Bongard Frederic , Darryl Sue. Diagnosis and Treatment Critical Care

Подождите немного. Документ загружается.



TRANSFUSION THERAPY

serve to promote rational use of blood products for the ben-

efit of patients as well as protecting the limited supply of

blood from waste. These trials have clearly influenced trans-

fusion practices in well-defined clinical situations. There are

many situations, however, where no empirical data exist, and

the clinician must determine the benefits of transfusion on

an individualized basis. Massive repeated transfusion in

patients with uncorrectable vascular defects or who are ter-

minally ill—and platelet and plasma transfusions for

patients without demonstrated response to such

transfusions—may deplete the blood supply without sub-

stantially improving the outcome for those patients.

The medical team caring for a patient with massive

uncontrollable bleeding or a terminal illness should make

every effort to discuss the limits of care with the patient and

family members, to establish long-term treatment goals and

expectations, and to decide when continued blood transfu-

sion is no longer of benefit to the patient. In addition, family

members should be strongly encouraged to donate blood to

help replace some of the units used. Ineffective therapy

should not be given prophylactically (eg, daily platelet trans-

fusions in patients with consumptive thrombocytopenia

without bleeding or plasma therapy in patients with multiple

severe coagulation deficiencies not corrected with large doses

of plasma).

REFERENCES

Boshkov LK: Transfusion-related acute lung injury and the ICU. Crit

Care Clin 2005;21:479–95. [PMID: 15992669]

Circular of Information for the Use of Human Blood and Blood

Components; from www.fda/gov/cber/dglns/crclr.pdf.

Corwin HL et al: Efficacy and safety of epoetin alfa in critically ill

patients. N Engl J Med 2007;357:965–976. [PMID: 17804841]

Davenport RD: Pathophysiology of hemolytic transfusion reactions.

Semin Hematol 2005;42:165–8. [PMID: 16041666]

Drews RE: Critical issues in hematology: Anemia, thrombocytope-

nia, coagulopathy, and blood product transfusions in critically ill

patients. Clin Chest Med 2003;24:607–22. [PMID: 14710693]

Gajic O et al: Pulmonary edema after transfusion: How to differ-

entiate transfusion-associated circulatory overload from

transfusion-related acute lung injury. Crit Care Med 2006;34:

S109–13. [PMID: 16617253]

Goodnough LT, Shander A, Brecher ME: Transfusion medicine:

Looking to the future. Lancet 2003;361(9352):161–9. [PMID:

12531595]

Goodnough LT: Risks of blood transfusion. Crit Care Med

2003;31(suppl 12):S678–86. [PMID: 14724466]

Hebert PC, Tinmouth A, Corwin HL: Controversies in RBC transfu-

sion in the critically ill. Chest 2007;131:1583–90. [PMID:

17494811]

Hebert PC et al: A multicenter, randomized, controlled clinical trial

of transfusion requirements in critical care. N Engl J Med

1999;340:409–17. [PMID: 9971864]

Hebert PC: Clinical consequences of anemia and red cell transfusion

in the critically ill. Crit Care Clin 2004;20:225–35. [PMID:

15135462]

Hellstern P: Solvent/detergent-treated plasma: Composition, effi-

cacy, and safety. Curr Opin Hematol 2004;11:346–50. [PMID:

15666659]

Hill SR et al: Transfusion thresholds and other strategies for guiding

allogeneic red blood cell transfusion. Cochrane Database Syst Rev

2002;2:CD002042. [PMID: 12076437]

Lauzier F: Fresh frozen plasma transfusion in the critically ill. Crit

Care Med 2007;35:1655–9. [PMID: 17522577]

McIntyre L et al: Is a restrictive transfusion strategy safe for resusci-

tated and critically ill trauma patients? J Trauma 2004;57:563–8.

[PMID: 15454803]

Napolitano LM, Corwin HL: Efficacy of red blood cell transfusion in

the critically ill. Crit Care Clin 2004;20:255–68. [PMID:

15135464]

Ohene-Frempong K: Indications for red cell transfusion in sickle cell

disease. Semin Hematol 2001;38(1 suppl 1):5–13. [PMID:

11206961]

Rao SV et al: Relationship of blood transfusion and clinical out-

comes in patients with acute coronary syndromes. JAMA

2004;292:1555–62. [PMID: 15467057]

Schiffer CA et al: Platelet transfusion for patients with cancer:

Clinical practice guidelines of the Society of Clinical Oncology. J

Clin Oncol 2001;19:1519–38. [PMID: 11230498]

Shander A: Anemia in the critically ill. Crit Care Clin 2004;20:

159–78. [PMID: 15135458]

Silliman CC, Ambruso DR, Boshkov LK: Transfusion-related acute

lung injury. Blood 2005;105:2266–73. [PMID: 15572582]

Klein HG, Spahn DR, Carson JL: Red blood cell transfusion in clini-

cal practice. Lancet 2007;370(9585):415–26. [PMID: 17679019]

Stanworth S et al: Prophylactic platelet transfusion for haemorrhage

after chemotherapy and stem cell transplantation. Cochrane

Database Syst Rev 2004;18:CD004269. [PMID: 15495093]

Stubbs JR: Alternatives to blood product transfusion in the critically

ill: Erythropoietin. Crit Care Med 2006;34:S160–9. [PMID:

16617261]



Critically ill patients are almost always given many different

drugs. For example, patients may require broad-spectrum

antibiotics, vasopressor agents, and antiarrhythmics. At the

same time, clinicians often must give sedatives and muscle

relaxants to patients in order to tolerate mechanical ventila-

tion, and there is strong evidence that ICU outcome is

improved by empirical prophylaxis for GI bleeding with pro-

ton pump inhibitors or other antacids and prophylaxis for

deep venous thrombosis with heparin.

One estimate is that ICU patients receive twice as many

different drugs as other hospitalized patients. It is no wonder,

therefore, that medication errors, a major focus of hospital

safety improvement, are increased in ICU patients. In fact, an

error in preparation, dosage, or rate of infusion has been

found in as many as 10% of intravenous medications.

Furthermore, complex changes in pharmacokinetics and

pharmacodynamics are seen frequently in these patients,

necessitating consideration of dosage, interaction, and organ

function.

Pharmacokinetics is the movement of a drug through the

body over time and is defined by the following: absorption,

distribution, metabolism, and elimination—basically, how

the patient affects a drug. Pharmacodynamics is the relation-

ship between drug concentration and drug effect—or how a

drug affects the patient. Both pharmacokinetic and pharma-

codynamic changes compared with baseline occur in critical

illness and the management of patients in the ICU.

PHARMACOKINETIC PARAMETERS

The most important pharmacokinetic parameters are clear-

ance and volume of distribution. The disposition of a drug,

once it has entered the body, depends on both clearance (CL)

and the volume of distribution (V

). Clearance is propor-

tional to the rate at which a drug is eliminated from the body.

Volume of distribution is a theoretical volume that relates

the concentration in the plasma to the total amount of drug

in the body. The elimination half-life (t

1/2

) depends on the

preceding independent variables and is described by the fol-

lowing mathematical relationship:

1/2

= (0.693 × V

/CL)

Hence either decreased clearance or an expanded volume

of distribution will result in an increased pharmacologic

half-life. A practical understanding of this mathematical

relationship is essential for developing optimal dosing regi-

mens in the critically ill patient.

PHARMACOKINETIC CONSIDERATIONS



Absorption

Drug absorption is influenced by a variety of factors, includ-

ing the site of absorption, the amount metabolized before

reaching the systemic circulation (“first-pass” effect for orally

administered drugs), and drug interactions. In critical illness,

the site of absorption is of utmost importance. The extent of

oral absorption—bioavailability—may be diminished as a

result of low cardiac output or shunting of blood from the

mesentery to the peripheral circulation. In patients who are in

the ICU for a longer duration, intestinal atrophy and motility

dysfunction may play a role. While highly encouraged in suitable

patients, enteral nutritional support may cause inadvertent mal-

absorption of some orally administered medications. Therefore,

therapeutic “failures” may be due to inadequate bioavailability

rather than absence of effect at the intended receptor site.Acutely

ill patients may have decreased perfusion of sites of parenteral

administration of drugs. These patients may have poor or unreli-

able absorption of subcutaneously or intramuscularly adminis-

tered medications. In general, the intravenous route is preferred

for critical medications, and oral, subcutaneous, or intramuscu-

lar administration should be avoided.

Pharmacotherapy

∗

Darryl Y. Sue, MD

∗

Jennifer H. Cupo Abbott, PharmD, and Maria I. Rudis, PharmD,

were the authors of this chapter in the second edition.



PHARMACOTHERAPY



Distribution

The distribution of drugs in the body depends on factors such

as blood flow, body composition, and plasma protein binding

(Table 4–1). For a single-compartment model, initial plasma

concentration (before elimination) equals the dose of drug

given divided by the apparent volume of distribution. In nor-

mal subjects, the volume of distribution of a drug is generally

well described and usually is related to some measure of

patient size (eg, total body water for a water-soluble drug and

extravascular volume for a large molecule). However, volume

of distribution is strongly influenced by drug lipid solubility,

protein binding, and intra- versus extracellular partitioning.

In critical illness, fluid overload can increase a patient’s

volume of distribution significantly, and insufficiently

effective drug concentrations may result. For drugs with

extensive tissue distribution (eg, digoxin), the volume of

distribution is approximately 500 L, or several times actual

body weight. Thus it is unlikely that changes in fluid status

can affect the distribution of digoxin significantly. For

drugs apparently distributed largely into extracellular

water, however, drug concentrations can be affected. It is

not uncommon for a critically ill patient to gain several

kilograms of extracellular water (ECW). Because this space

is roughly 20% of body weight, or about 12 L in a 60 kg

adult, gaining even 2 L of ECW would reduce expected con-

centration of such a drug by 15–20%. A drug such as gen-

tamicin exhibits concentration-dependent killing of

bacteria. A peak level of 25 μg/kg (dose = 5–7 mg/kg, nor-

mal V

= 0.25 L/kg) is expected. However, V

in critically ill

patients may be as high as 0.6 L/kg, resulting in a much

lower than expected serum level.

Body composition and the lipophilicity of a given drug

are also important factors to consider. In general, lipophilic

agents such as diazepam readily distribute into fat. For these

agents, dosing should be based on total body weight. Since

most agents used in the ICU setting are not lipophilic, such

as pressors and most antimicrobials, it may be more accu-

rate to use ideal body weight in dosing calculations.

However, an increasingly important question is how to

adjust medication dosages for obese patients because the

volume of distribution likely will change. Considerations

include unexpected changes in the proportion of adipose

tissue compared with total body weight, thereby changing

estimates of total body and extracellular water, and some-

times the difficulty in obtaining accurate scaled weights.

Analgesics, including opiates, and sedatives should be

titrated to desired clinical effect in the absence of studies of

weight-adjusted dosing. Antibiotic dosing should be guided

by therapeutic drug monitoring when possible, adjusting for

both desired antimicrobial effect and toxicity. Low-

molecular-weight heparin is a special situation because the

effect when dosed as milligrams per kilogram is no longer

predictable in the obese patient; in this case, measurement

of factor Xa activity may be useful.

Protein binding is another important determinant of dis-

tribution. Only the unbound fraction of a drug can diffuse or

be transported into tissues. Thus the influence of protein

binding is a limiting factor in drug distribution for highly

bound drugs (see Table 4–1). Serum pH, often abnormal in

critically ill patients, also may affect the degree of protein

binding of drugs. Phenytoin is a highly protein-bound drug

(~90%). The remaining 10% circulates as unbound or “free”

drug and is the fraction responsible for the pharmacologic

effect. If a patient with normal albumin has a phenytoin level

of 12 mg/L, the free fraction would be approximately 1.2 mg/L.

However, with decreased serum albumin, as is found in

patients with CNS trauma or those with end-stage liver dis-

ease, less serum protein is available to bind phenytoin. The

laboratory reports total phenytoin concentrations, which

includes bound and unbound drug. In a patient with hypoal-

buminemia, the total phenytoin level will be unchanged, but

the percentage of “free” drug (ie, pharmacologically avail-

able) will increase. For example, a patient with a serum albu-

min of 2 g/dL and a measured phenytoin plasma concentration

meas

) of 12 mg/L has more free drug, resulting in an adjusted

phenytoin plasma concentration (C

adj

) of 24 mg/L.

Drug Protein-Bound

Amphotericin B 90–95%

Ceftriaxone 93–96%

Chlordiazepoxide 94–97%

Clindamycin 93%

Diazepam 84–98%

Erythromycin 96%

Ethacrynic acid 95%

Furosemide 91–99%

Haloperidol 90–92%

Heparin >90%

Hydralazine 90%

Lorazepam 90%

Midazolam 94–97%

Nafcillin 70–90%

Nifedipine 89–92%

Oxacillin 89–94%

Phenytoin 90%

Prochlorperazine 90%

Rifampin 84–91%

Vecuronium 60–90%

Verapamil 90%

Table 4–1. Protein binding.



CHAPTER 4

The uremia accompanying renal failure also displaces

phenytoin from its binding sites because endogenous com-

petitors for binding accumulate. It is important to recognize

that malnourished patients and those with renal failure

will have a lower than normal “therapeutic range” for

phenytoin. For this reason, it is clinically more relevant to

monitor free phenytoin concentrations or to use these values

in calculation of dosage adjustments.



Drug Clearance (Elimination)

With limited exceptions, most pharmacologic agents are

eliminated either renally or hepatically, but drug clearance is

considered the effect of all pathways of elimination taken

together. Since multiorgan dysfunction is commonly

encountered in critically ill patients, drug accumulation and

toxicity are of concern. Specific dosage adjustment is often

required in the setting of renal or hepatic impairment. Some

of the most frequently used agents with predominantly renal

elimination are listed in Table 4–2. Most antimicrobials,

including aminoglycosides, vancomycin, beta-lactams, and

fluoroquinolones, are eliminated primarily via the kidneys.

Although some dosage adjustment is needed when these

antimicrobials are used in critically ill patients with renal

failure, studies have found both under- and overdosing of

antimicrobial agents because of renal function considera-

tions. Importantly, the initial concentration of an antimicro-

bial drug is related to its dose and volume of distribution, not

its elimination. Therefore, the “first dose” of an antimicrobial

need not be adjusted for renal failure, only subsequent doses.

Failure to recognize this concept leads to a delay in achieving

desired therapeutic levels.

Other drugs with primarily renal elimination used in the

critical care setting are the low-molecular-weight heparins.

Currently, dosage of enoxaparin is reduced 25–50% for

patients with a creatinine clearance of less than 30 mL/min,

but not for mild or moderate renal insufficiency. Some prac-

titioners advocate monitoring factor Xa activity for these

patients. Some drugs have mixed routes of elimination and

require adjustment for both renal and hepatic function for

proper dosing.

Renal Dysfunction

Medications used in the ICU are often eliminated by the kid-

neys, and dosages are adjusted in the face of renal insuffi-

ciency. Common causes of renal insufficiency in the ICU

include chronic kidney disease, acute renal failure from

shock or hypoperfusion, exposure to nephrotoxic drugs, and

obstructive uropathy.

Several studies have demonstrated that dosing of renally

excreted drugs (cleared by glomerular filtration) is

improved when creatinine clearance rather than serum cre-

atinine concentration alone is used to estimate renal func-

tion. This is especially true in the elderly and those who are

underweight, in whom a “normal” serum creatinine concen-

tration (0.8–1 mg/dL) may be associated with significantly

reduced renal function. Less commonly, patients with rhab-

domyolysis produce more than the expected amount of cre-

atinine; in these patients, elevated serum creatinine may not

indicate decreased glomerular filtration rate (GFR). Rarely,

some medications interfere with creatinine secretion, fur-

ther dissociating the relationship between serum creatinine

and GFR.

When the serum creatinine level is known, an estimate of

creatinine clearance can be obtained to assist with dosage

adjustment using the Cockroft-Gault equation:

(For females, multiply the numerator by a factor of 0.85.)

Although this equation and several others are used fre-

quently in ICU patients and generally are better estimates

than serum creatinine alone, some studies demonstrate that

they are far from perfect. In fact, even short-term collections

() )140

−×

age weight (in kg

serumcreatinine

Alb

mg/L

(0.2)(

ad j

meas

(.)( ) .





02 01

22.0 g/dL)

mg/L

 

Table 4–2. Drugs with primarily renal elimination.

Antimicrobials

Acyclovir Ciprofloxacin Meropenem

Amikacin Fluconazole Penicillin G

Ampicillin Flucytosine Piperacillin

Cefazolin Ganciclovir Ticarcillin-clavulanate

Cefepime Gatifloxacin Tobramycin

Cefotetan Gentamicin Trimethoprim-sulfamethoxazole

Cefoxitin Imipenem-cilastatin Vancomycin

Ceftazidime Levofloxacin

Ceftizoxime

Antihypertensives

Diazoxide

Methyldopa

Nitroprusside

Antiarrhythmic agents

Bretylium

Digoxin

Procainamide

Miscellaneous drugs

Ranitidine

Pancuronium



of urine for calculation of creatinine clearance are less accu-

rate than 12- or 24-hour collections, probably because of

variability in GFR over time.

For patients requiring hemodialysis, it is essential to

know the extent to which drugs can be removed by dialysis.

Knowledge of pharmaceutical properties such as

hydrophilicity, molecular weight, plasma protein binding,

and volume of distribution (Table 4–3) can help to distin-

guish agents that are dialyzable. Low-molecular-weight,

water-soluble drugs with low protein binding are highly

dialyzable. If they have a small volume of distribution, then

an appreciable amount of the drug can be eliminated by

dialysis. Conversely, drugs with extensive tissue distribution

such as digoxin or the calcium channel blockers and highly

lipophilic drugs are not affected by dialysis. Other impor-

tant considerations for drug clearance include the duration

and type of dialysis. Short dialysis sessions are less likely to

remove significant amounts of drug. New forms of dialysis

are more efficient and remove drugs previously thought

to be minimally dialyzable. Hemodialysis with high-flux

filters removes significant amounts of vancomycin, previ-

ously considered not to require replacement for drug lost to

dialysis. Continuous renal replacement therapy is a much

more efficient process than conventional hemodialysis

and equates to a clearance of approximately 30 mL/min for

creatinine, but not clearly comparable for other drugs. It

is essential to review the medication regimen closely

when patients are dialyzed with high-flux filters or are

switched from hemodialysis to continuous renal replace-

ment because the dosages of some drugs will have to be

increased.

Hepatic Dysfunction

Liver failure is a common problem in the ICU and may be

due to chronic liver disease (eg, cirrhosis or hemochromato-

sis) or acute liver disease (eg, acute hepatitis, drug-induced

hepatitis, alcohol, or toxins). Some drugs require dosage

adjustment in hepatic insufficiency (Table 4–4). However,

most commonly used tests of “liver function” describe the

degree of liver damage and not the liver’s capacity for drug

elimination. Determining the degree of hepatic dysfunction

is difficult because no quantitative equations exist.

Laboratory tests of hepatic synthetic function (eg, prothrom-

bin time, serum albumin, and conjugated bilirubin) are most

predictive of drug elimination. For drugs metabolized by the

liver, the route of metabolism is important in determining

the effects of liver disease on drug clearance. Some enzyme

systems are remarkably well preserved even in end-stage liver

disease. Drugs such as lorazepam that are metabolized pri-

marily by conjugation with glucuronic acid are minimally

affected in cirrhosis, so little dosage adjustment is required.

For drugs whose clearance depends on oxidative metabolism

(eg, metronidazole, theophylline, opioids, and sedative-

hypnotics), cirrhosis reduces their elimination. Generally

speaking, acute liver disease (eg, hepatitis) does not alter

drug clearance significantly.

Congestive heart failure may cause hepatic congestion

and decreased hepatic blood flow and decrease hepatic elim-

ination of drugs. The hepatic clearance of theophylline, for

example, is markedly reduced in patients with congestive

heart failure (CHF). Patients with chronic liver disease fre-

quently have hypoalbuminemia, thereby reducing the

amount of drug protein binding. For such highly protein-

bound drugs as phenytoin, the proportion of free unbound

drug rises with hypoalbuminemia.

PHARMACOTHERAPY

Table 4–3. Drugs significantly removed by hemodialysis.

Antimicrobials

Aminoglycosides Ceftizoxime Meropenem

Ampicillin Chloramphenicol Metronidazole

Cefazolin Ciprofloxacin Penicillin G

Cefepime Gatifloxacin Piperacillin

Cefotaxime Imipenem-cilastatin Quinupristin-

Cefotetan Levofloxacin dalfopristin

Cefoxitin Linezolid Trimethoprim-

Ceftazidime sulfamethoxazole

Antihypertensives

Diazoxide

Methyldopa

Nitroprusside

Antiarrhythmic agents

Bretylium

Digoxin

Procainamide

Miscellaneous drugs

Ranitidine

Pancuronium

Table 4–4. Drugs requiring dosage adjustment in severe

hepatic insufficiency.

Analgesics Antimicrobials

Acetaminophen Cefoperazone

Opioids Ceftriaxone

Salicylates Chloramphenicol

Antiarrhythmics Clindamycin

Lidocaine Erythromycin

Quinidine Isoniazid

Verapamil Metronidazole

Anticonvulsants Nafcillin

Phenobarbital Rifampin

Phenytoin Sedative-hypnotics

Antihypertensives Chlordiazepoxide

Hydralazine Diazepam

Labetalol Midazolam

Nitroprusside Miscellaneous

Haloperidol

Theophylline



CHAPTER 4



Therapeutic Drug Monitoring

Agents with a narrow therapeutic index have only a small dif-

ference between serum drug concentrations that produce

therapeutic and toxic effects, and monitoring of serum drug

concentrations is recommended (Table 4–5). Examples of

agents with low therapeutic indices are the aminoglycosides,

digoxin, theophylline, and phenytoin.

In many ICUs, routine blood samples are collected at a set

time. For therapeutic drug monitoring, this may not be

acceptable because the time the sample is drawn must be

related to the time since the last dose of the drug was given.

First, for some drugs, an estimate of peak concentration is

desired. This level should be obtained after distribution of the

dose into the volume of distribution is achieved. For example,

digoxin levels should be drawn about 4–6 hours after admin-

istration in order to distribute into its very large V

. If peak

aminoglycoside or vancomycin levels are sought, these are

usually reached at 30 minutes to 1 hour after administration.

Second, sometimes the “trough,” or lowest, value before

administration of the next dose is wanted. Obviously, the

sample is drawn just prior to administration. Finally, for

many drugs, dosing is predicted by formulas or nomograms

that use serum levels at designated times (eg, aminoglycosides

and vancomycin), such as 4–10 hours after dosing. Drug dis-

tribution is also of concern following dialysis. It is important

to allow at least 3 hours to elapse after dialysis to obtain drug

levels to allow for redistribution of drug from other tissues

into the main compartment (eg, intravascular space). This

is illustrated also in the case of hemodialysis for a toxic

ingestion of lithium. A lithium level of 10 meq/L (therapeutic

level is 0.5–2 meq/L) obtained before dialysis may decrease to

1 meq/L immediately after hemodialysis. However, a third

level obtained 3–4 hours after dialysis may rebound to a toxic

level of 5 meq/L, showing evidence of redistribution from the

CNS back into the main compartment. This indicates the

need for longer or more frequent hemodialysis.

Phenytoin serum levels are often used to adjust dosing.

Phenytoin is eliminated by first-order kinetics at low serum

levels, but elimination is saturable at higher levels, even

within the therapeutic range. Therefore, at these levels, small

increases in dosing may result in unexpectedly high levels.

Ethanol is eliminated by alcohol dehydrogenase and obeys

zero-order kinetics; thus a constant fall in serum level with

time is expected, usually 30-40 mg/dL per hour.



Drug Interactions

Given the number of drugs prescribed for critically ill

patients, the potential for drug interactions is high. Drug

interactions may occur as a result of pharmacodynamic, phar-

maceutical, or pharmacokinetic effects. Pharmacodynamic

interactions result from the drugs actions and may enhance or

antagonize a drug’s effects. Pharmaceutical interactions can

result from a number of causes, one of which is the relationship

between two drugs. The most striking interactions are phar-

macokinetic, which occur when one drug affects the absorp-

tion, distribution, or clearance of another.

Pharmacodynamic Interactions

Pharmacodynamic drug interactions can result in synergis-

tic, additive, or antagonistic pharmacologic effects. A benefi-

cial additive effect would be observed in a patient with

poorly controlled hypertension who receives a second anti-

hypertensive agent from a different class and then achieves

optimal blood pressure control. Synergistic combinations are

noted when the resulting pharmacologic effect with combi-

nation therapy is greater than the expected sum of drug

effects. This phenomenon occurs infrequently and is best

described for antimicrobial combinations. A beta-lactam

antimicrobial (eg, piperacillin or ceftazidime) in combina-

tion with an aminoglycoside may be more effective than a

beta-lactam alone and results in a lower incidence of

acquired bacterial resistance in the treatment of infections

with aerobic gram-negative organisms. On the other hand,

antagonism may be encountered when beta-blockers reverse

the pharmacologic benefit of beta-agonists in patients with

chronic obstructive pulmonary disease (COPD). While some

beta-blockers such as atenolol are more cardioselective at

lower doses, they still have the potential to antagonize bron-

chodilators such as albuterol and salmeterol. The concomitant

use of antimicrobials from the same class also carries the

potential for antagonism. For example, some beta-lactams

induce production of beta-lactamase. The combination of a

strongly inducing beta-lactam with a labile compound

Drug Therapeutic Range

Amikacin Peak: 25–35 mg/L

Trough: <10 mg/L

Amiodarone 0.8–2.8 mg/L

Gentamicin, tobramycin Peak: 8–12 mg/L

Trough: <1 mg/L

Digoxin 1–2 μg/L

Lidocaine 1–5 mg/L

Phenobarbital 10–30 mg/L

Phenytoin 10–20 mg/L

Procainamide 4–8 mg/L

-Acetylprocainamide <30 mg/L

Salicylates 100–300 mg/L

Theophylline (in COPD) 8–10 mg/L

Vancomycin Trough: 5–15 mg/L

Table 4–5. Therapeutic ranges for drugs commonly used

in critical care.



PHARMACOTHERAPY

(eg, piperacillin) for the treatment of infections owing to

Enterobacter species has been shown to produce antagonism

in vitro and in animal models of infection. Hence double

beta-lactam combinations that include a strong inducer

should be avoided.

Pharmaceutical Interactions

Pharmaceutical interactions may be caused by drug incompat-

ibilities or drug adsorption to catheters and to intravenous

administration materials. For example, intravenous adminis-

tration of nitroglycerin requires special equipment to decrease

the likelihood of adsorption. The complexity of drug regimens

in the critically ill patient coupled with limited intravenous

access makes intravenous drug compatibility a significant

issue. Although a great deal is known about the compatibility

of drug combinations, there are still many potential combina-

tions for which no such information is available.

Pharmacokinetic Interactions

Although pharmacokinetic interactions occur as a result of

alterations in drug absorption, distribution, metabolism, or

elimination, the effects on drug metabolism are the most

clinically significant. A commonly seen absorption interac-

tion occurs when fluoroquinolones are administered con-

comitantly with antacids, causing decreased quinolone

bioavailability. Similarly, enteral feeding should be withheld

2 hours before and after the administration of oral phenytoin

formulations because of the decreased and delayed absorp-

tion of phenytoin that occurs.

Drug interactions owing to altered distribution also may

occur. When two drugs compete for binding sites on plasma

proteins or tissues, the unbound or free serum concentration

of one or both drugs may increase. Although this theoreti-

cally may increase a drug’s effect, the enhanced pharmaco-

logic effect is usually transient because more unbound drug

is now available for elimination by the liver and kidney. Thus

the clinical significance of protein-binding displacement

interactions is minimal unless there is concomitant hepatic

or renal disease. However, warfarin and phenytoin may be

transiently displaced by a number of drugs.

Pharmacokinetic drug interactions are frequently due to

altered metabolism involving the cytochrome P450 (CYP)

enzyme system, which is largely responsible for oxidative

metabolism of drugs by the liver. These enzymes are a super-

family of microsomal drug-metabolizing enzymes that

degrade endogenous substances, chemicals, toxins, and med-

ications. The primary ones responsible for drug metabolism

are CYP3A4, CYP2D6, CYP1A2, and CYP2C. Examples of

commonly used drugs that are inducers and inhibitors of

CYP are shown in Table 4–6. The most potent drugs likely to

be encountered are phenobarbital, phenytoin, and rifampin,

with subsequently more rapid metabolism and lower serum

levels for cimetidine, phenytoin, theophylline, warfarin, cor-

ticosteroids, and quinidine. Cigarette smoking and chronic

ethanol use also increase CYP activity. This explains why

alcoholics may require surprisingly high doses of sedatives

(eg, diazepam and midazolam) or analgesics. CYP induction

does not occur immediately, but usually takes at least several

days. Therefore, effects of CYP may be immediate (eg, in a

Drug Affected CYP Inducer/Inhibitor Effect

Benzodiazepines (alprazolam) Inhibitor

Azole antifungal (fluconazole, itracona

zole, ketoconazole)

Increased benzodiazepine concentration

Cyclosporine Inducer

Rifampin, rifabutin, phenobarbital,

phenytoin

Inhibitor

Erythromycin, azole antifungal

Decreased cyclosporine levels

Increased cyclosporine levels

Theophylline Inhibitor

Fluoroquinolone

Inducer

Cigarette smoking

Increased theophylline levels

Decreased theophylline levels

Warfarin Inducer

Phenobarbital

Inhibitor

Cimetidine

Decreased warfarin levels

Increased warfarin levels

Table 4–6. Examples of cytochrome P450 (CYP) induction or inhibition by drugs.



CHAPTER 4

chronic smoker) or delayed (eg, after starting a potential

CYP inducer in the hospital). Drugs that inhibit CYP systems

may behave differently than those that are inducers because

the former can act immediately on CYP. The most common

CYP inhibitors in the ICU are allopurinol, amiodarone,

cimetidine, erythromycin, and fluconazole.

The importance of CYP induction and inhibition depend

on the therapeutic indices of the drugs whose metabolism

are affected. The narrower the therapeutic window (level

providing therapeutic effect compared with the level result-

ing in toxic effect), the greater is the likelihood that a CYP

inhibitor will lead to toxicity or an inducer will cause sub-

therapeutic levels.



Adverse Effects & Drug Toxicities

Drugs may adversely affect all organ systems, but the kidney,

liver, heart, CNS, and vascular system are most frequently

affected. In critically ill patients with multiple medical prob-

lems, it can be quite difficult to isolate drug toxicity as the

sole cause of organ failure. Some drug toxicities are dose-

dependent, so attention to dosing and elimination is impor-

tant, as well as to drug interactions that may increase drug

levels (eg, inhibition of cytochrome P450 enzymes). Other

adverse effects are allergic and depend on the host response

and prior exposure. For some adverse effects, the patient may

be more susceptible for genetic or other reasons (long QT

syndrome).

Nephrotoxicity

The most common causes of drug-induced nephrotoxicity

are listed in Table 4–7. Nephrotoxicity in critically ill patients

may be due to drug-induced causes or to hypoperfusion.

Because the mortality rate for ICU patients with acute renal

failure approaches 80%, efforts should be directed at remov-

ing all potential causes of nephrotoxicity. Adequate fluid

resuscitation and maintenance of renal perfusion are of

paramount importance for preventing prerenal acute renal

failure. Appropriate intravascular volume status and pre-

treatment with N-acetylcysteine or sodium bicarbonate

decrease the risk of nephrotoxicity from radiocontrast

agents.

Despite adequate preventive measures, up to 20% of all

cases of acute renal failure may be associated with drug toxi-

city. Drug-induced toxicity may take the form of acute tubu-

lar necrosis, interstitial nephritis, or glomerulonephritis. Of

those drugs associated with acute tubular necrosis, the most

notable are the aminoglycosides and amphotericin B. With

once-daily dosing of aminoglycosides (5–7 mg/kg per day)

and proper therapeutic drug monitoring, the incidence of

acute tubular necrosis is reduced significantly. Novel ampho-

tericin B formulations as well as the increased use of other

antifungals (eg, azoles and echinocandins) reduce the risk of

nephrotoxicity. Interstitial nephritis and glomerulonephritis

are due to hypersensitivity reactions or immune-complex

formation. The most common drugs leading to interstitial

nephritis are antibiotics, even though the most likely culprit,

methicillin, is no longer used.

Hepatotoxicity

While a number of drugs have been associated with altered

liver function tests, these changes are usually reversible on

discontinuation of the offending agent. Since acute hepatic

injury is classified according to morphology, drug-induced

hepatic injury may cause either direct hepatocellular necro-

sis, cholestasis, or a mixed presentation of both (Table 4–8).

Some drug combinations such as rifampin and isoniazid,

amoxicillin and clavulanic acid, and trimethoprim and sul-

famethoxazole also may increase the possibility of hepato-

toxic reactions. This may occur because one agent alters the

metabolism of the other, leading to the production of toxic

metabolites. Phenytoin induces both hepatic necrosis and

cholestasis in association, producing an immune response

manifested by a rash, eosinophilia, atypical lymphocytosis,

and serum IgG antibodies against phenytoin.

An increasingly important source of drug-induced hepa-

totoxicty is the use of herbal drugs. These may not be disclosed

Acute tubular necrosis

Acyclovir

Aminoglycosides

Amphotercin B

Iodinated contrast dyes

Foscarnet

Pentamidine

Interstitial nephritis

Allopurinol

Cimetidine

Furosemide

Methicillin

Phenytoin

Rifampin

Thiazides

Trimethoprim-sulfamethoxazole

Vancomycin

Glomerulonephritis

ACE inhibitors

Gold salts

Hydralazine

Penicillamine

Rifampin

Renal hemodynamics

ACE inhibitors

Cyclosporine

NSAIDs

Tacrolimus

Table 4–7. Nephrotoxic drugs.



PHARMACOTHERAPY

by patients without specific questioning. Toxicity may be

hepatocellular or cholestatic in nature. Some herbs may

inhibit or induce the CYP system (eg, St. John’s wart induces

CYP3A4, reducing concentrations of cyclosporine A), and

several herbal drugs affect metabolism of warfarin.

Cardiac Toxicity

Many drugs used in the ICU have potentially adverse cardiac

effects. These include drugs that cause tachycardia (eg, beta-

adrenergic agonists, dopamine, and theophylline), bradycardia

(eg, beta-adrenergic blockers and certain calcium channel

blockers), myocardial depression, and arrhythmias (eg,

digoxin, theophylline, and, surprisingly, antiarrhythmic drugs).

An important but unusual adverse effect is drug-induced

prolonged QT interval syndrome, sometimes associated with

a chaotic ventricular tachycardia (torsade de points). Both

cardiac and noncardiac drugs have been associated with this

syndrome, including quinindine and procaineamide;

antipsychotic drugs; antibiotics such as macrolides, fluoro-

quinolones, and ketoconazole; histamine-1-antihistamines;

and other drugs. In some cases, patients receiving multiple

drugs develop interactions that increase serum levels of the

contributing agent. It should be noted that some such drugs

have been withdrawn from the market because of the risk of

prolonging the QT interval. Drug-induced prolonged QT

syndrome and torsade de points are more common in

women, those with heart disease and electrolyte disorders,

and those with familial long QT syndrome. In one study, one

or more of these risk factors were present in the majority of

patients who had drug-induced torsade de points.

Electrolyte Abnormalities

Drugs may be associated with a variety of electrolyte and

acid-base abnormalities. Some of the effects are predictable,

such as hypokalemia induced by thiazide diuretics,

furosemide, glucocorticoids, insulin, and beta-adrenergic

agonists and hyperkalemia from spironolactone, tri-

amterene, or angiotensin-converting enzyme (ACE)

inhibitors. On the other hand, less often expected are hyper-

kalemia with heparin, potassium penicillin G, and trimetho-

prim. Hyponatremia may be a feature of thiazide diuretic

administration. Amphotericin B is associated with

hypokalemia, hypomagnesemia, and renal tubular acidosis.

MEDICATION ERRORS & PREVENTION

IN THE ICU

Reduction of medical errors is the focus of many hospital

quality improvement plans, and medication errors usually

make up the vast majority of medical errors. In the ICU,

attention to medication administration is very important,

and systems to reduce errors have been shown to be effective.

Medication errors arise at any point—from ordering to

administration. The most common errors are those of dosing

without due consideration for the patient’s age, size, renal or

hepatic function, or drug interactions. More rare is adminis-

tration of the wrong medication; this can result from tran-

scription errors or failure to match the medication with the

patient. Because critically ill patients often have multiple clini-

cians participating in their care and multiple medications

Table 4–8. Hepatotoxic drugs.

Type of Hepatic Injury Drug

Hepatocellular Acetaminophen

Allopurinol

Amiodarone

Isoniazid

Ketoconazole

Lisinopril

Losartan

Nevirapine

Paroxetine

Pyrazinamide

Rifampin

Risperidone

Ritonavir

Salicylates

Sertraline

HMG-CoA reductase inhibitors (“statins”)

Tetracycline

Trazodone

Valproic acid

Cholestatic Amoxicillin ⁄ clavulanate

Anabolic steroids

Azathioprine

Chlorpromazine

Clopidogrel

Cytarabine

Erythromycin

Estrogens

Oxacillin

Phenothiazines

Sulindac

Tricyclic antidepressants

Mixed Amitryptilline

Azathioprine

Captopril

Carbamazepine

Clindamycin

Cyprohepatadine

Enalapril

Ibuprofen

Nitrofurantoin

Phenobarbital

Phenytoin

Sulfonamides

Trazodone

Trimethoprim-sulfamethoxazole

Verapamil



CHAPTER 4

prescribed, it is not surprising that there may be duplication of

medication orders, inadvertent administration of two drugs

for the same purpose, and unrecognized drug interactions.

Systems for minimizing medication errors in the ICU

depend on careful, timely, and regular review of medications

and reconciliation of orders with administered medications.

Standardized intravenous mixtures, protocols for drug

administration (eg, sedation guidelines and IV insulin and

heparin protocols), computerized order entry, and auto-

mated review of potential drug interactions are effective

tools.

REFERENCES

Boucher BA, Wood GC, Swanson JM: Pharmacokinetic changes in

critical illness. Crit Care Clin 2006;22:255–71, vi. [PMID:

16677999]

Chang CY, Schiano TD: Review article: Drug hepatotoxicity. Aliment

Pharmacol Ther 2007;25:1135–51. [PMID: 17451560]

Erstad BL: Dosing of medications in morbidly obese patients in the

intensive care unit setting. Intensive Care Med 2004;30:18–32.

[PMID: 14625670]

Gabardi S, Abramson S: Drug dosing in chronic kidney disease. Med

Clin North Am 2005;89:649–87. [PMID: 15755472]

Kane-Gill S, Weber RJ: Principles and practices of medication safety

in the ICU. Crit Care Clin 2006;22:273–90, vi. [PMID: 16678000]

Lewis JH et al: Drug-induced liver disease. Curr Opin Gastroenterol

2006;22:223–33. [PMID: 16550036]

Mann HJ: Drug-associated disease: Cytochrome P450 interactions.

Crit Care Clin 2006;22:329–45, vii. [PMID: 16678003]

Schetz M et al: Drug-induced acute kidney injury. Curr Opin Crit

Care 2005;11:555–65. [PMID: 16292059]

Stedman C: Herbal hepatotoxicity. Semin Liver Dis 2002;22:

195–206. [PMID: 12016550]

Taber SS, Mueller BA: Drug-associated renal dysfunction. Crit Care

Clin 2006; 22:357–74, viii. [PMID: 16678005]

Trotman RL et al: Antibiotic dosing in critically ill adult patients

receiving continuous renal replacement therapy. Clin Infect Dis

2005;41:1159–66. [PMID: 16163635]

Vandendries ER, Drews RE: Drug-associated disease: Hematologic

dysfunction. Crit Care Clin 2006;22:347–55, viii. [PMID:

16678004]

Zeltser D et al: Torsade de pointes due to noncardiac drugs: Most

patients have easily identifiable risk factors. Medicine

(Baltimore). 2003;82:282–90. [PMID: 12861106]