua-tion o accuracy and utility o continuousmonitoring in broad patient populationsis needed beore these devices can be rec-

ommended or routine use. In studies opediatric postoperative cardiac surgicalpatients and pediatric medical/surgicalICU patients, correlation o continuousinterstitial glucose monitors with BGreadings is acceptable (i.e., mean absoluterelative dierence o 17.6% and 15.2%)and unaected by inotrope use, body tem-perature, body wall edema, patient size, orinsulin use (144, 145).

9. How should IV insulin be prepared andadministered?

We suggest continuous insulin inu-sion (1 unit/mL) therapy be initiated aterpriming new tubing with a 20-mL waste

volume.[Quality o evidence: moderate]Titration o insulin therapy to an

end point o tight GC requires the rapidresponse and immediate fexibility o acontinuous inusion. These inusionsshould be prepared in a standardized con-centration, with most protocols report-ing use o a 1 unit/mL solution o human

regular insulin, although 0.5 unit/mLsolutions may also be ound in the lit-erature. Insulin may be mixed with 0.9%sodium chloride, lactated Ringers injec-tion, Ringers injection, or 5% dextrose.Insulin may be prepared in glass or plas-tic containers (polyvinyl chloride [PVC],ethylene vinyl acetate, polyethylene, andother polyolen plastics), although loss

will occur through adsorption to contain-ers and to IV tubing and lters. Adsorptionis immediate upon contact, producing abioavailability o approximately 5060%in PVC with sustained stability or 168hrs (146). Factors such as storage tem-perature, concentration, and inusion rateinfuence the extent o adsorption. A trialo various priming volumes o 1050 mLconcluded that a 20-mL prime rom a100-mL polyvinyl chloride bag contain-ing regular insulin, 1 unit/mL, producedinsulin delivery through a 100-inch latex-ree polypropylene IV inusion set that was

not statistically dierent rom a 50-mLpriming volume (147). This maneuvershould be repeated each time new tubingis initiated to maintain consistent insulindelivery rates. The optimal priming vol-ume or syringe pump systems has notbeen reported.

Accurate insulin administration strat-egies include use o a reliable inusionpump or insulin administration, ideally

with saety sotware that prevents inad-vertent overdosing. The pump must beable to deliver insulin dose increments o

200 mg/dL triggered initiation o aninsulin inusion according to the nurse-managed protocol. Long-acting insulin

was added to the SQ regimen when ea-sible and appropriate. The patient-specictreatment protocol combining SQ andIV insulin regimens demonstrated saetyand ecacy in maintaining the BG con-centration predominately within the goalrange with excursions o BG > 180 mg/dL in 0.5 unit/hr, or stresshyperglycemia patients receiving insulininusion at a rate o >1 unit/hr (152156).

However, transition to SQ insulin

should be delayed until there are noplanned interruptions o nutrition orprocedures, until peripheral edema hasresolved, and until o vasopressors. Aprotocol or transition leads to better glu-cose control than nonprotocol therapy(157). Failure o SQ regimens to pro-duce or maintain GC (BG < 180 mg/dL)should trigger redesign o the regimen orresumption o insulin inusion therapy.

A retrospective review o 614 cardio-thoracic patients determined the eec-tiveness o an IV (in the ICU) ollowed by


17/26


SQ (outside the ICU) regimen on mor-bidity and mortality (158). The authorsound the SQ regimen to be less nursing-intensive and less costly in all patients,but only those with a preexisting diagno-sis o diabetes demonstrated signicantlylower rates o postoperative mortality.Protocolized transition to an SQ regi-men has been shown to decrease rebound

hyperglycemia ater inusion discontinu-ation (159).

B. We suggest that calculation obasal and bolus insulin dosingrequirements should be based onthe patients IV insulin inusionhistory and carbohydrate intake.

[Quality o evidence: very low]Several models have been proposed

or transition rom insulin inusion toSQ insulin therapy (156, 158161). Themajority o these models include a three-

component approach to insulin replace-ment: basal insulin, nutritional insulin,and correction insulin. Basal insulin isprovided as an injection o long-actinginsulin given every 24 hrs (e.g., glargine)or intermediate-acting insulin given every612 hrs (e.g., NPH). Basal insulin willbe needed in many diabetic patients onenteral eedings to achieve the desired BGgoal (162). The initial basal insulin doseis recommended at least 24 hrs beorestopping the insulin inusion when pos-sible to prevent rebound hyperglycemia

(28, 163). I this overlap is not easible,a simultaneous injection o rapid-actinginsulin (approximately 10% o the basaldose) may be given with the basal insu-lin injection when stopping the inusion(156). One group suggests calculating atotal daily dose (TDD) o IV insulin romthe mean hourly dose or at least the prior6 hrs as a guide to the basal insulin dose(28). As IV insulin delivery is reducedby adsorption to the container and tub-ing, the authors reduced the initial basaldose to 80% o the estimated TDD and

achieved their target or glucose controlmore readily than using smaller percent-ages o the TDD, although others haveshown acceptable glucose control using60%70% o the TDD (156, 158, 159,164). It is important to consider concur-rent changes in other drug therapy ornutritional regimens when planning atransition regimen.

Mixing insulin in a parenteral nutri-tion (PN) solution can replace a separateinsulin inusion or basal insulin injec-tions once the daily requirements are

stabilized. Additional correction doses canbe given to ne-tune GC every 36 hrs.

12. What are the nutritional consider-ations with IV insulin therapy inadult ICU patients?

A. We suggest that the amount andtiming o carbohydrate intakeshould be evaluated when calcu-

lating insulin requirements.B. We also suggest that GC proto-

cols should include instructionsto address unplanned discontinu-ance o any orm o carbohydrateinusion.

[Quality o evidence: low]Nutritional support requirements o

critically ill patients vary and are beyondthe scope o this discussion. Guidelinesor nutritional support o critically illpatients are available (165).

Consistent intake o nutrition appearsto simpliy glycemic management dur-ing an insulin inusion. Overeeding mayproduce hyperglycemia that necessitatesinsulin inusion therapy, and should beavoided.

Provision o 200300 g o dextrose perday was a component o the initial trialby Van den Berghe et al (1) in surgicalICU patients. The reduction o mortalityreported with achievement o BG valueso 80110 mg/dL has been suggested torefect minimization o complicationsrom PN, although similar calories were

provided in the medical ICU study, with-out the same impact on outcome (14).

While a meta-analysis o clinical trials,stratied by source o calories, suggestedthat tight GC is potentially more ben-ecial during PN regimen compared withenteral eeding, this was not conrmed ina prospective trial comparing early vs. latePN (33, 166). Tight GC (mean BG 100110mg/dL) was similarly achieved in patients

who received 34 g/kg/d o carbohy-drate (early) compared with 0.52 g/kg/d(late) over the rst 7 ICU days (166). The

patients on early PN required higher totalinsulin doses per day but ewer patientshad SH (2% vs. 3.5%,p = .001). Neverthe-less, the patients on late PN (who receivedcarbohydrates rom enteral nutrition and5% dextrose inusion or the rst week)had better overall outcomes. Thus, insu-lin inusion appears to be suitable orpatients regardless o the source o car-bohydrates, and GC alone is not enoughto reduce the apparent risks associated

with PN. The enteral route is preerredover the parenteral route or nutrition

support in the ICU setting when possible(165). However, due to several actorscommon to the ICU (e.g., gastric stasis,interruption o enteral nutrition or tests/procedures, and anatomical anomalies),the amount o eeding that can be deliv-ered enterally is generally less than theamount delivered parenterally. Interrup-tion o enteral eeds was ound to cause

the majority o the hypoglycemic events(62%) in the Leuven MICU trial, and simi-lar results were noted elsewhere (13, 167).Initiation o a 5% dextrose-containing IVsolution at the same rate as the discon-tinued enteral eeding solution appearsto prevent hypoglycemia (168). Dextrose(10%) solutions may be used to minimizethe volume o ree water.

Integration o an insulin protocolwith nutritional intervention has beensuggested to achieve a high level o GC.The Specialized Relative Insulin NutritionTables protocol titrates both eeding and

insulin doses to achieve tight glucosecontrol and was more eective at achiev-ing the BG target than a retrospectivecontrol (169, 170). Insulin was admin-istered with hourly bolus injections andcould be supplemented by an inusion oup to 6 units/hr. The rate o enteral eed-ing was also adjusted to acilitate GC, butresulted in delivery o only 50% o thepredicted caloric requirement, and thusmay not be an optimal long-term nutri-tional strategy.

Bolus doses o IV insulin may be

administered or nutritional insulin ther-apy during an insulin inusion when car-bohydrates are delivered intermittently,based on carbohydrate ratio, as previouslydiscussed. Consistent oral intake shouldtrigger transition to SQ insulin therapyand a consistent carbohydrate diet plan.Glucose monitoring should be scheduledto avoid measurement o postprandial BGconcentrations.

13. What actors should be consideredor sae insulin therapy programs in

the adult ICU?

We suggest that insulin is a high-riskmedication, and that a systems-basedapproach is needed to reduce errors.

[Quality o evidence: very low]Insulin is a high-alert, high-risk medi-

cation due to the risk o hypoglycemia,complexity o therapeutic regimens,and availability o multiple products inpatient-care areas. It is in the top vehigh-risk medications that account orabout one third o all major drug-related,


18/26


injurious medication errors. One analy-sis indicated that 33% o errors caus-ing death within 48 hrs involved insulintherapy (171). Strategies to reduce sucherrors have been suggested and shouldbe applied to the ICU setting (172). Theseinclude standardized protocols or insu-lin dosing and monitoring, computer-ized provider order entry, minimizingavailable insulin products, avoidance oabbreviations such as U or units, stor-ing insulin away rom other medications,and detailed multiproessional analysis oactual errors and near-miss events. Strat-egies to improve insulin saety includemandating an independent double-checko doses, requent BG monitoring, andprominent product labeling.

The limitations o BG monitoringequipment and methodology may alsoincrease the risk o error. For example,actitious elevations in BG occur whenicodextrin peritoneal dialysis solutions

or maltodextrin-containing medications(selected immune globulin products) areadministered and monitored with a glu-cose dehydrogenase monitoring system(173). Also, a dextrose solution adminis-tered via a pressurized fush system pro-duced actitious elevations in BG valuesdrawn through an arterial line, and sub-sequent inappropriate insulin administra-tion led to atal neuroglycopenia (174).

Saety in insulin administration meth-odology is also important, and a systems-based approach is needed to reduce

insulin errors. Complex insulin therapyprotocols with multiple patient-specicexceptions and the need or a high-leveltraining or accurate use are common.

A standardized protocol should be utilizedonly ater adequate education and pro-cesses are implemented to monitor out-comes. Routine and requent assessmento glucose metrics, as will be described,should be perormed. Failure to achieveadequate glucose control or requentepisodes o hypoglycemia should triggerrapid reassessment o the protocol and

monitoring system.

14. What are the characteristics o anoptimal insulin dosing protocol orthe adult ICU population?

We suggest that ICUs develop a pro-tocolized approach to manage GC.Components include a validated insu-lin administration protocol, appropri-ate stang resources, use o accuratemonitoring technologies, and a robustdata platorm to monitor protocol peror-mance and clinical outcome measures.

A standard insulin inusion protocolshould include a requirement or con-tinuous glucose intake, standardized IVinsulin inusion preparation, a dosingormat requiring minimal bedside deci-sion-making, requent BG monitoring,provisions or dextrose replacement ieedings are interrupted, and protocolizeddextrose dosing or prompt treatment o

hypoglycemia.[Quality o evidence: very low]A standard protocol or insulin admin-

istration and monitoring is essential orconsistency and saety. Comparison oexisting protocols is dicult due to sig-nicant dierences in processes and out-come measures, but key eatures will bediscussed.

Computerized decision-support sys-tems achieved better glucose control thanthat achieved with paper-based systemsusing ithen decision model (175).

Although paper-based systems may be

adequate, they may be more complex andtime-consuming and lack a reminder sys-tem to ensure timely BG measurement.Most o the studies comparing protocolsemployed pre- and post-interventioncohort design, limiting the ability to con-clude i the new protocol was the cause oimproved results. However, several RCTsdemonstrated avorable eatures o com-puterized insulin inusion protocols vs.paper-based systems (148, 176178). Gly-cemic control metrics and hypoglycemiarates have been consistently better with

computerized protocols. Reminder alertslead to more consistent and timely BGassessments. Commercial systems havelicensing ees that may be a barrier toutilization, although several institutionshave developed custom computer-basedsystems (96, 97). The largest trial, NICE-SUGAR, had a computer-assisted proto-col, but dosing was based on a complexdecision tree, rather than a specic set oormulas (16). It should be noted that thisprotocol ailed to achieve an average BGlevel within the goal range o 80110 mg/

dL.Numerous cohort reports describe the

utility and eectiveness o paper-basedprotocols as they evolve over time, com-pared with historical controls (23, 151,152, 179182). The reports are o low qual-ity due to small study size, single-centerexperience, use o historical controls, and

variable outcome measures (includingsurrogate measures such as BG resultsrather than patient outcomes). These pro-tocols vary in insulin dosing intensity andcomplexity. Some contain insulin bolus

doses, and others require multiple stepsto alter insulin dosing, which can lead tomarkedly dierent insulin doses in a sim-ulated patient model (183).

The original protocol published by Vanden Berghe et al (1) (Leuven protocol) wasrelatively unstructured, although it wassuccessully administered in a researchsetting with trained providers. Subse-

quent use by bedside providers in otherICU settings has produced hypoglycemiarates that were deemed to be excessive(15, 17).

Advantages o paper-based protocolsinclude easy bedside access, insulin ratechanges are made only when outside ogoal BG ranges, and sometimes separatescales or diering levels o insulin sen-sitivity. The major disadvantages o theseprotocols include their complexity (withmultiple recommendations on the samepage), a lack o fexibility with majorclinical changes, and lag time to respond

to BG trends (may recommend a doseincrease or a persistently high BG, eveni the BG level has actually declined).

A more straightorward approach is touse an algebraic ormula to calculate theinsulin rate based on the BG and a multi-plier (M) that relates to insulin sensitivity(insulin dose [unit/hr] = [BG 60] M)(9597, 184). This calculation can becomputerized, assisted by a tabular or-mat, or calculated manually (185, 186).The multiplier increases or BG abovethe target range and decreases when

the BG is below the goal. Advantages tothis approach include rapid determina-tion o the new insulin dose without theneed or extensive judgment or trainingo the bedside caregiver and constanttitration based on the BG trend. It hasresulted in some o the lowest reportedrates o severe hypoglycemia (177, 182,183). Disadvantages include the need ora bedside computer and the potential orexaggerated increases in insulin inusionrate in response to an elevated BG value,especially with a high multiplier. The

multiplier may need to be reset to a lowervalue, especially ollowing a signicantchange in nutritional intake or change inclinical status. More sophisticated com-puterized protocols have also been devel-oped and have similarly been shown toperorm better than conventional proto-cols (169, 170, 178, 187, 188). Computer-ized programs can also collect data on theperormance o the program and calculatea variety o metrics.

A source o error with virtually all insu-lin protocols is incorrect transcription o


19/26


BG values into a reestanding computerprogram, which may occur approximately5% o the time (189). Similarly, protocol

violations are reported with paper-basedsystems (190). The amount o practitionerlatitude in deviating rom the protocolrecommendations should be predenedand evaluated as a component o qualityassurance programs.

With many published protocols avail-able, there is no need to reinvent the

wheel to implement an insulin inusionprotocol. The local barriers to sae insulintherapy must be identied and addressed,including availability o adequate andappropriate testing equipment, consid-eration o workorce impact, and a teamapproach to education and implemen-tation (191). Tight levels o BG controlshould not be attempted when a new pro-tocol is initiated, to minimize hypoglyce-mia risk during the initial learning curve.

Systematic and requent assessment oresults is needed. Feedback to providersis essential when protocol violations oradverse events occur. In addition, a proto-col is only eective i used in a consistentashion. Automatic triggers or protocolinitiation are more ecient than waitingor prescriber recognition o hyperglyce-mia and appropriate response throughpatient-specic orders.

Other keys to a successul glycemicmanagement program include the avail-ability o a reliable methodology or BG

testing, with an adequate number odevices to minimize delays and wastedtime obtaining the device. The data shouldbe recorded in the electronic medicalrecord promptly and be displayed along

with insulin dosing adjustments toassess protocol perormance and allowevaluation o variances. In addition,the glycemic management programshould be coordinated with nutritionsupport interventions to minimize therisk o hyperglycemia or hypoglycemia

with addition or interruption o nutri-tional intake. Concurrent medicationsdosed intermittently should be mixedin sodium chloride solutions to reduceglucose variation induced by episodicdextrose administration. While patientsshould receive a consistent carbohydrateintake, the need or insulin may be mini-mized by limiting the inusion o exces-sive quantities o dextrose solutions.

15. What is the impact o GV on out-comes o critically ill patients?

Glycemic variability has been indepen-dently associated with mortality in severalcohorts o critically ill patients; however,there is no consensus regarding the appro-priate metric or mathematically deningGV. We suggest that the simplest toolssdo each patients mean BG and coeciento variation (sd/mean)be reported in allpublished interventional studies.

[Quality o evidence: very low]Glucose metrics are important toevaluate the overall results o a GC pro-gram. In clinical trials, glucose variabilityhas been suggested as a better end pointto assess the impact o blood sugar onpatient outcome during insulin inusioncompared with other measures, such asmean morning BG, mean o all BG values,or time-weighted average value. Higherlevels o GV have been independentlyassociated with mortality in adult cohortso mixed medicalsurgical patients (83,86), surgical ICU patients (192), patients

admitted with sepsis (193), as well asin critically ill pediatric patients (194).However, the most appropriate metricto describe GV has not yet been dened.Relatively simple measures to calculate

variability include sd, coecient o varia-tion, and mean daily delta (maximum minimum BG). More complex measuresthat have been evaluated in dierent stud-ies include mean amplitude o glycemicexcursion, the glycemic lability index,maximal glucose change, and the vari-ability index (195, 196).

A recent review summarized thebiologic basis or the deleterious eecto increased GV (196). One purportedmechanism is the oxidative stress thatoccurs at the cellular level induced byrapid changes in the BG level (93, 197,198). Fluctuations in BG levels may leadto changes in serum osmolality that causeinjury at the cellular and organ levels(199). Finally, wide excursions may maskoccult hypoglycemia, which has beenrecognized as a risk actor or mortalityin the critically ill (82). It is not known

yet whether eorts to minimize GV willdecrease the mortality rate in criticallyill patients, but this remains a promisingavenue or uture research.

16. What metrics are needed to evaluatethe quality and saety o an insulininusion protocol and GC program inthe adult ICU?

Measures o overall glucose controlshould include mean (sd) and median(IQR) BG levels as well as ICU-level run

charts o percentage BG < 150 mg/dL and180 mg/dL. We suggest that hypoglycemicevents should be monitored regularly andreported as events per patient, as a per-centage o all BG values, and events per100 hrs o insulin inusion.

[Quality o evidence: very low]This is a consensus suggestion to

improve the saety and ecacy o GC and

insulin therapy. Data on the perormanceo an insulin inusion protocol shouldbe assessed multiple times throughoutthe year (e.g., at least quarterly). Poten-tial measures o protocol eectivenessinclude global measures o BG control,such as mean and median BG per patient,measures o glucose variability, and timeto specic end points, including mean andmedian time required to reach the des-ignated glycemic target as well as meanand median time spent within the desiredglycemic range, reported as a percentageo total time in range (200202). Patients

with diabetic ketoacidosis and hyper-glycemic hyperosmolar coma should beexcluded rom this analysis.

Protocol saety should be regularlyassessed through metrics relating tohypoglycemia, which should be dened assevere (


20/26


measure o the consistency o glucosecontrol (88). This tool scores glucose val-ues based on the degree o excursion romthe goal, making it a more dynamic mea-sure o the variability o glucose values ina single patient. A higher value indicatesewer values within the goal range. Thistool has been used to compare insulininusion protocols, but not to evaluatepatient outcome. The HyperglycemicIndex measures the area under the curveo BG values above the upper limit o thegoal range vs. time (204). This method hasshown a signicant association with mor-tality when used or retrospective analy-sis o BG values or surgical ICU patients.This metric is most meaningul when thedaily number o BG values is consistentrom patient to patient.

17. What are the economic and work-orce impacts o a GC program in theadult ICU?

A. We recommend that programs tomonitor and treat hyperglycemiain critically ill patients be imple-mented to reduce hospital costs.

[Quality o evidence: moderate]

B. We suggest implementation oprograms to monitor and treathyperglycemia in diabetic patientsollowing cardiovascular surgeryto reduce hospital costs.

[Quality o evidence: low]The cost implications o implemen-

tation o programs to monitor and treathyperglycemia in hospitalized patientshave been studied in a variety o dierentpatient populations. Complications asso-ciated with poor GC have the potential toincrease total hospital costs. A reductionin sternal wound inections was associ-ated with improved GC and producedlower costs (205). This single-centerinvestigation estimated that each 50 mg/dL increase in mean BG level was associ-

ated with an excess o $2,824 in the cost ohospitalization. Promulgation o a hospi-tal-wide inpatient diabetes managementprogram produced a reduction in LOSthat resulted in over $2 million in savingsto another acility (206). However, totalcost is not the only important measure othe impact o GC programs. Aragon eval-uated the nursing work burden imposedby an IV insulin protocol on our dier-ent ICUs within a single academic insti-tution (207). A mean o 4.7 (1.1) mins

was needed or each hourly analysis o

BG, which extrapolated to nearly 2 hrs onursing time each day or insulin inusionmanagement. The design o this observa-tional study did not include calculation ototal paid nursing hours. Another timemotion study noted a marked dierencein the time required or GC activities witha paper protocol, depending on clinicalurgency. Malesker et al (208) reported

a mean o 2.24 (1.67) mins rom BGto therapeutic action and 10.55 (3.24)mins or hyperglycemia, although multi-tasking by nurses makes discreet evalua-tion o this activity more challenging. Thecomplete time rom meter acquisition tocompletion o documentation might havebeen as long as 33 mins or adjustment oinusion therapy, and longer or inusioninitiation.

There are ew published studies o theeect o tight GC implementation on ICUcosts. Van den Berghe et al (209) per-ormed an analysis o the 1,548-patient

cohort rom their landmark surgicalICU study. The methodology consistedo a cost accounting o the componentso care ound to change signicantly asa result o intensive insulin therapy: thedirect cost o insulin administration, ICUdays, mechanical ventilation, and the useo vasopressors, inotropes, IV antibiotics,and blood transusion. The total savingsper patient associated with the intensiveinsulin protocol was $2,638 per patient.The mean LOS in the conventional treat-ment and intensive treatment groups was

8.6 and 6.6 days, respectively, accountingor over 80% o the cost per patient.

The cost implications o a 1,600-patientpre- and post-intervention cohort studyo tight GC were implemented in a mixedmedicalsurgical ICU o a university-aliated community teaching hospital(210). These investigators attempted toquantiy all major components o the costo care: ICU days, mechanical ventilationtime, laboratory testing, pharmacy, diag-nostic imaging, and days in the hospitalon the regular wards ater discharge rom

the ICU. The net savings per patient was$1,580. The 17% decrease in ICU meanLOS (rom 4.1 to 3.4 days) accountedor 28% o the savings, but there werealso substantial savings associated withdecreased use o mechanical ventilation,diagnostic imaging, laboratory testing,and days in the hospital ater dischargerom the ICU.

A third report rom Sadhu and col-leagues (211) used a dierence-in-di-erences (quasi-experimental) designto measure an association between a

multi-ICU glycemic management pro-gram and hospital and patient outcome

variables. The participating ICUs dem-onstrated a reduction in mean BG com-pared with nonparticipating units inthe hospital. Outcomes were comparedin the groups to address the impact osecular time trends and patient char-acteristics that might have altered the

results in this beore and ater study.The glycemic management protocol wasassociated with an average reductiono 1.19 days o ICU care per admission(p .05) and a trend toward lower mor-tality and resource use including a reduc-tion o $4,746 in total costs per patient($10,509 to $1,832).

18. What are the implications o hyper-glycemia in pediatric critically illpatients?

In the absence o compelling data, no

recommendations could be made or oragainst the use o tight GC in pediatriccritical care patients.

Hyperglycemia is highly prevalentin pediatric critical care. While studiesshow an independent association betweenhyperglycemia and morbidity and mortal-ity rates, the paucity o data has resultedin practice variability (194). As in adults,children develop critical illness hypergly-cemia with no history o premorbid diabe-tes or insulin resistance related to severityo illness. Although most pediatric inten-

sivists believe that hyperglycemia maycause harm in their patients and supportthe concept o avoiding hyperglycemia,most are reluctant to practice routineGC (212, 213). An RCT o 700 criticallyill pediatric patients was completed in asingle center in Leuven, Belgium, whichestablished that insulin inusion titratedto a goal o 5080 mg/dL in inants and70100 mg/dL in children, compared

with insulin inusion only to prevent BG>215 mg/dL, improved short-term out-comes (214). The absolute risk o mor-

tality was reduced by 3% (conventional5.7% vs. interventional 2.6%, p = .038),and insulin therapy also reduced the ICULOS and C-reactive protein (the primaryoutcome variable). The study was notableor its rst proo o principle that tighterlevels o GC produce clinical benet. It

was also remarkable or its low target BGranges in the intervention groups, which

were described as age-adjusted normo-glycemia (5080 mg/dL in those 1 yr old).

Although several outcomes in this trial


21/26


were avorable, there were extremely highrates o SH (


22/26


versus glycemic control. Crit Care Med2003;

31:359366

14. Van den Berghe G, Wilmer A, Hermans G,

et al: Intensive insulin therapy in the medical

ICU.N Engl J Med2006; 354:449461

15. Preiser JC, Devos P, Ruiz-Santana S, et al:

A prospective randomised multi-centre con-

trolled trial on tight glucose control by inten-

sive insulin therapy in adult intensive care

units: The Glucontrol study. Intensive Care

Med2009; 35:17381748

16. The NICE-SUGAR Investigators: Intensive

versus conventional glucose control in

critically ill patients. N Engl J Med 2009;

360:12831297

17. Brunkhorst FM, Engel C, Bloos F, et al;

German Competence Network Sepsis

(SepNet): Intensive insulin therapy and pen-

tastarch resuscitation in severe sepsis.N Engl J

Med2008; 358:125139

18. De La Rosa G, Donado JH, Restrepo AH, et al:

Strict glycemic control in patients hospital-

ized in a mixed medical and surgical intensive

care unit: A randomized clinical trial. Crit

Care 2008; 12:R120

19. Arabi YM, Dabbagh OC, Tamim HM, et al:Intensive versus conventional insulin ther-

apy: A randomized controlled trial in medical

and surgical critically ill patients. Crit Care

Med2008; 36:31903197

20. Farah R, Samokhvalov A, Zviebel F, et al:

Insulin therapy o hyperglycemia in intensive

care.Isr Med Assoc J2007; 9:140142

21. Grey NJ, Perdrizet GA: Reduction o nosoco-

mial inections in the surgical intensive-care

unit by strict glycemic control.Endocr Pract

2004; 10(Suppl 2):4652

22. Mackenzie IM, Ercole A, Blunt M, et al:

Glycaemic control and outcome in general

intensive care: The East Anglian GLYCOGENIC

study.Br J Intensive Care 2008; 18:12112623. Krinsley JS: Eect o an intensive glucose

management protocol on the mortality o

critically ill adult patients. Mayo Clin Proc

2004; 79:9921000

24. Furnary AP, Gao G, Grunkemeier GL, et al:

Continuous insulin inusion reduces mor-

tality in patients with diabetes undergoing

coronary artery bypass grating. J Thorac

Cardiovasc Surg 2003; 125:10071021

25. Treggiari MM, Karir V, Yanez ND, et al: Inten-

sive insulin therapy and mortality in critically

ill patients. Crit Care 2008; 12:R29

26. Krinsley JS: Glycemic control, diabetic status,

and mortality in a heterogeneous populationo critically ill patients beore and during the

era o intensive glycemic management: Six

and one-hal years experience at a university-

aliated community hospital.Semin Thorac

Cardiovasc Surg 2006; 18:317325

27. Scalea TM, Bochicchio GV, Bochicchio KM,

et al: Tight glycemic control in critically

injured trauma patients. Ann Surg 2007;

246:605610; discussion 610

28. Furnary AP, Wu Y: Eliminating the diabetic

disadvantage: The Portland Diabetic Proj-

ect. Semin Thorac Cardiovasc Surg 2006;

18:302308

29. Tot P, Jrgensen HS, Toennesen E, et al:

Intensive insulin therapy to non-cardiac ICU

patients: A prospective study.Eur J Anaesthesiol

2006; 23:705709

30. Schultz MJ, Harmsen RE, Spronk PE: Clini-

cal review: Strict or loose glycemic control

in critically ill patientsImplementing best

available evidence rom randomized con-

trolled trials. Crit Care 2010; 14:223

31. Wiener RS, Wiener DC, Larson RJ: Benets

and risks o tight glucose control in critically

ill adults: A meta-analysis. JAMA 2008; 300:

933944

32. Griesdale DE, de Souza RJ, van Dam RM,

et al: Intensive insulin therapy and mortality

among critically ill patients: A meta-analysis

including NICE-SUGAR study data. CMAJ

2009; 180:821827

33. Marik PE, Preiser JC: Toward understanding

tight glycemic control in the ICU: A system-

atic review and metaanalysis. Chest 2010;

137:544551

34. Ingels C, Debaveye Y, Milants I, et al: Strict

blood glucose control with insulin during

intensive care ater cardiac surgery: Impact

on 4-years survival, dependency on medicalcare, and quality-o-lie. Eur Heart J 2006;

27:27162724

35. Zerr KJ, Furnar

(inglés) guía infusión insulina para manejo hiperglicemia pctes críticos [2012]

Documents