Добавил:
Sekretar
kiopkiopkiop18@yandex.ru
t.me/Prokururor I Вовсе не секретарь, но почту проверяю
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз:
Предмет:
Файл:Ординатура / Хирургия / Библиотека им академика М.И. Перельмана / Книга_1134_Библиотеки_им_академика_М_И_Перельмана.pdf
28 Venous Thromboembolism in the Intensive Care Unit
337
Table 28.1 Risk factors: venous thromboembolism in the intensive
care unit
VTE risk factors commonly acquired in the ICU
Respiratory failure requiring mechanical ventilation
Cardiac failure (New York Heart Association Class III/IV)
End-stage renal disease
Sepsis, severe sepsis, and septic shock
Vasopressors
Pharmacologic sedation
Immobilization
Central venous catheters
Platelet transfusion
Thrombophilia (e.g., heparin-induced thrombocytopenia)
Other major VTE risk factors
Malignancy
Personal history of previous VTE
Family history of VTE
Prolonged surgical procedure (>2 h)
Major general surgery
Major traumatic injury
Hip or leg fracture
Hip or knee replacement
Acute spinal fracture
Acute spinal cord injury (<1 month)
Acute stroke (<1 month)
Pregnancy/postpartum (up to 6 weeks)
Known thrombophilia (e.g., factor V Leiden, lupus anticoagulant,
anticardiolipin antibodies, antithrombin defi ciency, protein C or S
defi ciency, etc.)
Other minor VTE risk factors
Older age
Immobility from prolonged sitting (e.g., airplane travel or
prolonged car travel)
Laparoscopic surgery
Infl ammatory bowel disease
Obesity
Pregnancy/antepartum
Acute infection
Varicose veins
Arteriovenous malformations
Tobacco use
Estrogen/selective estrogen receptor modulators (e.g., tamoxifen)
Contraceptives
VTE venous thromboembolism, ICU intensive care unit
American Academy of Orthopaedic Surgeons (AAOS),
respectively [ 26 , 27 ]. Evidence-based best practice VTE pro-
phylaxis in the ICU varies based on the primary service (e.g.,
medicine, surgery, etc.) and other patient-specifi c risk factors.
Pharmacologic Prophylaxis
Most protocols use subcutaneous (SC) injection of unfractionated heparin or low molecular weight heparins (LMWHs)
such as enoxaparin, dalteparin, or fondaparinux for VTE
prophylaxis. Trauma and orthopedic literature typically supports the use of LMWH over unfractionated heparin [ 26 ].
Patients with unstable renal function or creatinine clearance
less than 30 mL/min should receive unfractionated heparin
instead of LMWH due to risks associated with bioaccumulation of some LMWHs in patients with reduced renal clearance. In ICU patients, LMWH may be preferable to
unfractionated heparin. The PROTECT study was a randomized controlled trial comparing unfractionated heparin and
LMWH as VTE prophylaxis in ICU patients [ 28 ]. There was
no signifi cant difference in proximal DVT between the two
groups, but patients treated with LMWH had fewer PE
events. A recent meta-analysis pooled data from eight randomized controlled trials (including PROTECT) to evaluate
the use of LMWH versus unfractionated heparin prophylaxis
in ICU patients [
29 ]. This study concluded that LMWH was
preferred over unfractionated heparin for VTE prophylaxis
in ICU patients: the risk of any DVT, any PE, major bleeding, and/or mortality was decreased by 10 % among patients
receiving LMWH versus unfractionated heparin (RR 0.90,
95 % CI 0.83–0.97, p = 0.01). However, when looking at each
of these outcomes separately, LMWH was associated with a
signifi cantly decreased rate of DVT but no signifi cant difference in PE, major bleeding, or mortality.
Most protocols recommend VTE prophylaxis throughout
the inpatient hospitalization, but some literature supports
extending prophylaxis to the outpatient setting for a limited
duration after discharge from the hospital. This may be of
particular use in patients at high risk for perioperative VTE
including orthopedic surgery patients or those with major
abdominopelvic oncologic resections. Dosing of unfractionated heparin is typically 5,000 units SC every 8 h for most (if
not all ICU) patients, while the less frequent dosing every
12 h regimen may be appropriate for some patients at lower
risk. Dosing for a common LMWH, enoxaparin, is typically
once daily with 40 mg SC. VTE prophylaxis is typically
administered 1–2 h before any major surgical procedure and
resumed 12–24 h postoperatively. Contraindications to pharmacologic prophylaxis include active bleeding, high risk of
bleeding, systemic anticoagulation, coagulopathy with international normalized ratio (INR) ≥1.5, or thrombocytopenia
(platelet count <50,000).
Mechanical Prophylaxis
Mechanical prophylaxis may include sequential compression devices (SCDs) and thromboembolic deterrent stockings (TEDS). SCDs are preferred over TEDS alone, and
TEDS may be associated with ulcers or skin breakdown,
especially in patients with peripheral vascular disease or
chronic lower extremity wounds and in ICU patients [
30 ].
338
L.M. Kodadek and E.R. Haut
Patients with lower extremity wounds, casts, external fi xation devices, or immobilizers may be unable to utilize SCDs
or TEDS. Finally, compliance with these devices in surgical
patients is poor even without any specifi c contraindications.
Although very little data support its use, ambulation has been
suggested as an effective adjunct to VTE prophylaxis when
feasible [
acceptable replacement to pharmacologic and/or mechanical
prophylaxis in hospitalized patients.
31 ]. However, this should never be considered an
Prophylactic Inferior Vena Cava Filters
Inferior vena cava (IVC) fi lters have been used as prophylaxis in certain high-risk patients without VTE who are
unable to receive pharmacologic prophylaxis. The strongest
data for this indication come from the trauma literature [
EAST offers a level III recommendation (based on retrospective data and/or expert opinion) that a prophylactic IVC fi lter
may be considered in very high-risk trauma patients who are
unable to receive pharmacologic VTE prophylaxis. This recommendation may apply to patients with both increased
bleeding risk and an injury pattern rendering them immobile
for a prolonged period such as severe closed head injury
(Glasgow Coma Scale <8), incomplete spinal cord injury
with paraplegia or quadriplegia, complex pelvic fracture
with associated long bone fracture, or multiple long bone
fractures [
lactic IVC fi lters for primary prevention of VTE [ 25 ].
Prophylactic IVC fi lters are associated with higher mortality
and higher risk of DVT in patients undergoing bariatric surgery [ 33 ]. IVC fi lters may be easily placed at bedside in the
ICU using portable fl uoroscopy and/or intravascular ultrasound techniques. Many IVC fi lters are retrievable and
should be removed as soon as the patient’s acute risk of VTE
decreases. However, fi lter endothelialization may occur as
soon as 3 weeks after placement, and many patients do not
return for IVC fi lter removal, rendering the device effectively
permanent.
26 ]. ACCP recommends against the use of prophy-
32 ].
doses of VTE prophylaxis, VTE may still occur [ 35 ]. One
approach to improve documentation of VTE risk status and
compliance with evidence-based guidelines is to utilize a
mandatory computerized clinical decision support tool
within the institution’s provider order entry system. This
approach has demonstrated dramatic improvements in
prescription of risk-appropriate VTE prophylaxis for medical and surgical patients [
19 , 36 ].
Diagnosis
Signs and symptoms of VTE are nonspecifi c and may include
fi ndings common among critically ill patients including
tachycardia, hypoxia, and fever. Furthermore, physical exam
and history are not useful to rule out a diagnosis of VTE in
the ICU because most critically ill patients with VTE are
“clinically silent” and not detected by history or physical
examination techniques [ 11 ].
DVT may cause local symptoms secondary to partial or
complete occlusion of venous outfl ow including pain, edema,
discoloration, or erythema of the affected area. PE may manifest with symptoms of dyspnea, tachypnea, substernal chest
pain, diaphoresis, hemoptysis, tachycardia, agitation, hypotension, syncope, and/or cardiac arrest. Pleuritic chest pain is
characteristic of smaller emboli, which travel more distally
to cause pleurisy. Large, proximal emboli do not generally
cause pleuritic chest pain. Other signs may include narrowed
pulse pressure, jugular venous distension, acute pulmonary
hypertension, or electrocardiographic evidence of acute right
ventricle strain. A new right bundle branch block or an
S1Q3T3 pattern on electrocardiogram may be indicative of
PE, but the most common fi nding on electrocardiogram is
sinus tachycardia. Any patient with clinical suspicion for
VTE requires further workup to establish or rule out this lifethreatening diagnosis.
Duplex Ultrasonography
Prescription and Administration Compliance
Despite evidence-based guidelines, many ICU patients are
not prescribed and/or administered VTE prophylaxis.
Current efforts focus on ensuring that healthcare providers
prescribe optimal prophylaxis and nurses administer all prescribed doses. Even missing one dose of VTE prophylaxis is
associated with VTE events [ 34 ]. One study recently showed
that only 42 % of patients diagnosed with DVT during a hospitalization had received VTE prophylaxis [ 7 ]. It is impor-
tant to note that not all VTE is preventable. Even when
patients are appropriately prescribed and administered all
DVT was historically diagnosed with invasive contrast
venography, but in current practice, DVT is almost exclusively diagnosed with noninvasive duplex ultrasonography.
DVT may be noted on contrast-enhanced CT scan or magnetic resonance imaging (MRI), but these tests are not frequently specifi cally used to diagnose DVT. Duplex
ultrasonography makes use of B-mode imaging, color
Doppler, and pulsed Doppler spectral analysis [ 37 ]. Acute
and chronic DVT are easily distinguished utilizing duplex
ultrasonography. Acute DVT demonstrates a noncompressible vein with hypoechoic thrombus, spongy texture, and
increased vein diameter due to acute venous hypertension.
Flow may be present around the acute thrombus, suggesting
28 Venous Thromboembolism in the Intensive Care Unit
339
incomplete attachment and possibility for embolization.
Chronic DVT is hyperechoic, fi rmly attached to the vein
wall, and often associated with valvular refl ex. The vein
remains noncompressible or partially compressible with
chronic DVT. Complete duplex examination for lower
extremity DVT involves the superfi cial and deep veins of
both lower extremities.
Computed Tomography Angiography
Invasive pulmonary angiography via right heart catheterization was historically employed to diagnose PE. This
invasive and costly procedure has been replaced with
contrast- enhanced computed tomography (CT) angiography for the diagnosis of PE. Current multidetector helical
CT angiography allows highly accurate diagnosis of PE
[ 38 ]. Furthermore, improvements in imaging modalities
allow visualization of segmental and subsegmental pulmonary arteries, although the clinical importance of treating
peripheral pulmonary emboli is not certain. CT angiography may also identify radiologic parameters of interest. For
example, increased right ventricular/left ventricular diameter ratio on transverse CT images may predict PE-related
mortality [ 39 ].
Echocardiography
Severity of PE may be evaluated with transthoracic or
transesophageal echocardiography. Echocardiography may
also be appropriate for evaluating patients with suspected
PE who are too unstable for transport to radiology for diagnosis by CT. Transthoracic echocardiography is noninvasive and easily applied at the bedside in the ICU, even in
urgent settings with hemodynamically unstable patients.
While the left main pulmonary artery is often obscured by
the air-fi lled left main stem bronchus, Doppler techniques
allow for estimation of pulmonary artery systolic pressure.
Other echocardiographic fi ndings suggestive of pulmonary
embolism include right ventricular dilatation, right atrial
dilatation, displacement of the intraventricular septum into
the left ventricular cavity during systole, and pulmonary
artery dilatation. Although it may be diffi cult to distinguish
acute from chronic pulmonary hypertension based on these
fi ndings, other evidence of chronic disease such as right
ventricular hypertrophy or valvular disease may suggest
other underlying etiologies or comorbid conditions.
Transesophageal echocardiography may not be feasible in
patients with acute hemodynamic collapse, but this technique is also useful and often allows direct visualization of
intraluminal thrombotic material in the main pulmonary
artery or at its bifurcation.
Other Diagnostic Modalities
Ventilation/perfusion scan (V/Q scan) is a nuclear medicine
test sometimes used to diagnose PE in patients who are unable
to undergo contrast-enhanced CT secondary to renal insuffi ciency or severe contrast allergy. D-dimer assay is commonly
used in emergency department patients and outpatients to rule
out VTE due to its high sensitivity. Fibrin D-dimer measures
the fi nal product of the plasmin-mediated degradation of fi brin
and is often elevated in patients with acute VTE. However,
D-dimer is also common in many other conditions associated
with fi brin production including malignancy, trauma, infection, infl ammation, and postoperative state. As such, D-dimer
has poor specifi city and has little predictive value for ICU
patients [ 11 , 40 ]. A negative D-dimer can help rule out the
diagnosis, but a positive test is certainly not confi rmatory for
VTE. Both V/Q scan and D-dimer assay must be utilized in
conjunction with a pretest probability assessment such as the
Wells score or the Geneva score to be clinically useful.
Screening in Asymptomatic Patients
Screening of high-risk asymptomatic patients remains a point
of controversy, and practices among surgeons may vary signifi cantly [ 41 ]. ACCP does not recommend routine screening
for DVT in critically ill patients [ 25 ]. EAST recognizes that
some patients at high risk may benefi t from routine screening
for DVT [ 26 ]. However, the clinical importance of asymptom-
atic DVT detected by routine screening remains unclear.
Supporters of routine screening see benefi t in performing a
relatively inexpensive and noninvasive test (duplex ultrasonography), in order to diagnose and treat asymptomatic DVT
before it progresses to symptomatic or fatal PE. Others feel
that increased medical testing, associated costs, and treatment
of asymptomatic DVT (which may never have come to clinical attention otherwise) incur not only the risk associated with
anticoagulation, but also unnecessary costs. Surveillance bias
(“the more you look, the more you fi nd”) is a common concern
when screening asymptomatic patients for VTE. Studies have
clearly shown that increasing screening is associated with
increasing rates of VTE [ 42 – 45 ]. While national and regional
bodies recognize low incidence of VTE as a marker of quality,
this is a biased measurement since hospitals that less commonly screen patients for VTE are going to identify fewer
VTE events regardless of associated healthcare quality.
Treatment of DVT
The mainstay of treatment for DVT is systemic anticoagulation. Anticoagulation prevents worsening of acute symptoms
and sequelae including recurrent DVT, PE, and post- thrombotic
340
L.M. Kodadek and E.R. Haut
syndromes. Most protocols recommend rapid initiation of
either weight-based intravenous unfractionated heparin infusion (ideally with a loading bolus) or subcutaneous
LMWH. Long-term anticoagulation can continue with either
LMWH or warfarin. Duration of therapy ranges from 3 months
to lifelong therapy based on individualized patient characteristics and risk factors. Provoked DVT, or those cases where a
clear risk factor such as traumatic injury or major surgery is
present, may only require 3 months of anticoagulation.
Spontaneous DVT without a clear risk factor is usually treated
for 3–6 months. Patients with recurrent VTE, or those with an
ongoing hypercoagulable state such as known thrombophilia
or malignancy, should probably continue anticoagulation
indefi nitely. Rarely, patients with distal DVT and no predisposing risk factors for VTE may be managed with compression stockings alone, although this is most appropriate for
ambulatory outpatients, rather than the ICU population.
Repeat duplex ultrasonography after 2 weeks should be performed to ensure resolution of the clot. Propagation of the clot
to the level of the popliteal vein or more proximal warrants
transition to anticoagulation.
Thrombolysis or thrombectomy has been proposed for
certain subsets of patients at low operative risk with proximal iliofemoral or femoral DVT, especially among young
patients at high risk for post-thrombotic syndrome. A large,
ongoing, multicenter, prospective randomized controlled
trial (the ATTRACT study) should help resolve the question
of whether pharmacomechanical catheter-directed thrombolysis benefi ts patients with large DVT [
46 ]. Patients with a
threatened limb from phlegmasia cerulea dolens or phlegmasia alba dolens should undergo thrombolysis or thrombectomy for the purpose of limb salvage. In patients with DVT
and absolute contraindication to anticoagulation or in
patients with recurrent DVT on adequate anticoagulation,
IVC fi lter placement is indicated.
secondary to PE, extracorporeal membrane oxygenation
(ECMO) has been suggested [
47 ]. Similar to long-term treat-
ment for DVT, provoked PE is often treated with
anticoagulation for 3 months, spontaneous PE is frequently
treated for 3–6 months, and patients with ongoing risk factors are most often treated indefi nitely.
Impact
Prevention of VTE remains one of the most important patient
safety practices in hospitalized patients, in particular for
those in the ICU. However, even when patients are prescribed
and administered VTE prophylaxis according to best practice guidelines, VTE may still not be preventable in as many
as 50 % of cases [ 35 ]. National bodies including the Centers
for Medicare and Medicaid Services and regional entities
impose fi nancial penalties when hospitalized patients
develop VTE. Policy changes at the regional and national
level should focus on a more impactful approach. Rather
than measuring incidence of VTE alone, some experts argue
for a pure process measure approach or combined process
and outcome measure instead [ 43 , 48 , 49 ]. A true benchmark
of patient safety and quality care should measure how frequently patients are prescribed and administered VTE prophylaxis according to best practice guidelines.
Conclusion
VTE is common among surgical patients with critical illness
and represents a major source of morbidity and mortality. All
ICU patients without contraindications require riskappropriate VTE prophylaxis because all ICU patients are at
risk for VTE. It is important to recognize risk factors, provide effective prophylaxis, and provide timely and accurate
diagnosis and treatment for patients with VTE in the ICU.
Treatment of PE
Treatment of PE in the hemodynamically stable patient
begins with initiation of either weight-based intravenous
unfractionated heparin infusion (ideally with loading bolus)
or subcutaneous LMWH. However, standard VTE treatment
with anticoagulation alone is not adequate for many patients
with massive and submassive PE. In the setting of hemodynamic instability and/or right ventricular dysfunction, other
aggressive and invasive therapies may be indicated. Systemic
thrombolytic therapy such as intravenous alteplase or
catheter- directed thrombolysis or embolectomy may be warranted. In patients with contraindications to anticoagulation
such as intracranial hemorrhage or active bleeding, surgical
embolectomy and/or IVC fi lter may be necessary. In some
cases, particularly those with acute cardiorespiratory failure
References
1. US Department of Health and Human Services. Surgeon gener-
al’s call to action to prevent deep vein thrombosis and pulmonary
embolism 2008. Available at:
NBK44178/
2. Shojania KG, Duncan BW, McDonald KM, Wachter RM, Markowitz
AJ. Making health care safer: a critical analysis of patient safety practices. Evid Rep Technol Assess (Summ). 2001;i–x:1–668.
3. Maynard G, Stein J. Preventing hospital-acquired venous thrombo-
embolism: a guide for effective quality improvement. Prepared by
the Society of Hospital Medicine. AHRQ Publication No. 08–0075.
Rockville: Agency for Healthcare Research and Quality. Aug 2008.
4. Shekelle PG, Pronovost PJ, Wachter RM, et al. The top patient
safety strategies that can be encouraged for adoption now. Ann
Intern Med. 2013;158:365–8.
5. Shekelle PG, Wachter RM, Pronovost PJ, et al. Making Health
Care Safer II: An Updated Critical Analysis of the Evidence for
Patient Safety Practices. Comparative Effectiveness Review No.
211. (Prepared by the Southern California-RAND Evidence-
. Accessed 15 Sept 2015.
http://www.ncbi.nlm.nih.gov/books/
28 Venous Thromboembolism in the Intensive Care Unit
341
based Practice Center under Contract No. 290-2007-10062-I.)
AHRQ Publication No. 13-E001-EF. Rockville, MD: Agency for
Healthcare Research and Quality. 2013.
fi ndings/evidence-based-reports/ptsafetyuptp.html
6. Cohen AT, Tapson VF, Bergmann J, et al. Venous thromboembolism risk and prophylaxis in the acute hospital care setting
(ENDORSE study): a multinational cross-sectional study. Lancet.
2008;371(9610):387–94.
7. Goldhaber SZ, Tapson VF, DVT FREE Steering Committee. A prospective registry of 5,451 patients with ultrasound-confi rmed deep
vein thrombosis. Am J Cardiol. 2004;93(2):259–62.
8. Attia J, Ray JG, Cook DJ, Douketis J, Ginsberg JS, Geerts
WH. Deep vein thrombosis and its prevention in critically ill adults.
Arch Intern Med. 2001;161(10):1268–79.
9. Minet C, Potton L, Bondadona A, et al. Venous thromboembolism
in the ICU: main characteristics, diagnosis and thromboprophylaxis. Crit Care. 2015;19(1):287.
10. McLeod AG, Geerts W. Venous thromboembolism prophylaxis in
critically ill patients. Crit Care Clin. 2011;27:765–80.
11. Crowther MA, Cook DJ, Griffi th LE, et al. Deep venous thrombosis: clinically silent in the intensive care unit. J Crit Care.
2005;20:334–40.
12. Velmahos GC, Spaniolas K, Tabbara M, et al. Pulmonary embolism
and deep venous thrombosis in trauma: are they related? Arch Surg.
2009;144(10):928–32.
13. Kearon C. Natural history of venous thromboembolism. Circulation.
2003;107(23 Suppl 1):I22–30.
14. Ibrahim EH, Iregui M, Prentice D, Sherman G, Kollef MH, Shannon
W. Deep vein thrombosis during prolonged mechanical ventilation
despite prophylaxis. Crit Care Med. 2002;30(4):771–4.
15. Kaplan D, Casper TC, Elliott CG, et al. VTE incidence and risk
factors in patients with severe sepsis and septic shock. Chest. 2015.
10.1378/chest.15-0287 .
doi:
16. Cook D, Crowther M, Meade M, et al. Deep venous thrombosis in
medical-surgical critically ill patients: prevalence, incidence, and
risk factors. Crit Care Med. 2005;33:1565–71.
17. Merrer J, De Jonghe B, Golliot F, et al. Complications of femoral
and subclavian venous catheterization in critically ill patients: a
randomized controlled trial. JAMA. 2001;286:700–7.
18. Maynard G, Morris T, Jenkins I, et al. Optimizing prevention of
hospital acquired venous thromboembolism: prospective validation
of a VTE risk assessment model. J Hosp Med. 2010;5(1):10–8.
19. Streiff MB, Carolan HT, Hobson DB, et al. Lessons from the Johns
Hopkins multi-disciplinary venous thromboembolism (VTE) prevention collaborative. Br Med J. 2012;344:e3935.
20. Caprini JA, Arcelus JI, Hasty JH, et al. Clinical assessment of
venous thromboembolic risk in surgical patients. Semin Thromb
Hemost. 1991;17 Suppl 3:304–12.
21. Barbar S, Noventa F, Rossetto V, et al. A risk assessment model for
the identifi cation of hospitalized medical patients at risk for venous
thromboembolism: the Padua Prediction Score. J Thromb Haemost.
2010;8:2450–7.
22. Rogers Jr SO, Kilaru RK, Hosokawa P, et al. Multivariable predictors of postoperative venous thromboembolic events after general
and vascular surgery: results from the patient safety in surgery
study. J Am Coll Surg. 2007;204(6):1211–21.
23. Spyropoulos AC, Anderson FA, FitzGerald G, IMPROVE
Investigators, et al. Predictive and associative models to identify hospitalized medical patients at risk for VTE. Chest.
2011;140(3):706–14.
24. Obi AT, Pannucci CJ, Nackashi A, et al. Validation of the caprini venous thromboembolism risk assessment model in critically ill surgical patients. JAMA Surg. 2015. doi:
jamasurg.2015.1841
25. Guyatt GH, Akl EA, Crowther M, et al. Antithrombotic therapy
and prevention of thrombosis, 9th ed: American College of Chest
.
www.ahrq.gov/research/
.
10.1001/
Physicians evidence-based clinical practice guidelines. Chest.
2012;141:7S–47.
26. Rogers FB, Cipolle MD, Velmahos G, Rozycki G, Luchette
FA. Practice management guidelines for the prevention of venous
thromboembolism in trauma patients: the EAST practice management guidelines work group. J Trauma. 2002;53(1):142–64.
27. Johanson NA, Lachiewicz PF, Lierberman JR, et al. American
academy of orthopaedic surgeons clinical practice guideline
on prevention of symptomatic pulmonary embolism in patients
undergoing total hip or knee arthroplasty. J Bone Joint Surg Am.
2009;91(7):1756–7.
28. Cook D, Meade M, Guyatt G, et al. Dalteparin versus unfractionated heparin in critically ill patients. N Engl J Med.
2011;364:1305–14.
29. Beitland S, Sandven I, Lill-Kristin K, Sandset PM, Sunde K, Eken
T. Thromboprophylaxis with low molecular weight heparin versus unfractionated heparin in intensive care patients: a systematic
review with meta-analysis and trial sequential analysis. Intensive
Care Med. 2015;4:1209–19.
30. The CLOTS (Clots in Legs Or sTockings after Stroke) Trial
Collaboration. Thigh-length versus below-knee stockings for deep
venous thrombosis prophylaxis after stroke: a randomized trial.
Ann Intern Med. 2010;153(9):553–62.
31. Lau BD, Streiff MB, Kraus PS, et al. No evidence to support ambulation for reducing postoperative venous thromboembolism. J Am
Coll Surg. 2014;219(5):1101–3.
32. Haut ER, Garcia LJ, Shihab HM, et al. The effectiveness
of prophylactic Inferior Vena Cava (IVC) fi lters in trauma
patients: a systematic review and meta-analysis. JAMA Surg.
2014;149(2):194–202.
33. Brotman DJ, Shihab HM, Prakasa KR, et al. Pharmacologic and
mechanical strategies for preventing venous thromboembolism
after bariatric surgery: a systematic review and meta-analysis.
JAMA Surg. 2013;148(7):675–86.
34. Louis SG, Sato M, Geraci T, et al. Correlation of missed doses
of enoxaparin with increased incidence of deep vein thrombosis in trauma and general surgery patients. JAMA Surg.
2014;149(4):365–70.
35. Haut ER, Lau BD, Kraus PS, et al. Preventability of hospital- acquired
venous thromboembolism. JAMA Surg. 2015;150(9):912–5.
36. Haut ER, Lau BD, Kraenzlin FS, et al. Improve prophylaxis
and decreased rates of preventable harm with the use of a mandatory computerized clinical decision support tool for prophylaxis for venous thromboembolism in trauma. Arch Surg.
2012;147(10):901–7.
37. Barleben A, Bandyk DF. Interpretation of peripheral venous duplex
testing. Semin Vasc Surg. 2013;26(2–3):111–9.
38. Quiroz R, Kucher N, Zou KH, et al. Clinical validity of a negative
computed tomography scan in patients with suspected pulmonary
embolism: a systematic review. JAMA. 2005;293:2012–7.
39. Meinel FG, Nance Jr JW, Schoepf UJ, et al. Predictive value of
computed tomography in acute pulmonary embolism: systematic
review and meta-analysis. Am J Med. 2015;128(7):747–59.
40. Crowther MA, Cook DJ, Griffi th LE, et al. Neither baseline tests
of molecular hypercoagulability nor d dimer levels predict deep
venous thrombosis in critically ill medical surgical patients.
Intensive Care Med. 2005;31(1):48–55.
41. Haut ER, Schenider EB, Patel A, et al. Duplex ultrasound screening
for deep vein thrombosis in asymptomatic trauma patients: a survey
of individual trauma surgeon opinions and current trauma center
practices. J Trauma. 2011;70(1):27–34.
42. Pierce CA, Haut ER, Kardooni S, et al. Surveillance bias and deep
vein thrombosis in the national trauma data bank: the more we look,
the more we fi nd. J Trauma. 2008;64:932–7.
43. Haut ER, Pronovost PJ. Surveillance bias in outcomes reporting.
JAMA. 2011;305(23):2462–3.
342
L.M. Kodadek and E.R. Haut
44. Haut ER, Noll K, Efron DT, et al. Can increased incidence of Deep
Vein Thrombosis (DVT) be used as a marker of quality of care in the
absence of standardized screening? The potential effect of surveillance bias on reported DVT rates. J Trauma. 2007;63(5):1132–7.
45. Bilimoria KY, Chung J, Ju M, et al. Evaluation of surveillance bias
and the validity of the venous thromboembolism quality measure.
JAMA. 2013;310(14):1482–9.
46. The ATTRACT study. Available at:
Accessed 12 Oct 2015. [Or insert citation once published].
http://attract.wustl.edu/# .
47. Yusuff HO, Zochios V, Vuylsteke A. Extracorporeal membrane
oxygenation in acute massive pulmonary embolism: a systematic
review. Perfusion. 2015;30:611–6. pii: 0267659115583377.
48. Bilimoria KY. Facilitating quality improvement: pushing the pendulum back toward process measures. JAMA. 2015;314(13):1333–4.
49. Aboagye JK, Lau BD, Schneider EB, Streiff MB, Haut ER. Linking
processes and outcomes: a key strategy to prevent and report harm
from venous thromboembolism in surgical patients. JAMA Surg.
2013;148(3):299–300.
Glycemic Control and Insulin Resistance
Richard N. Lesperance and Oscar D. Guillamondegui
2 9
Type 2 diabetes has become an epidemic in the developed
world. The US Centers for Disease Control and Prevention
estimates that 29 million (9.3 %) US residents have diabetes.
Up to 37 % of the US population (80 million) are pre-diabetics as measured by elevated hemoglobin A1C levels [
These pre-diabetics demonstrate increased peripheral insulin
resistance and pancreatic β(beta)-cell dysfunction and are at
increased risk for renal and vascular complications [ 2 ].
Additionally, patients without diagnosed diabetes may frequently experience hyperglycemia during critical illness. In
epidemiological studies, 75 % of adult ICU patients demonstrated either hyperglycemia or insulin resistance [ 3 ].
Cytokines and other soluble infl ammatory modulators cause
decreased glycolysis and increased peripheral insulin resistance in critically ill patients [ 4 ]. Stress hormones such as
catecholamines and glucocorticoids are increased during
critical illness, or frequently administered exogenously, and
promote hyperglycemia [ 5 ] (Fig. 29.1 .) Additionally, glu-
cose is a common component of IV infusions in the ICU and
a major component of total parenteral nutrition, thus promoting hyperglycemia.
Acute hyperglycemia is common in critically ill patients
admitted to the intensive care unit (ICU). Previously, hyperglycemia was thought to be an expected response to critical
illness and not aggressively treated until glucose levels
exceeded 200 mg/dl [ 6 , 7 ]. However, over the last decade,
hyperglycemia has been increasingly correlated with worse
outcomes in a wide variety of critically ill patients, including
those with myocardial infarctions [ 8 ], after non-cardiac sur-
gery [ 9 ], and in the trauma population [ 10 ]. Elevated blood
glucose has been shown to elevate the risk of surgical site
infections and prolong hospital stays in postoperative
patients [ 9 , 11 ], presumably due to interference with normal
neutrophil function [ 12 – 14 ]. The apparent deleterious effect
R. N. Lesperance , MD • O. D. Guillamondegui , MD, MPH (*)
Department of Surgery , Vanderbilt University Medical Center ,
Nashville , TN 37212 , USA
Richard.n.lesperance@vanderbilt.edu; oscar.
e-mail:
guillamondegui@vanderbilt.edu
1 ].
of hyperglycemia in diverse ICU patients led to interest in
strictly controlling serum glucose levels and the landmark
randomized controlled trial (RCT) from the Leuven group in
Belgium [
15 ].
Overview of Evidence Supporting Strict Glucose Control in the ICU
In 2001, researchers from the University of Leuven published the results of their RCT [ 15 ] comparing strict glucose
control (80–110 mg/dl) to their conventional standard of care
(180–200 mg/dl), in a primarily surgical ICU with a large
proportion of cardiac surgery patients. Their study was prematurely halted after the planned interim analysis when the
strict glucose control group demonstrated superior outcomes.
The overall randomization of approximately 1500 patients
demonstrated that the strict control group almost universally
required insulin infusion to maintain euglycemic control,
while in the conventional group, only 39 % required intravenous insulin infusion to meet glucose targets.
The improvements in the strict glucose control group
were notable: a reported 42 % relative risk reduction for ICU
mortality (4.6 % vs. 8 %) as well as a decrease in overall inhospital mortality (7.2 % vs. 10.9 %) favoring intensive glucose control. Despite the overall cohort being heavily
weighted toward cardiac surgery patients, the majority of
mortality benefi t occurred in patients remaining in the ICU
for greater than 5 days. This group was heavily weighted to
non-cardiac admissions. The mortality in this group was
10.6 % for the strict control group, as opposed to 20.2 % for
the conventional arm. They also reported signifi cant improvements in other aspects of intensive care management including rates of septicemia, time on ventilator support, and need
for renal replacement therapy.
However, when the Leuven group used a similar protocol
in a medical ICU [ 16 ], they found no improvement to overall
mortality. Interestingly, when they looked at a (predefi ned)
subgroup of patients staying in the ICU for 3 or more days,
© Springer International Publishing Switzerland 2016
N.D. Martin, L.J. Kaplan (eds.), Principles of Adult Surgical Critical Care, DOI 10.1007/978-3-319-33341-0_29
343
344
R.N. Lesperance and O.D. Guillamondegui
Glucocorticoids, catecholamines
(endogenous and exogenous)
Circulation and Electrolytes
Fluid depletion
Hypoperfusion
Electrolyte loss
Inflammation,
cytokines
Insulin resistance,
beta-cell dysfunction
Hyperglycemia
Sepsis
Impaired wound healing
Neuromyopathy
Dextrose
(IV, enteral)
Lipolysis
(fat cells)
Gluconeogenesis
(liver)
Cellular Effects
Mitochondrial injury
Neutrophil dysfunction
Endothelial dysfunction
Molecular Effects
Oxidant injury
Protein glycation
Complement inhibition
Fig. 29.1 Causes and effects of stress hyperglycemia (From Kavanagh and McCowen [ 73 ]. Copyright ©2010 Massachusetts Medical Society.
Reprinted with permission from Massachusetts Medical Society)
there appeared to be a survival advantage (43 % vs. 52.5 %).
This benefi t, however, was offset by increased mortality
among patients staying in the ICU for fewer than 3 days, a
fi nding the authors could not explain in their study. Regardless
of the lack of overall mortality benefi t, patients receiving
strict glucose control did show an improvement in the number of ventilator days, need for renal replacement therapy,
and ICU length of stay.
Unfortunately, since the publication of those two singlecenter studies, multiple other investigators have been unable
to replicate those benefi ts in larger multi-center trials, several
of which are recounted here. The studies are summarized in
Table 29.1 .
The VISEP trial [ 17 ] was conducted among 18 academic
centers in Germany. It examined both the use of intensive
glucose control and pentastarch resuscitation, among
patients presenting with sepsis and septic shock. The study
was halted early, at the fi rst planned safety analysis, due to
an increased rate of hypoglycemic events (glucose ≤40 mg/
dl) among patients randomized to intensive glucose control
(17 % vs. 4.1 %.) Although the authors were unable to identify any of the hypoglycemic events as having directly
caused death or disability, regression analysis did identify
hypoglycemia as an independent risk factor for death by
any cause. They did not fi nd any mortality benefi t to inten-
sive glucose control, and they were unable to demonstrate
improvements in ventilator time, ICU length of stay, or
need for renal replacement therapy. The trial also elicited
more “severe adverse events” in the strict glucose control
group.
The Glucontrol trial [ 18 ] was another large multi-center
randomized trial conducted mainly in European mixed medical/surgical ICUs. They compared strict glucose control
(80–110 mg/dl) to a control group with a slightly tighter
standard (140–180 mg/dl) than that utilized in the original
Leuven studies. This study was also halted early, unfortunately, due to the high rate of “protocol violations” for glucose control. For example, only 39 % of the recorded blood
glucose values were actually in the target range for the strict
control group. Since these investigators analyzed all of their
recorded glucose values, as opposed to the Leuven trials [ 15 ,
16 ] (which only analyzed the admission and morning val-
ues), no direct comparison of their accuracy compared to
Leuven can be made.
Despite halting their trial early, Glucontrol did accrue
over 500 patients per group. When they analyzed their data,
they found no benefi t of strict glucose control on mortality,
organ failure, ventilator days or ICU duration. There was no
distinguishable benefi t noted with the patient subsets within
or outside the determined glucose targets.
29 Glycemic Control and Insulin Resistance
Table 29.1 Comparison of major randomized controlled trials of intensive glucose control
Study Population Intensive group Control group Hypoglycemia rates Results
Leuven 2001 [
Leuven 2006 [
Glucontrol [
VISEP [
NICE-SUGAR [
ICU intensive care unit, LOS length of stay, RRT renal replacement therapy, SAE serious adverse event
15 ] Single-center surgical ICU,
1500 patients
16 ] Single-center medical ICU,
1200 patients
18 ] 21 academic medical/surgical
ICUs, 1100 patients
17 ] 18 academic medical/surgical
ICUs, 537 patients with
sepsis or septic shock
19 ] 42 medical/surgical ICUs,
6100 patients
80–110 180─200 5.1 % vs. 0.8 % Mortality benefi t 4.6 % vs. 8 %,
also better septicemia, ventilator
LOS, and need for RRT
80–110 180–200 18.7 % vs. 3.1 % No mortality benefi t, better
ventilator LOS, and need for
RRT
80–110 140–180 8.7 % vs. 2.7 % Study halted early due to
protocol violations. No mortality
benefi ts
80–110 180–200 17 % vs. 4 % Trial halted early for increased
rate of SAEs (11 % vs. 5 %) in
intensive group
81–108 144–180 6.8 % vs. 0.5 % Increased mortality for intensive
glucose control, 27.5 % vs.
24.9 %
345
The final large, multinational randomized trial was
NICE- SUGAR [ 19 ], involving over 6,000 patients in
mixed medical- surgical ICUs in Australia, New Zealand,
and Canada. Patients were randomized to strict control of
81–108 mg/dl or a conventional group with targets
between 144 and 180 mg/dl. Their primary outcome was
a 90-day mortality. These investigators were able to
define a mortality difference but, contrary to the Leuven
trials, one that favored the conventional group. Patients
receiving intensive glucose control had a 27.5 % 90-day
mortality, as opposed to 24.9 % in the conventional group.
Despite this study being conducted in mixed ICUs, when
the results were analyzed by predefined subgroups, surgical patients benefited the most from looser blood glucose
targets with a 31 % improvement in survival (odds ratio
of 1.31).
To try and resolve these discrepant results, two large
meta-analyses were performed including several smaller
randomized studies in different environments. The fi rst
meta- analysis [
20 ] did not include the data from the ongo-
ing NICE-SUGAR study. The pooled results showed no
mortality benefi t to intensive glucose control and no
improvement in organ failure, although there was a
decreased risk of septicemia. The second meta-analysis
included NICE-SUGAR data, as well as slightly different
inclusion criteria for other studies, resulting in approximately 13,000 included patients [
21 ]. They also found no
overall benefi t to mortality for the strict control group. In
contradiction to the previous meta- analysis, this study
showed the only group identifi ed to receive a survival benefi t to strict glucose control was those patients managed in
surgical ICUs (odds ratio of 0.63 favoring intensive control,
95 % CI 0.44–0.91.) Both analyses identifi ed a higher incidence of hypoglycemic events in patients receiving intensive glucose control.
Resolving the Differences Between Studies of Intensive Glucose Control
There have been several theories advanced to explain the discrepant results between the Leuven studies and subsequent
studies. The initial Leuven study [ 15 ] had a much higher rate
of IV glucose administration (200–300 g/day) than the
Glucontrol, VISEP, and NICE-SUGAR studies provided.
A second theory is that the control group of the subsequent studies targeted a more physiologically appropriate
blood glucose range (140–180 mg/dl) than the liberally set
range of the initial Leuven study (which allowed patients to
reach 215 mg/dl before starting therapy). The true benefi t of
the subsequent studies may not be the achievement of an artifi cial “normoglycemia” but simply the avoidance of excessive (>180 mg/dl) hyperglycemia. The exception to the lower
conventional target range was the VISEP study, which used
a range of 180–200 mg/dl, but since that study was halted
early, it may have been underpowered to detect a difference.
Another possibility is a higher than expected mortality
rate among patients in the initial Leuven study. The mean
APACHE II score for both the intensive and conventional
control groups was 9. For postoperative patients, this should
yield an in-hospital mortality of 3.9 % [ 22 ], yet the (hospital)
mortality in the two groups was 7.2 and 10.9 %. There are
criticisms about the use of APACHE II, however, in comparing mortality estimates between different facilities, the
APACHE II system was originally derived using a North
American population [ 23 ], and this may be less refl ective of
the European cohort of critically ill patients. This may be
related to patient selection bias. Additionally, entering data
for the calculation of APACHE II scores is heavily dependent on medical staff training [ 24 ], and this may be affected
by systematic differences in North American and European
healthcare delivery systems. One similarity among all the
346
R.N. Lesperance and O.D. Guillamondegui
studies referenced above was the consistency of hypoglycemia identifi ed within the strict glucose control group.
Universally, the rate of hypoglycemia was higher than the
conventional/liberal glucose control groups. In the initial
Leuven trial, the strict glucose control group had a 5.1 % rate
of hypoglycemia. In the subsequent trials, the rate varied
from 6.8 to 19 % [
mia in the ICU is frequently correlated with mortality.
A fi nal potential difference between the Leuven results
and those achieved by subsequent investigators is the accuracy of the blood glucose measurements. Inaccurate measurements of blood glucose might result in higher rates of
actual (if not measured) hypoglycemia (see the section on
Glucose Measurement in the ICU, below). In the initial
Leuven study, only arterial blood samples were used and
were assayed on blood gas analyzers. In both the second
Leuven trial and the subsequent larger RCTs described
above, a convenience mix of blood gas analyzers and pointof- care fi nger stick monitors were used for the measurements
of samples of both arterial and capillary origin.
4 , 16 – 19 ]. As described below, hypoglyce-
The Role of Hypoglycemia and Glucose Variability in ICU Mortality
As stated above, a consistent fi nding in studies of intensive
glucose control is an increased rate of hypoglycemia among
patients targeted for tighter control. Hypoglycemia was not
felt to contribute to mortality by the authors of the original
Leuven study [ 15 ]. Since then, there have been increasing
concerns that the impact of hypoglycemia reduces or eliminates any benefi t obtained from intensive glucose control.
The neuroglycopenic effects of hypoglycemia are well
known [ 25 ]. Severe hypoglycemia causes brain neuronal
death in a pattern distinct from cerebral ischemia [ 26 ] and
appears to worsen after reperfusion with glucose [ 27 ]. Even
in the absence of overt hypoglycemia, patients with traumatic brain injury who undergo intensive glucose control
have decreased brain glucose measured by microdialysis,
accompanied by increased markers of cellular distress [ 28 ].
Hypoglycemia also interferes with adrenocortical responsiveness to ACTH during stress states [ 29 , 30 ], which could
presumably interfere with response to septic insults.
Additionally, episodes of hypoglycemia decrease the adrenergic responsiveness to subsequent hypoglycemic insults.
This progressive effect could interfere both with attempts to
return to normoglycemia and also with response to sepsis, by
decreasing endogenous catecholamine and corticosteroid
response [ 31 , 32 ].
Hypoglycemia may cause harmful cardiovascular effects
through several mechanisms. Overnight asymptomatic hypoglycemia in diabetics has been associated with prolongation
of the QTc interval and other conduction abnormalities and
arrhythmias [ 33 , 34 ]. Hypoglycemia causes an overall pro-
infl ammatory and pro-thrombotic state [
mechanisms involved in cardiovascular risk may be from
induction of platelet aggregation [
function and mitochondrial oxidative stress from interference with nitric oxide signaling [ 38 , 39 ].
The fi rst Leuven study [
glycemia in the intensive control group, but the authors felt
there was no harm from these episodes. The subsequent studies, however, have correlated hypoglycemic events with an
increased risk of mortality, even after correcting for disease
severity [ 40 , 41 ]. This suggests that hypoglycemia is not
simply a marker of a sicker patient. Even relatively mild
hypoglycemia (72–81 mg/dl) in ICU patients has been independently associated with death from cardiovascular or septic causes [ 42 ].
Aside from hypoglycemia, blood glucose variability is
becoming increasingly appreciated as a marker of mortality.
Most major studies of intensive glucose control evaluate
their impact by looking at the mean serum glucose values,
but this may obscure time spent hyper- or hypoglycemic [ 4 ]
(see Fig. 29.2 .) Standard deviation (SD) is used to express
variability about a mean value in statistics. Several investigations have correlated higher SD in blood glucose values with
ICU mortality, including patients who had “normal” mean
glucose values [ 43 , 44 ]. The standard deviation, however,
does not discriminate between gentle changes in glucose levels over time, as opposed to vigorous fl uctuations of hyperand hypoglycemic states. Other investigators have used
different measures of variability to account for this and found
that they correlate better with mortality [ 45 , 46 ]. It remains
to be seen whether glucose fl ux is a possible target for intervention or simply another marker of underlying illness severity. Conversely, it is possible that lower glucose variability
simply refl ects more attentive medical and nursing care, and
mortality benefi ts accrue from this “nursing Hawthorne
effect” [ 47 ].
15 ] found increased rates of hypo-
35 , 36 ]. Some of the
37 ], and endothelial dys-
What Is the Appropriate Target for Glucose Control?
It is clear that the ICU practitioner must make every effort to
avoid hypoglycemia, whether due to disease process or iatrogenic. The question of what blood glucose level to target
however remains diffi cult, given the harmful effects of
uncontrolled hyperglycemia. While it is obvious that blood
glucose levels <81 mg/dl have been associated with poor
outcomes, it is not certain at which level hyperglycemia
becomes detrimental. Certainly, as outlined earlier, levels
>200 mg/dl should be avoided. The results of the NICESUGAR study suggested that patients allowed to remain
slightly hyperglycemic (144–180 mg/dl) had better out-