http://surgerybook.net/

pulmonary edema or interstitial lung disease) can be treated effectively by exogenous administration of

oxygen. The A-a gradient is normal in hypoxemia due to pure hypoventilation and increased in V/Q

mismatch. V/Q mismatch, the cause of most hypoxemia in the ICU, occurs with decreased airflow to the

alveoli in relation to the amount of pulmonary capillary blood flow. In the normal lung, there is V/Q

mismatch because perfusion and ventilation are heterogeneous. Both ventilation and perfusion are

greater in the bases than in the apices. However, the difference between apical and basilar ventilation is

less than the difference between apical and basilar perfusion. Therefore, the V/Q ratio is higher in the

apices than in the bases. In the diseased lung, V/Q mismatch increases because heterogeneity of both

ventilation and perfusion worsen resulting in hypoxemia. Common causes of hypoxemia due to V/Q

mismatch include obstructive lung diseases, pulmonary vascular diseases, and interstitial diseases.

Shunting occurs when there is adequate ventilation of the alveoli but decreased pulmonary capillary

blood flow. Shunting may be anatomic as with intracardiac shunts and hepatopulmonary syndrome or

physiologic as when nonventilated alveoli are perfused as with pneumonia. Hypoxia due to shunting

with venous admixture as with severely depressed cardiac output or anemia cannot be overcome by

administration of oxygen (Fig. 10-4).

Figure 10-4. West lung zones. Zone 1: Not observed in healthy lung. Alveolar pressure exceeds pulmonary blood vessel pressure

leading to alveolar dead space. Zone 2: Located approximately 3 cm above heart. Pulmonary vessel pressure exceeds alveolar

pressure in a pulsatile fashion and allows pulmonary blood flow and alveolar distention. Zone 3: Majority of healthy lungs with

continuous blood flow and oxygenation. PA

, Alveolar Pressure; Pa

, arterial pressure; PV

, venous pressure.

Ventilation and CO2 Removal

Ventilation refers to the exchange of air from the atmosphere to the lungs and alveoli. The rate at which

alveolar ventilation occurs determines the amount of carbon dioxide (CO2

) excretion, a readily

diffusible substance. The amount of CO2 produced and requiring elimination is related to the respiratory

quotient (RQ). RQ is the ratio of CO2 produced to O2 consumed and is normally 0.8, depending on the

diet. Chemoreceptors in the medulla are responsive to pH changes and when patients become acidotic

they hyperventilate to decrease PaCO2

to maintain homeostasis at 40 mm Hg. End tidal CO2 may be

lower than PaCO2

if substantial lung is ventilated but not perfused (dead space).

Pulmonary Mechanics

The relationship between ventilatory volumes and pressures is referred to as pulmonary mechanics and

depends on compliance. The functional lung in acute lung injury is smaller with heterogeneous disease,

necessitating a protective lung strategy of ventilation that will be described below.

Figure 10-5. Flow–time scalar showing autoPEEP.

VENTILATOR MODES AND GRAPHICS

One of the initial steps in evaluating a patient with respiratory failure is to establish a patent and

307

http://surgerybook.net/

protected airway. The ICU team must determine if the patient has an acute issue causing the respiratory

failure that can be treated without mechanical ventilation. Through the use of physical examination,

bedside ultrasound and radiography, the ICU team may diagnose a pneumothorax or acute pulmonary

edema that can be treated without intubation. If the patient has acute respiratory failure without an

easily treatable process, intubation and mechanical ventilation may be safest. When setting up the

mechanical ventilator there are several options and modalities from which to select.

Mechanical ventilation improves oxygenation and ventilation by attempting to improve V/Q

mismatch by decreasing the amount of shunt. Current recommendations are for low tidal volume (Vt)

settings of 6 to 8 mL/kg for the prevention of barotrauma in surgical ICU patients.42 Respiratory Rate

(RR) is typically set at 12 to 16 breaths/min. RR is altered to achieve an optimal PaCO2

. Positive end

expiratory pressure (PEEP) is applied to prevent significant alveolar collapse during expiration and is

particularly important in low lung volume ventilation strategies. PEEP is typically started at 5 cm H2O

but can be titrated up to improve oxygenation. The initial setting for patients with respiratory failure is

an FiO2 of 1.0. After initial check of ABG the FiO2

is weaned down to maintain a minimal PaO2 of 60

mm Hg and O2 saturation of >90% with goals of FiO2 <0.4 to prevent O2

toxicity.

Ventilatory mechanics can be displayed graphically and include scalars and loops. Scalars are plots

against time – either flow, pressure, or volume. Several loops are helpful clinically and include the

pressure–volume and flow–volume loops. Each graphic can be important in demonstrating a physiologic

anomaly and displays a certain pattern depending on the ventilator mode. The flow versus time scalar is

helpful for identifying air trapping or auto-PEEP that can accompany certain ventilator modes (Fig. 10-

5). This can result in barotrauma or hemodynamic instability. Measures to decrease auto-PEEP include

decreasing tidal volume, respiratory rate, or inspiratory time (depending on ventilator mode) or

increasing flow rate. The pressure time scalar allows the practitioner to view and calculate compliance

(Fig. 10-6). Further, one can observe elevations in peak airway pressure that may occur as a result of a

large pneumothorax or kinked endotracheal tube or a sharp decrease that may represent an airway leak

or disconnection. The volume versus time scalar allows one to identify an airway leak as well (Fig. 10-

7). Ventilators often display all three scalars simultaneously. A spontaneously breathing patient is

identified by the negative deflection in the pressure–time scalar at breath initiation (Fig. 10-8). Distinct

ventilator modes have unique methods of cycling, triggering, and limitations of breaths. In controlled

mandatory ventilation, breaths are volume-cycled, time-triggered, and flow-limited (Fig. 10-9). In

pressure-controlled ventilation, breaths are time-cycled, time–triggered, and pressure-limited. Note the

distinct difference in the flow and pressure versus time scalars compared to CMV (Fig. 10-10). Pressure

supported breaths are flow-cycled, patient-triggered, and pressure-limited (Fig. 10-11). Pressure and

volume controlled ventilation can be set in an assist control (AC) mode or intermittent mandatory

ventilation (IMV) mode. With the AC mode the patient receives a set number of breaths delivered per

minute whether with pressure or volume. Patients who initiate breaths on IMV will have tidal volumes

determined by their own respiratory effort and not the ventilator. Patients who are critically ill may

require AC for full support but these patients should be weaned to an IMV or PSV mode quickly to

prevent respiratory muscle atrophy.43

Figure 10-6. Pressure–time scalar.

ACUTE RESPIRATORY FAILURE

Diagnosis

6 Past consensus definitions of adult respiratory distress syndrome (ARDS) and acute lung injury (ALI)

308

http://surgerybook.net/

categorized them by acute onset hypoxemia, bilateral infiltrates on chest radiography, and absence of

left atrial hypertension (i.e., noncardiac in etiology).44 ALI and ARDS were differentiated by the degree

of hypoxia, with ALI typified by a ratio of PaO2/FiO2 of less than 300 and ARDS 200. Newer definitions

make an attempt at describing timing relative to a known insult and acknowledge that cardiac failure is

now often diagnosed without the aid of a pulmonary artery catheter. Most importantly, ALI is now part

of ARDS spectrum that is classified as mild, moderate, or severe if the PaO2/FiO2

ratio is between 200

and 300, between 100 and 200, and less than 100, respectively.45 Further, this gradation of disease is

consistent with therapy, wherein less acute patients might be successfully treated with noninvasive

ventilation, moderately ill patients with higher PEEP and possibly neuromuscular blockade, and

severely ill patients possibly requiring prone positioning therapy and extracorporeal support.

Figure 10-7. Volume–time scalar.

Figure 10-8. Pressure–time scalar with alternating ventilator-assisted breaths and spontaneous breaths.

Ventilator Management

Conventional Modes

The ventilatory standard of care was set forth as a result of an early study from the NIH ARDS Network.

By all metrics, the patients ventilated at 6 mL/kg had better outcome than those who received 12

mL/kg. Thus, as a result, the standard of care for ventilating ARDS patients became the use of low tidal

volumes, or low stretch therapy to diminish the incidence of ventilator-associated barotrauma.46

Rescue Therapies

Involving different ventilator modes share the fact that they achieve their end by raising mean airway

pressure while preventing elevation of peak airway pressure. How the disparate modes of ventilation

achieve this varies.

High PEEP. Ideally, to prevent atelectasis and avoid overdistension, PEEP is titrated to the safe window

between the lower and upper inflection points of the respiratory pressure–volume curve (Fig. 10-12).

The precise PEEP level to optimize outcome from ARDS, particularly in the face of low stretch

ventilation, has been the subject of several trials. The Assessment of Low Tidal Volume and Elevated

End-Expiratory Pressure to Obviate Lung Injury (ALVEOLI) trial did not demonstrate a benefit to adding

high PEEP to those ventilated with lower tidal volumes.47 Two other trials (LOVS and EXPRESS)

reached similar conclusions.48,49 A systematic review and meta-analysis showed no statistically

significant difference in hospital mortality overall but worse in those with severe established ARDS

treated with high PEEP.50

309

http://surgerybook.net/

Figure 10-9. Controlled mandatory ventilation. Breaths are volume-cycled, time-triggered, and flow-limited.

Perhaps the more valuable variable to consider preventing barotrauma and atelectasis is not the PEEP

level, but rather the transpulmonary pressure. Transpulmonary pressure is the difference between

airway pressure and pleural pressure. As airway pressure is normally slightly positive and pleural

pressure slightly negative, transpulmonary pressure is typically zero. If transpulmonary pressure were

negative, it would favor atelectasis, whereas a positive value would promote barotrauma and

hemodynamic compromise. In ICU patients, there are many forces extrinsic to the lung that can raise

pleural pressure and promote atelectasis. These include obesity, intra-abdominal hypertension, and

anasarca. Thus, in these circumstances, it might be beneficial to utilize higher PEEP levels. Although it

is not possible to measure pleural pressure directly, esophageal pressure is a reasonable surrogate.

Although titrating PEEP to esophageal pressure has not yet been proven clinically efficacious, it may in

the future.

Figure 10-10. Pressure controlled ventilation: Breaths are time-cycled, time–triggered, and pressure-limited.

310

http://surgerybook.net/

Figure 10-11. Pressure support (PS). Breaths are flow-cycled, patient–triggered, and pressure-limited. Pressure support breaths are

second and fourth breaths.

High-Frequency. Modes include high-frequency jet and oscillatory ventilation. Both utilize ultrafast

subtidal ventilation and can be employed for rescue or treatment of bronchopleural fistulae. In general,

heavy sedation and/or paralysis are required. The ventilator decouples oxygenation ventilation such

that unique settings are used to manipulate either one. The ventilatory circuit is heavy and cumbersome

in this mode and secretion clearance is problematic. Importantly, two large trials failed to show

improvement in ARDS patients, with one demonstrating an increased mortality in the high-frequency

group.51,52

Airway Pressure Release Ventilation (APRV). takes advan-tage of spontaneous breathing (and hence

the ability to keep patients more awake) with CPAP and a brief pressure release, generating a very high

inspiratory to expiratory (I:E ratio). The extremely short phase time of pressure release builds up

intrinsic PEEP to prevent atelectasis. There are increasing studies in the literature that demonstrate that

APRV is at least equivalent to low stretch ventilation in terms of safety and outcome.53,54 As APRV may

be better tolerated than low tidal volume ventilation, some would say this provides an advantage. There

is also some suggestion that routine use of APRV may decrease the incidence of and mortality from

ARDS.55

Figure 10-12. Pressure–volume curve with added PEEP. V, volume; P, pressure; LIP, lower inflection point; UIP, upper inflection

point. Ideal PEEP is set between LIP and UIP.

Volumetric Diffusive Respiration (VDR). Is a hybrid mode that combines convective gas delivery

similar to CPAP, diffusive gas delivery similar to oscillatory modes and a unique percussive mode to

promote secretion removal. As such, it has been used extensively in the burn patient population.

However, the technology is not widely available and large trials of use in rescue therapy of ARDS

patients have not been completed.

Extracorporeal Support. via extracorporeal membrane oxygenation (ECMO) was thought to provide a

beneficial effect to those patients in the European Conventional Ventilation or ECMO for Severe Adult

Respiratory Failure (CESAR) trial.56 However, it was apparent on further analysis that patients with

311

http://surgerybook.net/

severe ARDS benefit most from transfer to a center that had ECMO capabilities, whether or not the

therapy was actually used. Data from recent viral pneumonia epidemics are similarly confusing.

Regardless of the effect on mortality, it is apparent that ECMO may be thought of as a last ditch effort

to improve oxygenation, but should not be employed after lung fibrosis and irreversible organ failure

have transpired. For sure, current technology involving smaller access catheters, more portable

oxygenators and venovenous deployment has made ECMO far less morbid than initial experience with it

decades ago with prior mortality rates in excess of 50%. The definitive trial without crossover treatment

to validate the use of ECMO for severe ARDS has not been performed.

Ventilator Associated Events

Ventilator-associated pneumonia (VAP) is the most common life-threatening healthcare associated

infection, although there is some evidence that it is less morbid in surgical than in medical patients.57

VAP is associated with increased ICU and hospital length of stay and costs, and may often be

preventable. Nonetheless, it can be extraordinarily difficult to define, particularly in injured patients

who may be febrile and have pulmonary infiltrates for reasons other than pneumonia. Further, if

attempts are made to diagnose the entity predominantly by microbiologic criteria (using

bronchoalveolar lavage to culture >105 organisms), then the incidence of VAP may be overstated.

Thus, the CDC has worked with various critical care societies to define the surveillance definition of

ventilator-associated events (VAE) which are publically reportable and include probable and possible

VAP, which are not reportable. Briefly, VAE require patients to have a baseline period of more than 2

calendar days on the ventilator followed by a sustained decrement in oxygenation. Note there are no

longer radiographic criteria for VAE or VAP. An infection-related VAE requires a change in temperature

or white blood cell count and the initiation of a new antibiotic that is continued for 4 days. The

differentiation between possible and probable VAP involves positive cultures as previously prescribed.58

Liberation from Ventilation

Automated, computerized liberation protocols significantly shorten time on ventilator and success rate

compared to those directed by physicians. As opposed to the concept of “weaning” patients from

support, it is now apparent that the need for ventilation can be approached in a more binomial, on–off

approach. Thus, current practice is to complete daily spontaneous breathing trials (SBTs) of no longer

than 30 minutes to assess patients’ readiness for liberation from ventilation. Of course, these trials

should not be conducted in patients with hypoxia, hemodynamic instability or neurologic disease. SBTs

may be accomplished using a T-piece trial or CPAP with or without low PSV (<10 mm Hg), the latter

to enable ventilatory monitoring of work of breathing and respiratory effort. If patients develop

tachypnea, tachycardia, hypoxia, hypercapnia, and/or use accessory muscles during the SBT they are not

ready to be extubated. Typically, the parameter that predicts extubation success most accurately is the

rapid shallow breathing index or the tidal volume divided by the respiratory rate with a level of less

than 105 most predictive of success.59 Further, the requirement for airway support, as might be

expected in those with burns, head injury, or anasarca must be considered separately from the

requirement for ventilatory support. Checking for a cuff leak, although conceptually of benefit in

identifying patients with airway edema, may not be an accurate metric of those with a stable airway.

Trials in the medical literature suggest that a short course of steroids may benefit some to diminish

airway edema; these have not been replicated in surgical patients.60 Finally, some patients, such as

those with COPD, may benefit from extubation to noninvasive modes of ventilation such as bilevel

positive airway pressure (BiPaP) or CPAP. Additionally, the use of high-flow nasal cannula oxygen,

typically delivered at 40 L/min or greater may prevent reintubation.

Adjunctive Therapies

Inhaled Nitric Oxide and Prostaglandins. Inhaled Nitric Oxide (iNO) affords selective vasodilation of

the pulmonary vasculature without systemic effect. It improves oxygenation and lowers pulmonary

artery pressures. If used, methemoglobinemia can result, thus, monitoring should be conducted. Five

early meta-analyses showed improved oxygenation without a mortality benefit, with three

demonstrating increased renal failure.61 A more recent meta-analysis failed to show a mortality

advantage for iNO use for mild, moderate, or severe ARDS.62 Inhaled and/or intravenous prostacyclins

such as epoprostenol can also lower pulmonary artery pressures and improve oxygenation. A 2010

Cochrane review demonstrated no safety or outcome differences between iNO and epoprostenol;

however, at present, the latter is substantially cheaper. Although neither agent improves mortality, it

312

http://surgerybook.net/

seems reasonable to employ prostaglandins in ARDS patients with concomitant pulmonary

hypertension.63

Positional Therapy. Prone positioning therapy can be done in conjunction with many other therapies

previously described. It can be accomplished on a conventional ICU bed or one specially equipped to

provide this therapy. Either way, extreme caution must be taken to avoid tube and line dislodgement

and development of pressure-related skin complications. The ideal amount of time to keep a patient in

the prone position is of debate. Often, a patient is turned every shift or if a procedure is required. A

recent large study in French ICU patients demonstrated a significant improvement in mortality in ARDS

patients treating with prone positioning therapy.64 Similarly, a systematic review and meta-analysis

reinforced the mortality benefit of nearly 25% by prone positioning therapy for ARDS.65

Fluid Management. Another large trial conducted by the ARDSnet group Fluid and Catheter Treatment

Trial (FACTT) demonstrated that ARDS patients have better outcome if managed with conservative fluid

strategy that need not be guided by a PAC.38 Of course, these findings must be reconciled with

extensive data supporting improved outcome in septic patients who received more robust fluid

resuscitation.66,67

Steroids. Trials regarding the use of steroids in ARDS are old and heterogeneous. At present, the use of

steroids cannot be recommended, but future trials controlling for other factors may indicate otherwise.

Sedation and Paralysis. Recently released pain, agitation, and delirium guidelines underscore several

features regarding the sedation of ICU patients. First, most ventilated ICU patients benefit from light

sedation and continued emphasis on mobility (even if ventilated) and screening for delirium. If sedative

agents are required, a strategy that employs narcotic analgesics first is preferred. If additional agents

are necessary, there is substantial evidence that a strategy that relies heavily on use of benzodiazepines

will result in a higher incidence of delirium, time on ventilator, and greater costs.68 Ideal sedatives

should be easy to titrate, short acting with rapid onset, and without accumulation with prolonged use.

These sedatives should also have minimal adverse effects, minimal metabolism, and no active

metabolites. Traditionally used sedatives and analgesics morphine, propofol, and benzodiazepines have

well-known adverse effects, active metabolites and tend to accumulate with prolonged use. Fentanyl is

widely used in ventilated surgical ICU patients due to its excellent pain control, rapid onset, short

duration of action, and lack of active metabolites. Dexmedetomidine is a sedative with alpha 2-

adrenoreceptor agonist properties that is used as an alternative for analgesia and sedation in the surgical

ICU. Dexmedetomidine has been shown to decrease the incidence of delirium and decreased duration of

mechanical ventilation versus benzodiazepines while not being inferior to benzodiazepines and propofol

in maintaining comfort and sedation.69 Using sedation scales in the ICU such as the Richmond AgitationSedation Scale (RASS) provide a universal numerical scale to communicate between ICU team members

the target and actual sedation. Although we should not use long-term neuromuscular blockade (NMB) in

those with mild ARDS, a recent French study demonstrated a significant benefit to short-term use in

terms of mortality.70 A much larger propensity-matched retrospective study documented a 4.3%

absolute reduction in mortality with short-term use of NMB in severe ARDS.71

VENOUS THROMBOEMBOLISM PROPHYLAXIS IN THE ICU

Much attention has been focused on the prevention of venous thromboembolism (VTE) in hospitalized

patients, particularly in the United States, Canada, and the United Kingdom where national initiatives

are present. VTE prophylaxis is a quality metric promoted by such endeavors as the SCIP discussed

elsewhere in this text. The surgical ICU patient is a particular target for VTE and has unique features

rendering prophylaxis more challenging than it may be in a ward patient. VTE is a common cause of

preventable morbidity and mortality in critical care patients. A large amount of VTE complications

occur following surgery and injury.12,72 Further, the risk of bleeding in the multitraumatized, immobile,

and/or freshly postoperative patient may be heightened in the ICU patient. Finally, safe administration

of a chemoprophylactic agent as described below must take into account drug metabolism that may be

altered with renal or hepatic failure seen in ICU patients and desire to rapidly reverse the medication if

frequent procedures are needed.

313

http://surgerybook.net/

Table 10-7A Caprini Risk Assessment for Venous Thromboembolism (VTE) in

Surgical Patients

Incidence

The risk for developing asymptomatic lower extremity deep venous thrombosis (DVT) depends on the

assessment of general risk factors described below and the type of surgical patient. The incidence can

vary from less than 10% (in mobile, thin patients undergoing brief, minor surgery) to over 80% (in

those with spinal cord injury or elderly, obese hip fracture patients) in the absence of prophylaxis.

However, the true incidence may be unknown. It has become clear that upper extremity DVT (which is

not routinely screened) is more common than previously appreciated and may result in pulmonary

embolism at a higher rate than thought in the past,73 thus dramatically affecting VTE rates. The

incidence of fatal pulmonary embolism in surgical patients ranges from less than 1% in most undergoing

elective general surgery to as high as 5% to 8% in those receiving operations for hip fractures.74 The

projected VTE incidence of the surgical ICU would, thus, depend on the composition and the use of

screening for asymptomatic patients.

Risk Factor Scoring

Surgical critical care patients are at increased risk of VTE via Virchow triad of stasis, vessel wall injury,

and hypercoagulability. Acquired risk factors in the surgical critical care patients are a consequence of

surgery, trauma, immobilization, obesity, malignancy, advanced age, and invasive monitoring devices.

The American College of Chest Physicians (ACCP) 2012 guidelines

75 stratify surgical patients into four

risk categories: very low, low, moderate, and high. The most widely used tool to assist in classification

is the Caprini score (Table 10-7).76 A quick glance at the assessment would reveal that the vast majority

of surgical ICU patients over the age of 40 would fall into at least the moderate risk category; if a

central venous catheter were present, then they would be considered high risk.

Approaches to Prevention of VTE

Screening – Secondary Prevention

This strategy involves the early detection of subclinical venous thrombosis by regular screening tests.

These tests include Duplex ultrasonography and laboratory assays such as evaluation of d-dimer,

discussed in further detail elsewhere in this textbook. As most ICU patients will fall into moderate and

high risk categories for VTE (in whom d-dimer assays are universally elevated), there is virtually no

role for utilizing d-dimer acquisition for diagnostic purposes. Further, although many groups feel

differently and perform routine screening lower extremity Duplex ultrasonography, the authors believe

that there is little role for this strategy if screening upper extremity studies are not performed

concomitantly. Further, as will be discussed below, the safety of early administration of

chemoprophylactics is now being proven in even the highest-risk patients (e.g., neurotrauma).77

Table 10-7B Caprini VTE Percentage Risk Based on Score

314

http://surgerybook.net/

Treatment – Primary Prophylaxis

Interventions for the prevention of VTE include elastic stockings, intermittent pneumatic compression

devices, low-dose unfractionated heparin (UFH), low–molecular-weight heparin (LMWH), indirect

factor Xa inhibitors such as fondaparinux, direct thrombin and Xa inhibitors, aspirin, and inferior vena

cava (IVC) filters.

Mechanical Devices

Alone are utilized for surgical ICU patients at the lowest risk for VTE. Intermittent compression devices

may afford additional protection from VTE in those in high-risk groups if used in concert with chemical

agents.78

Chemical

7 UFH and LMWH. Few studies have performed a head-to-head analysis of these agents for VTE

prophylaxis in surgical ICU patients. A study of injured patients completed by Geerts and published in

199679 demonstrated that use of LMWH resulted in a significantly lower incidence of VTE (31% vs.

44%). This likely drove the 2009 ACCP recommendation that moderate and high risk ICU patients

should preferentially receive LMWH for prophylaxis (with a IA strength and level of recommendation),

However, in the interim, Dr. Geerts and colleagues performed a multicenter, RCT (PROTECT) that did

not demonstrate a difference in the development of DVT or proximal DVT in those who received LMWH

in the form of dalteparin compared to those administered twice-daily UFH.80 There was no difference in

bleeding rate, but a higher incidence of PE was noted in the UFH patients (and a nonsignificant increase

in heparin induced thrombocytopenia). This study, undoubtedly, drove the 2012 ACCP recommendation

that UFH or LMWH can be used interchangeably (2C recommendation).75 However, it is vital to

mention that the authors excluded injured, orthopedic and neurosurgical patients from this study. eerts

himself provides guidelines suggesting that ICU patients at high risk for VTE including injured and

major surgery patients should receive LMWH preferentially. UFH should be reserved for moderate-risk

ICU patients. High-risk patients at risk for bleeding should receive mechanical prophylaxis until the risk

for bleeding is abrogated, and then LMWH should be initiated.81 Literature suggests that this may be

substantially earlier than we might have thought in the past.82 Finally, in those with renal dysfunction,

consideration must be given to avoiding LMWH, or at least monitoring Xa levels, although there is little

literature to offer guidance in this regard.

Indirect Xa inhibitors include fondaparinux and have been studied extensively in the orthopedic (nonICU) and general surgery population. In both these populations, it may be more efficacious than LMWH

without an increase in bleeding risk.75 How this extrapolates to the ICU population is unknown;

however, it may be helpful in those with HIT.

There is even less experience using oral Xa (rivaroxaban and apixaban) and thrombin II (dabigatran)

inhibitors in the ICU. These agents may be less than helpful because of a long half-life and inability to

reverse (particularly dabigatran that may be best removed by hemodialysis). Nonetheless, we should

appreciate that all three agents had a lower rate of significant bleeding such as intracranial hemorrhage

in outpatients who were being treated for atrial fibrillation. Rivaroxaban is metabolized renally and

apixaban both hepatically and renally. Due to similar concerns of time and ability to reverse, the use of

aspirin and warfarin may be limited in the ICU.

IVC Filter. At present, the current ACCP guidelines do not recommend insertion of prophylactic IVCF in

those who cannot be anticoagulated. They recommend initiating chemoprophylaxis as soon as is

practical.75

TRANSFUSION OF THE ICU PATIENT

8 Transfusion of packed red blood cells, ostensibly to promote oxygen delivery, is extraordinarily

315