Drug Dosing Renal Failure

ANALYTIC REVIEWS

Drug Dosing in Critically Ill Patients with Renal Failure: A Pharmacokinetic Approach
Ronald J. DeBellis, PharmD,*† Brian S. Smith, PharmD,‡ Pauline A. Cawley, PharmD,§ and Gail M. Burniske, PharmD¶
DeBellis RJ, Smith BS, Cawley PA, Burniske GM. Drug dosing in critically ill patients with renal failure: a pharmacokinetic approach. J Intensive Care Med 2000;15:273–313.

Accurate pharmacotherapy management in the intensive care unit (ICU) patient is crucial to minimize adverse drug events. Pharmacokinetic principles including absorption, distribution, metabolism, and excretion (ADME) all play an important role in determining the fate of medications used in the critical care setting. Renal failure in this setting further alters pharmacokinetic parameters, resulting in drug dosing changes. This article highlights and applies principles of drug dosing in normal patients and in the pharmacokinetically challenging environment of critically ill patients with renal failure. Specific drug dosing tables serve as a guide for the clinician to renally adjust medication doses in the critically ill patient with renal failure.

From the *Massachusetts College of Pharmacy and Health Sciences, †University of Massachusetts School of Medicine, ‡University of Massachusetts Memorial Health Care, Worcester, MA, §Regional Medical Center at Memphis, Memphis, TN, and ¶University of Maryland Medical Center, Baltimore, MD. Received Jun 2, 2000, and in revised form Jul 17, 2000. Accepted for publication Jul 18, 2000. Address correspondence to Ronald J. DeBellis, Department of Pharmacy, University Campus, 55 Lake Ave. N, Worcester, MA 01655, or e-mail: [email protected]

Critically ill patients often have multiple medical problems and commonly receive complex medication regimens. These factors make critically ill patients highly susceptible to adverse drug events. Adverse drug events have been shown occur in 10–12% of intensive care unit (ICU) patients and occur approximately two times more frequently when compared to patients on general medicine units [1,2]. Patients who are critically ill are also at increased risk for developing renal failure. Acute renal failure occurs in 7–25% of all patients admitted to ICUs and increases the average mortality from approximately 15% to more than 60% [3–9]. Renal failure is a risk factor for adverse drug events and likely contributes to the high rate of adverse drug events in critically ill patients. Up to 45% of patients with an estimated creatinine clearance less than 40 ml/min receive medications that are dosed 2.5 times higher than the maximum recommended dose [10]. In addition, adverse drug reactions have been shown to occur in 9% of patients with blood urea nitrogen (BUN) level less than 20 mg/dl versus 24% of patients with a BUN greater than 40 mg/dl [11]. Adverse drug events place critically ill patients at risk for morbidity and mortality, resulting in the utilization of tremendous financial resources. It has been estimated that each adverse drug event increases hospital costs by $2,000–$4,600 [12–14]. The cost of adverse drug events in patients admitted to an ICU have the potential to be higher due to the augmented cost of an ICU bed. Accurate drug dosing in a pharmacokinetically challenging environment is essential to ensure optimal pharmacotherapy in critically ill patients. This particularly applies to patients with renal failure. The following review will discuss some key concepts and theories of drug dosing in critically ill patients. In addition, practical guidelines for drug dosing in the critically ill patient with renal failure are delineated.

Pharmacokinetics and Pharmacodynamics
In order to comprehend the disposition of medications in patients with renal failure, the clinician must
Copyright ᭧ 2000 Blackwell Science, Inc.

273

274 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

apply general pharmacokinetic and pharmacodynamic principles in order to construct an effective and safe medication regimen. Pharmacokinetics describes the processes influencing a drug’s transport through the body to its site of action. The processes affecting a drug’s transport through the body are absorption, distribution, metabolism, and elimination (ADME). Clinical pharmacokinetics is the application of these principles to individualize drug therapy based on patient-specific information with the goal of maximizing therapeutic outcomes while minimizing the risk of toxicity. Pharmacodynamics describes the relationship between a drug’s concentration at the site of action and the ensuing pharmacologic response. The relationship between pharmacokinetics and pharmacodynamics can be described as the dose of a drug that provides a sufficient drug concentration at the site of action in order to elicit a biologic response (Fig 1). There are many factors in critically ill patients with renal failure that will alter the pharmacokinetics and pharmacodynamics of drugs and necessitate modification of the drug therapy regimen. Before we discuss the changes that occur in critically ill patients with renal failure, we will briefly review pharmacokinetic models and pharmacokinetic terminology.

Pharmacokinetic Models
Many mathematical models and equations have been developed to describe pharmacokinetic processes. The aim is to focus on concepts and ideas

that will provide the practitioner with an understanding of pharmacokinetics as it relates to drug dosing in the critically ill patient. Compartmental models are methods to mathematically describe and conceptualize pharmacokinetic processes. A drug can be described in terms of a one-, two-, or multicompartment model. Onecompartment models do not precisely represent the pharmacokinetics of most drugs, but the equations used to describe one-compartment models are clinically most useful. Two- and multicompartment models more accurately describe the pharmacokinetics of most drugs, but the equations needed to describe these models are very complex and are not clinically practical. We will review the concepts of a one-compartment model because the concepts are clinically useful and they can be applied to twoand multicompartment models. In a one-compartment model, the drug enters the central or vascular compartment either by direct intravenous injection or by other routes requiring absorption. Drugs administered intravenously will have 100% bioavailability or 100% of the administered dose reaches the systemic circulation. Bioavailability refers to both the rate and extent of absorption and is defined as the relationship between the total amount of drug and how fast that drug is absorbed from a nonintravenous route compared to the same dose administered intravenously. Drug absorption can occur through many different membranes including the gastrointestinal tract, skin, subcutaneous or intramuscular tissue, lungs, or any other membrane. The bioavailability of drugs

Fig 1. The relationship between pharmacokinetics and pharmacodynamics. (Adapted from Chernow B, ed. Critical care pharmacotherapy. Baltimore: Williams & Wilkins, 1995:4.)

DeBellis et al.: Drug Dosing in Renal Failure 275

administered by any route requiring absorption can be highly variable. Once a drug is in the central compartment, it can bind to plasma proteins, primarily albumin or ␣1-glycoprotein, and may be pH dependent. The drug will then achieve a state of equilibrium between the unbound and bound state. Only unbound or free drug is available to exert a pharmacologic effect, and to be metabolized and/ or eliminated. Unbound drug is primarily metabolized and eliminated from the central compartment via the liver and kidneys, the major metabolic and elimination pathways (Fig 2). In two- and multicompartment models, all the same processes of a one-compartment model occur, but the drug will also distribute to peripheral compartments such as adipose tissue, muscle tissue, or the central nervous system. When the amount of unbound drug going into and coming from various compartments is equal, a state of equilibrium is achieved. The phase where drug is achieving a state of equilibrium between compartments is called the distribution phase. Similar to one-compartment models, unbound drug in the central compartment is often metabolized and eliminated through the liver and kidney, though some drugs can be metabolized and/or eliminated in peripheral compartments as well.

Zero- and First-Order Kinetics
Most drugs used in critically ill patients will follow principles associated with first-order or linear pharmacokinetics, however, some drugs demonstrate zero-order or nonlinear kinetics. For example, phenytoin follows zero-order pharmacokinetic principles or Michaelis–Menten kinetics. Drugs are described as exhibiting zero-order pharmacokinetics when a constant quantity or amount of drug is removed per unit of time. As the plasma concentration of the drug decreases or increases, the quantity or amount eliminated remains the same. Zero-order kinetics is a result of metabolism by a saturated enzyme system eliminating drug at a constant rate despite the serum concentration of the drug. Clinically this means small increases in the drug’s dose can lead to large increases in the plasma concentration, hence the term nonlinear pharmacokinetics (Fig 3). Because there are few drugs that follow zero-order pharmacokinetics, attention will be directed toward drugs following principles of firstorder or linear pharmacokinetics. Medications that demonstrate first-order or linear pharmacokinetics eliminate or dispose of a constant percentage of drug from the plasma per unit of

Fig 2. Compartmental drug model. a Drugs administered intravenously enter the central compartment directly. b Drugs administered by any route other than intravenously must be absorbed before entering the central compartment. c Distribution to peripheral compartments only occurs in two or multicompartment models. d In a one-compartment model, drug interacts with its receptor directly from the central compartment.

276 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

Fig 3. Effect of increasing daily dose on average steadystate drug concentrations for drugs undergoing nonlinear or zero order pharmacokinetic modeling ( ). Effect of increasing daily dose on average steady-state drug concentrations for drugs undergoing linear or first order pharmacokinetic modeling ( _ _ _ _ _ ).

time required for a drug to reach equilibrium between the central and peripheral compartments. Drug such as gentamicin demonstrate an alpha elimination phase. It is this phase that is important in causing acute tubular necrosis. This is important to keep in mind because most drugs take time to distribute. When drawing blood levels, it is important to know the time required for the distribution phase to complete. Obtaining plasma drug concentrations prior to completion of distribution will yield falsely elevated levels. This will ensure clinical decisions are not made on falsely elevated drug levels that have not fully distributed to other tissues. Once distribution is complete, the slope changes and is referred to as the beta or elimination phase. The slope of the elimination phase is the elimination rate constant (Kel) and can be used to determine the drugs half-life (t1/2).

Half-Life
time. As the plasma concentration increases, the amount of drug eliminated increases (a directly proportional relationship). As the plasma concentration decreases, the amount of drug eliminated decreases. In a clinical sense, if the dose of a drug is increased, the plasma concentration increases proportionally as well as the amount eliminated (Fig 3). Regardless of the plasma concentration, the percentage that is eliminated remains constant. If the logarithm of the plasma concentration versus time for a drug is plotted, one will see two different slopes (Fig 4). The upper portion is referred to as the alpha or distribution phase and represents the A drug’s half-life (t1/2) is a constant value determined by the function of the metabolizing and eliminating processes. The definition of a drug’s half-life is the amount of time required for the concentration of the drug to decrease by 50%. Half-life is often expressed in minutes or hours. The half-life of a specific drug will remain the same as long as the function of the metabolizing and eliminating processes remains constant. For example, if it takes 8 hours for the plasma concentration of a drug to decline from 10 mg/L to 5 mg/L, the plasma halflife is 8 hours. If, however, there is a change in renal function, the half-life can be significantly prolonged. The half-life of a drug can be used to determine the time required for a drug to be eliminated from the body, as well as the time required to reach steady state. From a pharmacokinetic perspective, it takes three to five half-lives to achieve 87.5–96.875% of steady-state drug concentration (the point where the amount of drug going into the body equals the amount being eliminated). Concurrently it takes the same period of time for a drug to be 87.5–96.875% eliminated from the body. This is important since a clinician should generally wait for steady state prior to drawing a blood level or increasing the dose of a medication. Knowing how long it will take before a drug is almost completely removed can help a clinician judge how long it should take for a pharmacologic or toxic effect to wear off [15]. It is important for a clinician to keep in mind that the pharmacodynamic behavior of some drugs may correlate with pharmacologic activity regardless of pharmacokinetic drug behavior. Some

Fig 4. Logarithm of plasma concentration (Cp) versus time plot for a drug following rapid intravenous injection delineating both the alpha distribution and beta elimination phase.

DeBellis et al.: Drug Dosing in Renal Failure 277

drugs can stay at receptor sites and have pharmacologic activity long after the plasma concentration has decreased. For example, consider the extended spectrum macrolide antibiotic azithromycin. A patient generally requires 5 days of therapy for most infections, however, administration of this drug for 5 days is the same as receiving 9–10 days of therapy from other antimicrobial agents. Understanding a drug’s pharmacokinetic and pharmacodynamic profile is necessary to appropriately use it’s halflife to dose and assess the response in critically ill patients with renal failure.

Elimination Rate Constant
As mentioned earlier, most drugs follow first-order models of elimination. The elimination rate constant (Kel) determines the rate of elimination from the body. With first-order elimination, a constant percentage of drug is removed from the plasma per unit of time and is often expressed as per minute or per hour. The elimination rate constant for a drug can represent total body elimination (Kel) or it can be broken down into the specific organs responsible for elimination, such as renal (Kr) or metabolic (Km) [15]: Kel ‫ ס‬Kr ‫ ם‬Km. (1)

The elimination rate constant for a drug can be determined by plotting the logarithm of the drug plasma concentration (Cp) over time (t) and determining the slope of the line after distribution has occurred (Fig 4). The following equation describes linear or first-order elimination [15]: Cp(0) ‫ ס‬Cp(t)e–(Kel)(t). (2)

The volume of distribution is most commonly expressed in terms of liters (L) or liters per kilogram (L/kg). It is important to remember that the volume of distribution is a theoretical volume, not a physiological volume. For example, if a 700 mg dose of a drug administered intravenously to a 70 kg patient results in a calculated maximum plasma concentration of 7 mg/L, it appears as if the drug is dissolved in 100 L of fluid. The volume of distribution would be 100 L or 1.429 L/kg. Obviously under normal physiologic conditions, a 70 kg adult does not have 100 L of fluid in their body. A drug can have such a high volume of distribution as a result of plasma protein binding and/or distribution to other compartments (intracellular space, lipid compartments, muscle). Protein binding or distribution to peripheral compartments leads to a larger volume of distribution by reducing the amount of measurable drug in the plasma. Drugs that are not highly bound to proteins and/or drugs that do not distribute out of the central (vascular) compartment will tend to have lower volumes of distribution closer to the intravascular volume. In clinical situations it is difficult to calculate a drug’s volume of distribution. The equation above assumes an intravenous bolus of a drug, instantaneous distribution, and the maximum plasma concentration that immediately results. This cannot be accomplished in clinical situations. The maximum plasma concentration (Cpmax) has to be calculated or back-extrapolated from a measured peak plasma concentration (Cppeak). The following equation is commonly used to calculate Cpmax [15]: Cpmax ‫ ס‬measured Cp/e –(Kel)(t). (5)

Cp(0) is the plasma concentration of the drug at time zero, Cp(t) is the plasma concentration of the drug at time (t ), and (Kel) is the elimination rate constant. The elimination rate constant is also inversely proportional the drug’s half-life (t1/2) [15]: t1/2 ‫ ס‬0.693/Kel. (3)

The time (t) is the time difference between when the dose was administered and when the peak plasma concentration (measured Cp) was drawn.

Clearance
Clearance (Cl ) is the term describing the volume of fluid cleared of drug over time, usually in milliliters per minute. Total body clearance (ClTB) represents all the processes involved in removing a drug from the body. Clearance can be broken down into the individual organs or processes that are responsible for the elimination of drug from the body, such as renal clearance (ClR), metabolic clearance (ClM), or any other process that eliminates drug (ClX) from the body. Total body clearance is the sum of all the clearance processes in the body [15]: ClTB ‫ ס‬ClR ‫ ם‬ClM ‫ ם‬ClX. (6)

The rate of elimination and half-life are constants and do not change unless the function of the metabolizing and/or eliminating processes change.

Volume of Distribution
The volume of distribution (Vd) is a parameter relating the dose of a drug to the maximum plasma concentration (Cpmax) [15]: Cpmax ‫ ס‬dose/Vd or Vd ‫ ס‬dose/Cpmax. (4)

278 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

Clearance through an organ is determined by the blood flow to the organ (Q) and the extraction ratio (ER) for the organ. Blood flow to the organ is expressed as a unit of volume per time, often milliliters per minute. The extraction ratio is the percentage or fraction of drug removed from the blood as a result of passing through the organ and has no units. The extraction ratio depends on the fraction of free drug presenting to the organ, the rate of blood flow through the organ, and the intrinsic ability of the organ to eliminate drug [15]: Cl ‫( ס‬Q)(ER). (7)

Glomerular Filtration
Plasma flow to the kidney is approximately 650–700 ml/min in a healthy adult. Of this amount, about 20% is filtered at the glomerulus. Glomerular filtration is the most common means of drug excretion by the kidneys [16–18]. Drug excretion via glomerular filtration is a passive, first-order process. Drug excretion is a function of the glomerular filtration rate (GFR) and the percentage of free unbound drug. The unbound drug is filtered through pores in the glomerular capillaries called fenestrae. The pores in the glomerular capillaries are much larger than pores found in other capillaries, making them much more permeable to solutes. The GFR for an average healthy adult is 100–125 ml/min [16–18]. The GFR depends on the hydrostatic pressure or renal plasma flow and osmotic pressure gradients between the glomerulus and Bowman’s capsule. There are many factors influencing the amount of drug filtered at the glomerulus. Table 1 lists these factors and how they influence drug filtration.

Changes in blood flow to the organ responsible for clearing the drug or any factor altering the extraction ratio of a drug will alter a drug’s clearance. For example, a patient experiencing septic or cardiogenic shock may have impaired blood flow to the liver or kidneys impairing the clearance of a particular drug. In addition, if a pharmacologic vasopressor is added to the therapy, blood flow to the gastrointestinal tract may be compromised, resulting in a decreased transport of drug to target site. Clearance is also equal to the product of the elimination rate constant (Kel ) and the volume of distribution (Vd ) [15]: Cl ‫( ס‬Kel)(Vd). (8)

Effects of Impaired Glomerular Filtration on Drug Elimination
Impairment of glomerular filtration can lead to a clinically significant accumulation of drug and/or its metabolites. To assess the likely impact of decreased glomerular filtration, it is important to know what fraction of a drug is renally eliminated, as well as the excretion method for any active or toxic metabolites [16,17]. Drugs excreted primarily by glomerular filtration will be filtered at a rate that is proportional to the patients GFR (first order) and the percentage of free drug in the plasma, since only free, unbound drug can be filtered:

Using pharmacokinetics to calculate Kel and Vd, it is possible to calculate a drugs clearance from the body. Again, it is easy to see that changes in the elimination rate constant and/or volume of distribution will affect a drug’s clearance from the body.

Renal Drug Excretion
The kidney is the primary organ responsible for the excretion of drugs and their metabolites. The three main processes by which the kidney excretes drugs include glomerular filtration, tubular secretion, and tubular reabsorption.

Rate of ‫( ס‬GFR) (free drug in plasma) (9) drug filtration

Table 1. Factors Influencing Glomerular Filtration of Drugs [16–18] Factor Hydrostatic pressure Plasma protein binding Volume of distribution Molecular size Glomerular integrity Number of functioning nephrons Effect on Glomerular Filtration Drug filtration decreases as hydrostatic pressure decreases Drug filtration decreases as plasma protein binding increases High volume of distribution decreases the amount of drug available to be filtered Drug filtration decreases as molecular size increases (MW less than 5 kDa and ˚) radii less than 15 A Drug filtration increases as membrane integrity decreases Drug filtration decreases as the number of functioning nephrons decreases

DeBellis et al.: Drug Dosing in Renal Failure 279

For example, if a drug is excreted solely by filtration and the GFR decreases by 50%, drug excretion will also decrease by 50%. There are many mechanisms by which GFR is altered. Table 2 summarizes the major factors and mechanisms.

Tubular Secretion
As mentioned before, 20% of the plasma flow is filtered at the level of the glomerulus. The remaining 425–600 ml/min of renal plasma flow not filtered at the glomerulus is directed to the peritubular capillaries, where drugs may be secreted. Tubular secretion is an active process where drugs are transported by membrane proteins from the interstitial fluid surrounding the proximal tubule and secreted into the lumen. Tubular secretion rate depends on the intrinsic activity of the transporter, proximal tubule blood flow, and the percentage of

free or unbound drug. There are two main transport systems for drugs in the proximal tubule. One transport system is for anions and the other transport system is for cations [16,22–27]. These secretory transport systems are not fully understood and there may be more than one transport subtype for each system responsible for eliminating different substances. Drugs can compete for secretion with other drugs and endogenous substances secreted by the same transporter, since they are saturable [24–27]. An example of competition for secretion via an anionic transporter is probenecid with penicillins or cephalosporins [28]. This combination has been used to prolong the half-life of penicillin. Table 3 is a list of drugs actively secreted by the kidney. It is difficult to study drug secretion interactions because most drugs are metabolized and eliminated by multiple processes. This makes it difficult to study the effects of secretion alone with all these other processes occurring simultaneously. Studies

Table 2. Factors Affecting Glomerular Filtration in Critically Ill Patients with Renal Failure [18–21] Factor Cardiac output Mechanism Decreased cardiac output leads to decreased GFR. Homeostatic mechanisms attempt to maintain blood flow to the heart, brain, and muscle at the expense of the kidney. Decreased plasma flow to the kidney resulting in proportional decrease in GFR. Increased permeability results in increased clearance rate of protein bound-drugs. For example, with membrane permeability in nephrotic syndrome the glomerular basement membrane loses negative charge, allowing albumin and other large molecules to cross the barrier. Tubular blockage with casts, cellular debris, or cellular swelling leading to decreased GFR. Decreased GFR with luminal fluid back-leak into the interstitium and renal venous blood; caused by damaged epithelium in moderate to severe acute renal failure. Decreased GFR due to either afferent arteriolar vasoconstriction or efferent arteriolar capillary hydraulic pressure vasodilation. Decreased GFR by drug-mediated prostaglandin inhibition.

Renal plasma flow Glomerular basement

Renal tubule obstruction Reabsorption via back-leak Decreased glomerular filtration Drug induced

Table 3. Drugs Secreted by Anionic and Cationic Transport Systems [22–27] Organic Anion Transport Acyclovir Acetazolamide Captopril Cephalosporins Folic acid Furosemide Moxalactam Nafcillin Penicillin G Phenobarbital Sulfonamides Thiazides Organic Cation Transport Acetylcholine Amantadine ␤-blockers Cimetidine Ephedrine Epinephrine Methadone Morphine Procainamide Pseudoephedrine Serotonin Ampicillin Cisplatin Ibuprofen Naproxen Probenecid Uric acid Amiloride Creatinine Ethambutol Nicotine Quinidine Ascorbic acid Clofibrate Indomethacin Nitrofurantion Quinolones Zidovudine Amphetamines Digoxin Famotidine Norepinephrine Quinine Benzylpenicillin Ethacrynic acid Methotrexate Oxalate Salicylates

Atropine Disopyramide Metformin Pindolol Ranitidine

280 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

evaluating changes in drug elimination in renal failure look at changes in total body clearance and/or increases in blood concentration rather than excretion changes via secretion alone [24–26].

Pharmacokinetic Changes in Critically Ill Patients with Renal Failure
There are many physiological changes that can occur in a critically ill patient that will alter the pharmacokinetics of drugs. Overall, studies looking at specific pharmacokinetic changes in drugs in critically ill patients are limited. Most pharmacokinetic studies are performed in healthy volunteers or in patients with a specific disease state who are not critically ill. Critically ill patients are very dynamic and often have multiple organ dysfunction or age-specific parameters potentially altering all aspects of drug therapy. Some ADME changes that can occur in critically ill patients, specifically those with renal failure, will be addressed.

Effects of Impaired Tubular Secretion on Drug Elimination
Tubular secretion is an extremely efficient process that is capable of eliminating relatively large amounts of drug substances from the blood in a short period of time [27]. When drugs are excreted from the body via tubular secretion, the rate of drug clearance is greater than drug clearance via filtration [26]. A decline in renal function impacts tubular secretion, as both endogenous and exogenous organic acids and bases accumulate and compete for transporters. This competition for the transporters required for active secretion may lead to either drug toxicity or lack of efficacy, however, it is not possible to predict which is most likely to occur or the extent of the interaction [19].

Absorption
There is a lack of abundant discussion concerning drug absorption in patients with renal failure [30]. Drug absorption in patients with renal failure may be altered secondary to gastrointestinal edema, nausea and vomiting due to uremia, and delayed gastric emptying [31,32]. Comorbid illnesses or conditions that commonly occur with renal disease such as diabetic gastroparesis may also have a significant effect on drug absorption. Patients receiving peritoneal dialysis may experience complications leading to peritonitis, which has been shown to decrease gastrointestinal peristalsis, thus impairing the absorption process [19]. In general, the average number of medications taken by a patient in renal failure is eight [33,34]. This setting provides the framework for multiple drug interactions. Specific drug interactions involving decreased absorption secondary to chelation manifest in patients taking phosphate binding antacids containing aluminum or calcium. Other enterally administered medications need to be spaced around the antacid by at least 2 hours to minimize chelation. Although these antacids are administered for the sole purpose of phosphate binding in renal failure patients, a subsequent increase in gastric pH

Tubular Reabsorption
Tubular reabsorption of drugs can occur by active and/or passive processes. When ultrafiltrate passes through the nephron, up to 99% of the filtered volume is reabsorbed. This can lead to a dramatic increase in a drug’s concentration in the tubule as the volume decreases. This high concentration gradient of drug between the renal tubule and plasma promotes passive diffusion from inside the tubule into the plasma. The properties that effect passive tubular reabsorption are listed in Table 4. Altering urine pH has long been used to decrease the amount of drug reabsorbed and enhance excretion. Alkalinizing the urine can be used to enhance the elimination of barbiturates (weak acids) by increasing the fraction of ionized drug, which decreases the amount available for reabsorption [29]. Table 5 lists some drugs with pH-dependent elimination.

Table 4. Factors Influencing Tubular Reabsorption [16,22,26] Factor Lipid solubility of the drug Degree of ionization of the drug Urine pH Urine flow Concentration gradient Effect on Tubular Reabsorption Increased reabsorption with increased lipid solubility Decreased reabsorption with increased ionization Variable depending on if drug is acidic or basic Decreased reabsorption as urine flow increases Increased reabsorption as concentration gradient increases

DeBellis et al.: Drug Dosing in Renal Failure 281

Table 5. Drugs with pH-Dependent Elimination [16] Weak Acids Phenobarbital Salicylates Sulfonamides Weak Bases Amphetamines Ephedrine N-acetylprocaninamide Procanamide Pseudoephedrine Quinidine Tocanide Tricyclic antidepressants

occurs [19]. The elevated gastric pH may impair the dissolution process of other enterally administered medications, leading to incomplete drug absorption, particularly with acidic drugs. In addition, the onset of drug action may be delayed secondary to decreased gastric emptying. The effects of enterally administered analgesics are often impaired in this situation. Caution must be exercised in renal failure patients with concomitant diabetic gastroparesis. Metoclopramide or erythromycin are frequently administered to enhance the motility of the gastrointestinal tract. When these agents are administered, enteral absorption of medications is often decreased due to an increase in gastrointestinal transit [35]. Bioavailability studies, for the most part, are lacking in critically ill patients. Most bioavailability studies are conducted in healthy adults versus critically ill patients, however, it is known that in a majority of medications, the bioavailability in patients with renal failure is either unchanged or increased (Table 6).

Distribution
Altered plasma protein binding in critically ill patients with renal failure can significantly change drug distribution. Drugs that bind to plasma proteins exist in a state of equilibrium between unbound (free) and bound drug (not free), and since the unbound drug exerts a pharmacologic effect, decreased binding increases the amount of drug available to exert a pharmacologic effect and therefore increases the risk of toxicity. Drug-drug interactions can occur when two highly plasma protein-bound drugs compete for binding with the same plasma protein. A drug is considered to be highly plasma protein bound when more than 90% is bound to plasma proteins. Drugs that are bound to plasma proteins less than 90% are not considered to be clinically significant binders. Anionic or acidic drugs tend to bind to albumin, while cationic or

basic drugs tend to bind to ␣1-glycoprotein. Drugs like warfarin, phenytoin, valproic acid, and salicylates are highly bound to albumin and can lead to displacement-mediated drug interactions if administered together [43–45]. Even though drug displacement interactions occur, their clinical significance tends to be low. Drug-drug interactions do not occur primarily due to alteration in plasma protein binding, but they also occur in patients with poor renal function due to changes in the configuration of albumin [46–48]. For example, the pharmacodynamic effects of phenytoin and warfarin are increased in patients with renal failure. The decreased binding of drugs to albumin in patients with renal failure is thought to be due to the accumulation of small acidic molecules displacing these drugs from binding sites or alterations in binding sites on the albumin molecules [49,50]. Critically ill patients often have low albumin due to malnutrition and/ or acute illness [52]. This can lead to higher free fractions of drugs and potentially increase the risk of toxicity. Drugs binding to ␣1-glycoprotein appear to be less affected in critically ill patients with renal failure, even though it is an acute-phase reactant that increases with trauma, surgery, or acute illness [49]. Any of these protein binding changes may alter a drug’s volume of distribution. For example, if plasma protein concentrations experience a sudden decrease, and a patient is taking warfarin, the volume of distribution for that medication becomes significantly smaller since warfarin is bound to plasma proteins more than 90%. This occurrence becomes significant by having warfarin exert more of a pharmacodynamic effect by having less drug bound to proteins that may result in an adverse event such as bleeding. In addition, fluid status can be highly variable in a critically ill patient with renal failure, leading to changes in a drug’s volume of distribution. Accumulation of fluid in renal failure patients can result in lower drug concentrations [52]. The clearance of drugs can be affected by changes in the volume

Table 6. Bioavailability of Drugs in Patients with Renal Disease [19,29–32,36–42] Decreased D-Xylose Furosemide Pindolol Unchanged Cimetidine Ciprofloxacin Codeine Digoxin Labetalol Trimethoprim Sulfamethoxazole Increased Bufuralol Dextropropoxyphene Dihydrocodeine Oxprenolol Propranolol Tolamolol

Adopted from [23].

282 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

of distribution and protein binding by altering the amount of unbound drug available to be metabolized and/or eliminated. Despite the potential for many changes in the distribution of drugs in critically ill patients with renal failure, it is often difficult or impossible to predict these interactions. It is important for the clinician to be aware of the potential for interaction and monitor for the signs of efficacy and toxicity so interactions are recognized and corrected.

Metabolism
Most drug dosage changes in patients with renal failure are necessary due to reduced renal elimination, however, some drugs can have altered metabolic elimination due to renal failure. The kidneys have been found to have many drug metabolizing systems, and it is likely that renal disease alters renal drug metabolism as well as hepatic metabolism [53–56]. The exact mechanisms are not completely defined, but one study suggested drugs oxidized by the cytochrome P-450 2D6 isozyme are more likely to be affected [57]. The clinical significance of these effects in critically ill patients with renal disease remains to be further explored. Critically ill patients often have impaired metabolic function from nonrenal causes; either from direct damage to the liver (cirrhosis), decreased blood flow to the liver (shock, elderly), or as a result of other medications that are enzyme inhibitors or inducers [58–60]. Clearance ‫ ס‬flow ‫ ן‬extraction ratio (Equation 7) mathematically delineates this concept. Careful drug dosing and monitoring is essential to ensure drug therapy is achieving the desired pharmacologic effects without causing adverse events.

Elimination
Studies to determine drug pharmacology and clearance in critically ill patients are usually performed on patients undergoing anesthesia. Occasionally the study population includes patients with chronic disease to one organ that is stable. It is therefore difficult to apply these study results to critically ill patients with unstable, multiple organ disease. Critically ill patients each have a unique combination of factors that can affect renal drug clearance [61]. In addition to renal failure, which is relatively common in this patient population, the impact of other organ dysfunction such as liver, cardiovascular, or respiratory failure, as well as malnutrition, must be assessed [62,63].

Acute renal failure is often accompanied by metabolic acidosis and respiratory alkalosis. Depending on the pKa value of drugs, the pH difference between plasma and tissue compartments may alter the ionization of drug molecules and therefore affect tissue redistribution versus clearance [64]. Renal failure also impacts total body water and therefore drug volume of distribution. Patients who have a low level of serum albumin via decreased hepatic synthesis, protein loss through increased vascular permeability, or malnutrition will have a corresponding decrease in plasma protein binding of drugs [65]. This can lead to an increase in clearance of drugs that are normally highly plasma protein bound. Plasma protein binding can also be reduced in conditions such as the nephrotic syndrome, proteinuria, conditions that alter the molecular structure of albumin, and with accumulation of uremic toxins that compete with drugs for protein binding sites [64,66]. Cardiovascular failure contributes to the reduction in renal drug clearance by two mechanisms: reduction in cardiac output and therefore reduction in renal plasma flow and increased hepatic congestion, and increased sympathetic drive leading to shunting of blood away from the kidney in order to protect blood flow to the heart, brain, and muscle. This reduction in the supply of oxygenated blood may further affect drug clearance if anaerobic metabolism and metabolic acidosis ensue, potentially causing changes in the ionization of drugs. Retention of fluid may then increase the drug volume of distribution, further reducing drug clearance [64,66]. Other conditions with profound vasodilation such as sepsis, systemic inflammatory response syndrome (SIRS), multiple organ dysfunction syndrome (MODS), pancreatitis, and liver failure will also cause a decrease in renal drug elimination due to decreased GFR and renal plasma flow [65]. Patients with respiratory failure who are placed on mechanical ventilation may have reduced cardiac output (due to increased mean intrathoracic pressure) and volume of distribution changes, in addition to possible alkalosis or acidosis which can affect drug disposition if clearance is pH sensitive [65,66].

Principles of Drug Elimination Via Dialysis
Drug elimination via dialysis deserves specific attention since there are many variables unique to dialysis that can affect drug clearance. Dialysis patients, on average, take more than eight different medications [33,34]. It is therefore extremely important to adjust the drug dosing schedules cor-

DeBellis et al.: Drug Dosing in Renal Failure 283

rectly according to the degree, if any, of residual renal function, together with the dialyzability of the drugs. The way in which drug dosing should optimally be adjusted depends on the particular drug characteristics and the type of dialysis to be used.

side of the membrane, a negative pressure on the dialysate side, or a combination of the two [68]. In our drug tables we provide dosing information for hemodialysis (HD) and continuos arteriovenous/ venovenous hemodialysis (CAVHD/CVVHD). Information on drug dosing in peritoneal dialysis is not included since this is not a common dialysis modality in the critically ill patient population.

Types of Dialysis
There are two main types of dialysis—diffusive and convective—with many variations or combinations of these principles (Table 7). Diffusive dialysis involves the system of dialysate and blood separated by a semipermeable membrane, with selective movement of substances down a concentration gradient. In this way drugs that are capable of being dialyzed can be cleared from the blood, and electrolytes can be simultaneously replaced from the dialysate if needed. In convective dialysis, however, solutes are removed from blood via solvent drag that is independent of concentration gradients and is not limited by drug molecule size. Conventional hemodialysis describes a process that is primarily diffusive with minimal convective losses, whereas hemofiltration describes primarily convective solute clearance. The terms ‘‘high efficiency’’ and ‘‘high flux’’ are used to indicate large membrane surface area and pore size, respectively [67]. Hemodiafiltration indicates one-third convection and two-thirds high-flux diffusion [34]. Ultrafiltration refers to removal of fluid volume from the patient [68]. In arteriovenous dialysis the driving force is the mean arterial pressure of the patient, whereas for venovenous dialysis the system relies on the use of a mechanical blood pump. The driving pressure for ultrafiltration is established by one of three ways: a positive pressure gradient on the blood
Table 7. Types of Dialysis Dialysis Types Hemodialysis Conventional hemodialysis (HD) Continuous arteriovenous hemodialysis (CAVHD) Continuous venovenous hemodialysis (CVVHD) Continuous arteriovenous hemofiltration (CAVH) Continuous venovenous hemofiltration (CVVH) Continuous arteriovenous hemodiafiltration (CAVHD) Continuous venovenous hemodiafiltration (CVVHD) Slow continuous ultrafiltration (UF)

Dialysis Drug Clearance
Drug clearance via hemodialysis can be estimated as follows: ClHD ‫( ס‬Clurea)(60/MWdrug) (10)

where ClHD is the drug’s clearance by hemodialysis, Clurea is the clearance of urea by the dialyzer (typically about 150 ml/min for standard dialyzers), and MWdrug is the molecular weight of the drug [33]. If a drug has negligible clearance via dialysis, then a postdialysis replacement dose is not necessary; however, if clearance is more efficient, then the percentage of the drug removed is usually calculated and a replacement dose provided. In addition to taking into account the amount of drug removal, it is also important to consider the degree of residual renal function of the patient, since additional dosage may need to be administered to take this into account.

Components of Dialysis System and Factors That Affect Drug Removal
There are three main components to the dialysis system: blood, dialysate, and membrane. Changes to each of these will affect drug removal. Table 8 summarizes how drug removal is affected by changes to the respective components [33,34,67,68]. Since information pertaining to drug dosing in patients receiving different types of dialysis is not always available, assessing the patient-specific dialysis method together with an analysis of the variables discussed above should provide a basis from which to make clinical decisions on drug dosing regimens.

Hemofiltration

Serum Drug Monitoring with Dialysis
If there is a known relationship between serum drug level and efficacy/toxicity, serum drug monitoring can be a useful tool in drug dosing for dialysis patients. If a replacement dose of drug is provided postdialysis then an appropriate period of time

Hemodiafiltration

Ultrafiltration

284 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

Table 8. Factors Affecting Drug Removal During Dialysis [33,34,67–69] Drug Characteristic Molecular weight Comments Larger MW decreases the likelihood that the drug will pass through the dialyzer membrane, depending on membrane type. This is one of the most predictive characteristics of a drug’s dialyzability. Generally if MW > 1,000 Da, then convection is required rather than diffusion to clear the drug. Increased PPB results in a decreased amount of free drug available for dialysis. Note that with dialysis heparinization, lipoprotein lipase is induced which increases levels of free fatty acids (FFA). These FFA compete with certain drugs for protein binding sites. Drugs with a Vd < 1 L/kg are more likely to be dialyzed than those with a higher Vd, providing the MW and PPB conditions are favorable. A drug with a Vd of 1–2 L/kg will be marginally dialyzable, and > 2 L/kg is unlikely to be dialyzable. Synthetic membranes tend to have higher ultrafiltration coefficients than those produced from biologic materials. Modified biologic materials include cellulose and cuprane derivatives. Synthetic materials include polysulfone, polycarbonate, polymethylmethacrylate, and polyacrylonitile-based materials. Increased surface area leads to increased efficiency of drug clearance. However, as the MW increases drug clearance becomes more reliant on convection techniques rather than diffusion. An increase in dialysate flow rate to > 500 ml/min has only a modest increase in solute clearance due to increased turbulence within the membrane. Longer duration of dialysis increases the likelihood of clearance; however, reequilibration of high Vd drugs must be considered.

Plasma protein binding

Volume of distribution

Membrane Membrane type

Surface area

Dialyzer system Flow rate Duration of dialysis

MW ‫ ס‬molecular weight, PPB ‫ ס‬plasma protein binding, Vd ‫ ס‬volume of distribution.

must be established before measuring the drug level to allow adequate tissue distribution. A peak level is typically drawn 1–2 hours after oral drug administration, and about 30 minutes after parenteral administration. A trough level is drawn at the end of a dosing interval. For example, if a drug is to be administered every 12 hours, a trough level should be drawn after 11.5 hours from the time that the dose was administered, or just prior to the next dosage administration. In most instances, if a maintenance dose has been given, then care must be taken to ensure that three to four doses are given before checking the drug level to ensure a steadystate level is established [33]. However, by definition, steady state is based on the half-life of the drug and the dosing schedule is partially based on the half-life. Given this, steady state can occur well before the third or fourth doses or well after. For example, digoxin can be administered on a once a day dosing interval. It takes 5–7 days to achieve steady state with this drug, drawing a level after the third dose will in fact not truly reflect steady state. It is important to note that the drug’s volume of distribution must also be taken into account when considering serum drug monitoring. Drugs with a high volume of distribution (e.g., digoxin) will be extensively distributed into tissues and therefore are not available in the circulation for clearance via dialysis. Thus intracellular drug con-

centrations may only decrease by 1–2% after dialysis, with intercompartmental reequilibration taking place after dialysis completion. This means that if serum drug levels are monitored, it is important to allow sufficient time for reequilibration before measuring the drug level, generally 4 hours after hemodialysis is complete [69]. High ultrafiltration can increase the level of reequilibration rebound.

Measuring Renal Function
Assessing the degree of renal function in a critically ill patient is a crucial step toward being able to select appropriate medication dosages for drugs that are renally eliminated. Since it is not possible to directly measure the GFR, indirect measurements to estimate GFR utilizing either exogenous or endogenous marker substances can be used. The ‘‘gold standard’’ for estimating GFR is the measurement of clearance of an exogenous substance called inulin. This sugar is considered to provide the most accurate GFR estimation because it is filtered freely through the glomerulus and is not subject to renal metabolism, reabsorption, or secretion. Clinically inulin is not commonly used because the administration technique is cumbersome and impractical. An intravenous bolus followed by a maintenance

DeBellis et al.: Drug Dosing in Renal Failure 285

infusion is provided to maintain a certain plasma concentration, and then serial venous blood and urine collections are taken at specific time intervals [69]. The alternative to using inulin in estimating GFR is to utilize the endogenous substance, creatinine. Creatinine is a product of muscle creatine decomposition. Daily, a constant rate of approximately 1.6–2.0% of the total amount of creatine (or approximately 20 mg/kg/day depending on age, gender, diet, and physical condition) in the body is converted spontaneously to creatinine [70]. This constant production means that urinary excretion rate varies by only a small amount in a healthy person. In addition, creatinine is freely filtered across the glomerulus. However, in contrast to inulin, creatinine undergoes a small degree of renal tubular secretion, which means that the use of creatinine as a GFR estimation marker is considered to be less accurate than inulin. Clinically, however, creatinine is commonly used to estimate GFR because measurement of either serum or urine creatinine are more practical. However, whether or not to measure creatinine in the blood or urine is subject to practitioner debate. Traditionally a 24-hour urine collection is conducted to measure the amount of creatinine collected over this time period and assessed with serum creatinine levels taken at the beginning and end of the 24 hours. The most common problem with this method is that urine collections are often incomplete, and therefore the amount of creatinine collected is inaccurate. One possible method to assess whether or not a urine

collection is complete is to compare the amount of creatinine collected to the usual rate of production of about 20 mg/kg/day to see how these two values compare. Perhaps a more reliable method of measuring urine creatinine is to collect consecutive carefully timed urine samples over time periods of a couple of hours instead of 24 hours. However, a shortened collection period may serve to increase the inaccuracy in measuring urine volume if the bladder is not completely emptied [70]. Since the early 1970s many researchers have developed nomograms or formulas to more easily estimate GFR using serum creatinine without the need to perform urine collections. There are many equations available to clinicians, and some are less cumbersome and more practical to use than others. Six of the main methods used clinically to estimate GFR, and therefore renal function in adults, are outlined in Table 9, with the relative positive and negative aspects of each of the formulas listed in Table 10. Clinically the Cockcroft and Gault equation (or a variation of this equation) is the most commonly used in both the clinical and research settings where urine collection is not deemed practical or necessary. When utilizing the Cockcroft and Gault equation to determine creatinine clearance, the value obtained is most likely an overestimation of the GFR, because between 10 and 60% of creatinine undergoes renal tubular secretion, and therefore appropriate clinical judgment should be exercised [70–72]. It is also important to note that the Jaffe reaction used to measure serum creatinine is based

Table 9. Equations Used to Estimate Creatinine Clearance [16,73] Equation Cockcroft–Gault (males)a Cockcroft–Gault (females)a Jelliffe (males)b Jelliffe (females)b Walsera,c Mawer (males)a Mawer (females)a Mawer (males)a Mawer (females)a Wagner (males)a Wagner (females)a Hull (males)d Hull (females)d Formula CrCl ‫([ ס‬140 - age) ‫ ן‬IBW]/SCr ‫ ן‬72 Male value ‫ ן‬0.85 CrCl ‫[ ס‬98 - 16 ‫( ן‬age - 20)/20]/SCr Male value ‫ ן‬0.90 (GFR)(3/ht2) ‫ ס‬a ‫ ם‬b (cr)‫מ‬1 ‫ ם‬c (age) ‫ ם‬d (wt) CrCl ‫ ס‬TBW ‫[ ן‬29.3 - (0.203 ‫ ן‬age)] CrCl ‫ ס‬TBW ‫[ ן‬25.3 - (0.175 ‫ ן‬age)] CrCl ‫ ס‬IBW [29.3 - 0.203(age)][1 - 0.03(SCr)]/[14.4(SCr)] CrCl ‫ ס‬IBW [25.3 - 0.175(age)][1 - 0.03(SCr)]/[14.4(SCr)] Log CrCl ‫ ס‬2.008 - 1.19 log SCr Log CrCl ‫ ס‬1.888 - 1.20 log SCr CrCl ‫ ס‬145 - age - 3/SCr Male value ‫ ן‬0.85

CrCl ‫ ס‬creatinine clearance; GFR ‫ ס‬glomerlar filtration rate; SCr ‫ ס‬serum creatinine; IBW ‫ ס‬ideal body weight (in kilograms); TBW ‫ ס‬total body weight (in kilograms); IBW males ‫ ס‬50 ‫ ם‬2.3(each inch > 60 inches); IBW females ‫ ס‬45 ‫ ם‬2.3(each inch > 60 inches). a CrCl in ml/min. b CrCl in ml/min/1.73 m2. c For the Walser equation: a ‫מ ס‬6.66 (males) or ‫מ‬4.81 (females); b ‫ם ס‬7.57 (males) or ‫ם‬6.05 (females); c ‫מ ס‬0.103 (males) or ‫מ‬0.080 (females); d ‫ם ס‬0.096 (males) or ‫ם‬0.080 (females); cr ‫ ס‬SCr in millimoles; wt ‫ ס‬kilograms; ht ‫ ס‬meters. d CrCl in ml/min/70 kg.

286 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

Table 10. Positive and Negative Aspects of Equations Used to Estimate Creatinine Clearance [16,73] Equation Cockcroft–Gault Jelliffe Walser Wagner Hull
CrCl ‫ ס‬creatinine clearance, IBW ‫ ס‬ideal body weight.

Positive Aspects Considered ‘‘standard method for estimating GFR’’ Not complicated, therefore rapid bedside estimation of CrCl Attempts to measure GFR and not CrCl

Negative Aspects Overestimates GFR; does not consider nonrenal clearance No adjustment for body size, estimates CrCl, urinary creatinine is constant Number of variables and correlation factors needed for each gender Does not consider age, IBW, nonrenal clearance Does not consider IBW and nonrenal clearance

on the red color that is produced when creatinine is complexed with alkaline picrate. Other substances present in the blood sample such as glucose, protein, and ascorbic acid are also picked up by the test and provide a reading that is approximately 15% higher than it should be. This inaccuracy can serve to offset some of the overestimation of GFR that is obtained when using the Cockcroft and Gault formula [73]. Limitations of glomerular filtration estimates using creatinine clearance calculations may exist in certain patient populations. Patients who are elderly, cachetic, obese, fluid overloaded, or burned require special attention. In these patients, body weight may not accurately reflect true muscle mass and/or creatinine production. In addition, clearance of aminoglycosides and vancomycin in patients with burns are often enhanced by extrarenal mechanisms and may not correlate well with creatinine clearance [74]. Estimated creatinine clearance may change due to many factors. One is diurnal fluctuations in serum creatinine of up to about 25%, therefore it is desirable to measure serum creatinine at the same time every day [73]. Certain drugs such as cimetidine, trimethoprim, and probenecid can also affect creatinine clearance because they interfere with tubular secretion [70]. Patient factors such as disease state and age (i.e., production of creatinine via muscle mass) need to be considered [17]. In addition, as renal function declines, tubular secretion and GFR are not altered at the same rate. A critically ill patient may experience rapid changes in renal function, making quantitative estimation of GFR extremely difficult. The estimation of creatinine clearance by the formulas listed assumes that creatinine production is stable and that extrarenal elimination of creatinine does not exist. A patient in renal failure does not meet these assumptive criteria [68]. In addition, there can be a considerable lag time between rapid renal function decline and a reflection in increased serum creatinine level [71]. The clinician should be aware of the limitations

of the currently available methods of GFR estimation and use of clinical judgment in order to assess the level of renal function that will be assumed in order to use the medication dosage tables in the appendix.

References
1. Leape LL, Cullen DJ, Dempsey Clapp M, et al. Pharmacist participation on physician rounds and adverse drug events in the intensive care unit. JAMA 1999;282:267–270 2. Bates DW, Cullen DJ, Laird N, et al. Incidence of adverse drug events and potential adverse drug events: implications for prevention. JAMA 1995;274:29–34 3. Wilkins RG, Faragher EB. Acute renal failure in an intensive care unit: incidence, prediction and outcome. Anaesthesia 1983;38:628–634 4. Chertow GM, Christiansen CL, Cleary PD, et al. Prognostic stratification in critically ill patients with acute renal failure requiring dialysis. Arch Intern Med 1995;155:1505–1511 5. Kraman S, Khan F, Patel S, et al. Renal failure in the respiratory intensive care unit. Crit Care Med 1979;7:263–266 6. Menashe PI, Ross SA, Cottlieb JE. Acquired renal insufficiency in critically ill patients. Crit Care Med 1988;16: 1106–1109 7. Brivet FG, Kleinknecht DJ, Loirat P, et al. Acute renal failure in the intensive care units—causes, outcome, and prognostic factors of hospital mortality: a prospective, multicenter study. Crit Care Med 1996;24:192–198 8. Groeneveld ABJ, Tran DD, van der Meulen J, et al. Acute renal failure in the medical intensive care unit: predisposing, complicating factors and outcome. Nephron 1991;59: 602–610 9. Maher ER, Robinson KN, Scoble JE, et al. Prognosis of critically-ill patients with acute renal failure: APACHE II score and other predictive factors. Q J Med 1989;72:857–866 10. Cantu TG, Ellerbeck EF, Yun SW, et al. Drug prescribing for patients with changing renal function. Am J Hosp Pharm 1992;49:2944–2948 11. Smith JW, Seidl LG, Cluff LE. Studies on the epidemiology of adverse drug reactions. V. Clinical factors influencing susceptibility. Ann Intern Med 1966;65:629–640 12. Bates DW, Spell N, Cullen DJ, et al. The costs of adverse drug events in hospitalized patients. JAMA 1997;277: 307–311 13. Classen DC, Pestotnik SL, Evans S, et al. Adverse drug events in hospitalized patients. Excess length of stay, extra costs, and attributable mortality. JAMA 1997;277:301–306

DeBellis et al.: Drug Dosing in Renal Failure 287

14. Bates DW, Leape LL, Cullen DJ, et al. Effect of computerized physician order entry and a team intervention on prevention of serious medication errors. JAMA 1998;280:311–316 15. Winter ME, ed. Basic clinical pharmacokinetics, 3rd ed. Vancouver: Applied Therapeutics, 1994:1–499 16. Regardh CG. Factors contributing to variability in drug pharmacokinetics. IV. Renal excretion. J Clin Hosp Pharm 1985;10:337–349 17. Talbert RL. Drug dosing in renal insufficiency. J Clin Pharmacol 1994;34:99–110 18. Brezis M, Rosen S, Epstein FH. Acute renal failure. In: Brenner BM, Rector WG, eds. The kidney. Philadelphia: WB Saunders, 1996:735–779 19. Reed WE, Sabatini S. The use of drugs in renal failure. Semin Nephrol 1986;6:259–295 20. Nissenseon AR. Acute renal failure: definition and pathogenesis. Kidney Int 1998;53(suppl 66):S7–S10 21. Power BM, Forbes AM, Vernon van Heerden P, et al. Pharmacokinetics of drugs used in critically ill adults. Clin Pharmacokinet 1998;34:26–48 22. Shargel L, Yu A. Drug elimination and clearance concepts. In: Shargel L, Yu A, eds. Applied biopharmaceutics and pharmacokinetics, 4th ed. Stamford, CT: Appleton and Lange, 1999:325–351 23. Matzke GR, Millikin SP. Influence of renal function and dialysis on drug disposition. In: Evans WE, Schentag JJ, Jusko WJ, et al., eds. Applied pharmacokinetics principles of therapeutic drug monitoring, 3rd ed. Vancouver: Applied Therapeutics, 1992:1–39 24. Bonate PL, Reith K, Weir S. Drug interactions at the renal level. Implications for drug development. Clin Pharmacokinet 1998;34:375–404 25. van Ginneken CAM, Russel FGM. Saturable pharmacokinetics in the renal excretion of drugs. Clin Pharmacokinet 1989;16:38–54 26. Bendayan R. Renal drug transport: a review. Pharmacotherapy 1996;16:971–985 27. Somogyi A. Renal transport of drugs: specificity and molecular mechanisms. Clin Exp Pharmacol Physiol 1996;23: 986–989 28. Brown GR. Cephalosporin–probenecid drug interactions. Clin Pharmacokinet 1993;24:289–230 29. Lindberg MC, Cunningham A, Lindberg NH. Acute phenobarbital intoxication. South Med J 1992;85:803–807 30. Matzke GR, Frye R. Drug administration in patients with renal insufficiency: minimizing renal and extrarenal toxicity. Drug Saf 1997;16:205–231 31. McNamee PT, Moore GW, McGeown MG, et al. Gastric emptying in chronic renal failure. Br Med J 1985;291: 310–311 32. Dressman JB, Bass P, Ritschel WA, et al. Gastrointestinal parameters that influence oral medications. J Pharm Sci 1993;82:857–872 33. Aronoff FR, Erbeck KM. Prescribing drugs for dialysis patients. In: Henrich WL, ed. Principles and practice of dialysis. Baltimore: Williams & Wilkins, 1994:89–97 34. Bennet WM, Golper TA. Drug usage in dialysis patients. In: Nissenson AR, Rine RN, Gentile D, eds. Clinical dialysis. Stamford, CT: Appleton and Lange, 1995:806–831 35. Manninen V, Melin J, Apajalahti A, et al. Altered absorption of digoxin in patients given propantheline and metoclopramide. Lancet 1973;1:398–400 36. Guay DR, Awai WM, Findlay JW, et al. Pharmacokinetics and pharmacodynamics of codeine in end stage renal disease. Clin Pharmacol Ther 1988;43:63–71 37. Balant LP, Dayer P, Fabre J. Consequences of renal insufficiency on the hepatic clearance of some drugs. Int J Clin Pharmacol Res 1983;3:459–474

38. Bianchetti G, Graziani G, Brancaccio D, et al. Pharmacokinetics and effects of propranolol in terminal uraemic patients and in patients undergoing regular dialysis treatment. Clin Pharmacokinet 1976;1:373–384 39. Gibson TP, Ciacomini KM, Briggs WA, et al. Propoxyphene and norpropoxyphene plasma concentrations in the anephric patient. Clin Pharmacol Ther 1980;27:665–670 40. Plaisance KI, Drusano GL, Forrest A, et al. Effect of renal function on the bioavailability of ciprofloxacin. Antimicrob Agents Chemother 1990;34:1031–1034 41. Barnes JN, Williams AJ, Tomson MJ, et al. Dihydrocodeine in renal failure: further evidence for an important role of the kidney in the handling of opioid drugs. Br Med J 1985; 290:740–743 42. Matzke GR, Flaherty FJ. Drug dosing in renal failure. In: Young LY, Koda-Kimble MA, eds. Applied therapeutics: the clinical use of drugs. Vancouver: Applied Therapeutics, 1988:571–586 43. Doucet J, Fresel J, Hue G, et al. Protein binding of digitoxin, valproate and phenytoin in sera from diabetics. Eur J Pharmacol 1993;45:577–579 44. Perucca E, Richens A. Drug interactions with phenytoin. Drugs 1981;21:120–137 45. Taylor JW, Alexander B, Lyon LW. Oral anticoagulant-phenytoin interactions. Drug Intell Clin Pharm 1980;14:669–673 46. MacKichan JJ. Influence of protein binding and the use of unbound (free) drug concentrations. In: Evans WE, Schentag JJ, Jusko WJ, eds. Applied pharmacokinetics, 3rd ed. Vancouver: Applied Therapeutics, 1992:1–48 47. Liponi DF, Winter ME, Tozer TN. Renal function and therapeutic concentrations of phenytoin. Neurology 1984;34: 395–397 48. Swan S, Bennett WM. Drug dosing guidelines in patients with renal failure. West J Med 1992;156:633–638 49. Reidenberg MM. The binding of drugs to plasma proteins and the interpretation of measurements of plasma concentrations of drugs in patients with poor renal function. Am J Med 1974;62:466–470 50. Gibaldi M. Drug distribution in renal failure. Am J Med 1974;62:471–474 51. Marik PE. The treatment of hypoalbuminemia in the critically ill. Heart Lung 1993;22:166–170 52. Klotz U. Pathophysiologic and disease induced changes in drug distribution volumes: pharmacokinetic implications. Clin Pharmacokinet 1976;1:204–218 53. Vree TB, Hekster YA, Anderson PG. Contribution of the human kidney to the metabolic clearance of drugs. Ann Pharmacother 1992;26:1421–1428 54. Davis BB, Mattammal MB, Zenser TV. Renal metabolism of drugs and xenobiotics. Nephron 1981;27:187–196 55. Gibson TP. Renal disease and drug metabolism: an overview. Am J Kidney Dis 1896;8:7–17 56. Anders MW. Metabolism of drugs by the kidney. Kidney Int 1980;18:636–647 57. Touchette MA, Slauhter RL. The effect of renal failure on hepatic drug clearance. DICP Ann Pharmacother 1991;25: 1214–1224 58. Romac DR, Albertson TE. Drug interactions in the intensive care unit. Clin Chest Med 1999;20:385–399 59. Nielson C. Pharmacologic considerations in critical care of the elderly. Clin Geriatr Med 1994;10:71–89 60. Westphal JF, Brogard JM. Drug administration in chronic liver disease. Drug Saf 1997;1:47–73 61. Park GR. Pharmacokinetics and pharmacodynamics in the critically ill patient. Xenobiotica 1993;23:1195–1230 62. Bodenham A, Shelly MP, Park GR. The altered pharmacokinetics and pharmacodynamics of drugs commonly used in critically ill patients. Clin Pharmacokin 1988;14:347–373

288 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

63. Mann HJ, Fuhs DW, Cerra FB. Pharmacokinetics and pharmacodynamics in critically ill patients. World J Surg 1987; 11:210–217 64. Power BM, Millar Forbes A, Van Heerden V, Hett K. Pharmacokinetics of drugs used in critically ill patients. Clin Pharmacokinet 1998;34:26–48 65. Horl WH, Druml W, Stevens PE. Pathophysiology of ARF in the ICU. Int J Artif Organs 1996;19:84–86 66. Morrison G. Drug dosing in the intensive care unit: the patient with renal failure. In: Irwin RS, Cerra FB, Rippe JM, eds. Intensive care medicine. Philadelphia: LippincottRaven, 1999:1023–1057 67. Shuler C, Golper RA, Bennett WM. Prescribing drugs in renal disease. In: Brenner BM, Rector WG, eds. The kidney. Philadelphia: WB Saunders, 1996:2653–2692 68. Lazarus JM, Enker Owen WF. Hemodialysis. In: Brenner BM, Rector WG, eds. The kidney. Philadelphia: WB Saunders, 1996:740–843 69. Danovitch GM, Wilkinson AH. Evaluation of renal function. In: Massry SG, Glassock RJ, eds. Textbook of nephrology. Baltimore: Williams & Wilkins, 1989:1619–1627 70. Levey AS, Perrone RD, Madias NE. Serum creatinine and renal function. Ann Rev Med 1988;39:465–490 71. Rose BD. Clinical assessment of renal function. In: Laufer RS, DeLeo HC, Fitzpatrick JJ, eds. Pathophysiology of renal disease. New York: McGraw-Hill, 1981:1–30 72. Goerdt PJ, Heim-Duthoy KL, Macres M, et al. Predictive performance of renal function estimate equations in renal allografts. Br J Clin Pharmacol 1997;44:261–265 73. Sladen RN. Accurate estimation of glomerular filtration in the intensive care unit: another holy grail? Crit Care Med 1993;21:1424–1427 74. Maderazo EG, Sun H, Jay GT. Simplification of antibiotic dose adjustments in renal insufficiency: the DREM system. Lancet 1992;340:767–770 75. Kullberg MP, Freeman GB, Biddlecome C, et al. Amrinone metabolism. Clin Pharmacol Ther 1981;29:394–401 76. Blackwell EW, Briant RH, Conolly ME, et al. Metabolism of isoprenaline after aerosol and direct intrabronchial administration in man and dog. Br J Pharmacol 1974;50: 587–591 77. Stroshane RM, Koss RF, Biddlecome CE, et al. Oral and intravenous pharmacokinetics of milrinone in human volunteers. J Pharm Sci 1984;73:1438–1441 78. Desjars P, Pinaud M, Bugnon D, et al. Norepinephrine therapy has no deleterious renal effects in human septic shock. Crit Care Med 1989;17:426–429 79. Hengstmann JH, Goronzy J. Pharmacokinetics of 3H-phenylephrine in man. Eur J Clin Pharmacol 1982;21:335–341 80. Doherty JE, Flanigan WJ, Murphy ML, et al. Tritiated digoxin. XIV. Enterohepatic circulation, absorption and excretion studies in human volunteers. Circulation 1970;42: 867–873 81. Mazur JE, Devlin JW, Peters MJ, et al. Single versus multiple doses of acetazolamide for metabolic alkalosis in critically ill medical patients: a randomized, double-blind trial. Crit Care Med 1999;27:1257–1261 82. Vidt DG. Mechanism of action, pharmacokinetics, adverse effects and therapeutic uses of amiloride hydrochloride, a new potassium sparing diuretic. Pharmacotherapy 1981;1: 179–187 83. Pillay VK, Schwartz FD, Aimi K, et al. Transient and permanent deafness following treatment with ethacrynic acid in renal failure. Lancet 1969;1:77–79 84. Bogaert MG, Rosseel MT, Boalaert J, et al. Fate of isosorbide dinitrate and mononitrates in patients with renal failure. Eur J Clin Pharmacol 1981;21:73

85. Vlasses PH, Larijani GE, Conner DP, et al. Enalapril, a nonsulfhydryl angiotensin-converting enzyme inhibitor. Clin Pharmacol 1985;4:27–40 86. Wolter K, Fritschka E. Pharmacokinetics and pharmacodynamics of quinaprilat after low dose quinapril in patients with terminal renal failure. Eur J Clin Pharmacol 1993; 44(suppl 1):S53–S56 87. Fillastre JP, Baguett JC, Dubois D, et al. Kinetics, safety and efficacy of ramipril after long term administration in hemodialyzed patients. J Cardiovasc Pharmacol 1996;27: 269–274 88. Esmail ZN, Loewen PS. Losartan as an alternative to ACE inhibitors in patients with renal dysfunction. Ann Pharmacother 1998;32:1096–1098 89. Rindone JP, Sloane EP. Cyanide toxicity from sodium nitroprusside: risks and management. Ann Pharmcother 1992; 26:515–520 90. Frishman WH, Michelson EL, Johnson BF, et al. Multi-clinic comparison of labetalol to metoprolol in treatment of mild to moderate systemic hypertension. Am J Med 1983; 75(suppl 4A):54–67 91. Dumas M, d’Athis P, Besancenot JF, et al. Variations of sotalol kinetics in renal insufficiency. Int J Clin Pharmacol Ther Toxicol 1989;27:486–489 92. Carlson RV, Bailey RR, Begg EJ, et al. Pharmacokinetics and effect on blood pressure of doxazosin in normal subjects and patients with renal failure. Clin Pharmacol Ther 1986;40:561–566 93. Titmarsh S, Monk JP. Terazosin: a review of its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy in essential hypertension. Drugs 1987;33:461–477 94. Kirch W, Ramsch KD, Duhrsen U, et al. Clinical pharmacokinetics of nimodipine in normal and impaired renal function. Int J Clin Pharmacol Res 1984;4:381–384 95. Kobayashi K, Maeda K, Koshikawa S, et al. Antithrombotic therapy with ticlopidine in chronic renal failure patients on maintenance hemodialysis. A multicenter collaborative double blind study. Thromb Res 1980;20:255–261 96. Bleich SD, Adgey AAJ, McMechan SR, et al. An angiographic assessment of alteplase: double-bolus and front-loaded infusion regimens in myocardial infarction. Am Heart J 1998; 136:741–748 97. Campbell TJ, Williams KM. Therapeutic drug monitoring: antiarrhythmic drugs. Br J Clin Pharmacol 1998;46:307–319 98. Waller ES. Pharmacokinetic principles of lidocaine dosing in relation to disease state. J Clin Pharmacol 1981;21: 181–194 99. Patel IH, Sugihara JG, Weinfeld RE, et al. Ceftriaxone pharmacokinetics in patients with various degrees of renal impairment. Antimicrob Agents Chemother 1984;25:438–442 100. Mihindu JCL, Scheld WM, Bolton ND, et al. Pharmacokinetics of aztreonam in patients with various degrees of renal dysfunction. Antimicrob Agents Chemother 1983;24: 252–261 101. Leo RJ, Ballow CH. Seizure activity associated with imipenem use: clinical case reports and review of the literature. Ann Pharmacother 1991;25:351–354 102. Calandra G, Lydick E, Carrigan J, et al. Factors predisposing to seizures in seriously ill infected patients receiving antibiotics: experience with imipenem/cilastatin. Am J Med 1988; 84:911–918 103. Balfour HH, McMonigal KA, Bean B. Acyclovir therapy of varicella-zoster virus infections in immunocompromised patients. J Antimicrob Chemother 1983;12(suppl B): 169–179 104. Shepp DH, Dandliker PS, de Miranda P, et al. Activity of 9(2-hydroxy-1-(hydroxymethyl) ethoxymethyl) guanine in

DeBellis et al.: Drug Dosing in Renal Failure 289

the treatment of cytomegalovirus pneumonia. Ann Intern Med 1985;103:368–373 105. Szeto HH, Inturrisi CE, Houde R, et al. Accumulation of normeperidine—an active metabolite of meperidine in patients with renal failure or cancer. Ann Intern Med 1977; 86:738–741 106. Portenoy RK, Foley KM, Stulman J, et al. Plasma morphine and morphine-6-glucuronide during chronic morphine therapy for cancer pain: plasma profiles, steady-state con-

centrations and the consequences of renal failure. Pain 1991;47:13–19 107. Buckley MMT, Brogden RN. Ketorolac: a review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential. Drugs 1990;39:86–109 108. Janssens J, Peeters TL, Vantrappen G, et al. Improvement of gastric emptying in diabetic gastroparesis by erythromycin: preliminary studies. N Engl J Med 1990;322:1028–1031

Appendix: Creatinine Clearance Calculated According to Cockroft and Gault
CrCl (30–50 ml/min) CrCl (10–30 ml/min) CrCl Hemodialysis (<10 ml/min) (HD) Continuous Hemoperfusion (CAVHD/ CVVHD)

Drug

Normal Dose

Notes

Pharmacologic vasopressors Dobutamine (Dobutrex)௡ 2.5–15 ␮g/kg/ min titrate for response No change No change No change No change No change Liver metabolized; increased doses may lead to more frequent atrial and ventricular arrhythmias Increased doses may lead to more frequent atrial and ventricular arrhythmias Used as a second- or thirdline agent in refractory hypotension

Dopamine (Intropin)௡

2–5 ␮g/kg/min (low); 5–10 ␮g/ kg/min (medium); titrate for response up to 50 ␮g/kg/min

No change

No change

No change

No change

No change

Epinephrine (Adrenaline)௡

0.1–1 ␮g/kg/ No change min in refractory hypotension; 2–10 ␮g/min in septic or cardiogenic shock; all doses to be titrated for response 2–10 ␮g/min; titrate for desired heart rate 2–12 ␮g/min; titrate for blood pressure No change

No change

No change

No change

No change

Isoproterenol (Isuprel)௡

No change

No change

No change

No change

50–80% of drug is excreted renally, no dose adjustment data available [76] Doses may drastically exceed 12 ␮g/ min; titrate to desired blood pressure [78] 80–86% excreted renally; no data on dose adjustment [79] 40% of amrinone is excreted unchanged in urine [78]

Norepineprhine (Levophed)௡

No change

No change

No change

No change

No change

Phenylephrine (Neosynephrine)௡

25–60 ␮g/min; titrate for blood pressure

No change

No change

No change

No change

No change

Inotropes Amrinone (Inocor)௡ 0.75 mg/kg load; 5–10 ␮g/ kg/min; titrate for effect No change No change 0.37 mg/kg load; 2.5–5 ␮g/kg/min; titrate for effect No data No change

290 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

Appendix: Continued
CrCl (30–50 ml/min) Same LD; IV/ PO MD 0.125–0.375 mg every 36 hr CrCl (10–30 ml/min) Same LD; IV/ PO MD 0.125–0.375 mg every 36 hr CrCl Hemodialysis (<10 ml/min) (HD) 0.625 mg LD; No IV/PO MD supplement 0.125–0.375 for HD mg every 48 hr Continuous Hemoperfusion (CAVHD/ CVVHD) Normal LD; IV/ PO MD 0.125–0.375 mg every 36 hr

Drug Digoxin (Lanoxin)௡

Normal Dose IV 0.4–1 mg/day LD, 0.125–0.375 mg/day MD; PO 0.75–1.25 mg/ day LD, 0.125–0.375 mg/ day MD (IV and PO loading doses are to be administered every 6–8 hr) 1–5 mg IV bolus every 30–60 min, or 1–10 mg/hr IV infusion 50–75 ␮g/kg load over 10 minutes, 0.375–0.75 ␮g/ kg/min infusion based on clinical response 500 mg IV ‫ ן‬1 dose

Notes The Vd of digoxin can decrease by up to 50% in patients with renal failure, necessitating dose adjustment [80]

Glucagon

No change

No change

No change

No change

No change

Milrinone (Primacor)௡

0.33–0.43 ␮g/ kg/min based on CrCl 30–50 ml/min, respectively

0.23–0.28 ␮g/ kg/min based on CrCl 30–50 ml/min, respectively

25–50 ␮g/kg load, 0.19– 0.56 ␮g/kg/ min infusion based on clinical response Not effective

No data

Use normal dose

85% of dose eliminated unchanged in urine within 24 hr [80]

Preload reducers (diuretics/nitrates) Acetazolamide (Diamox)௡ No change No change Not effective Not effective Use in nonchloride responsive metabolic alkalosis [81] Diuretic effect plateaus when doses exceed 40 mg/day [82] Maximum dose not to exceed 10 mg intermittent, or 12 mg/ infusion/24 hr in patients with normal renal function Thiazide diuretics are not recommended for use alone where CrCl is <30 ml/min; demonstrate synergism with loop diuretics in renal failure Ototoxicity prevalent in patients with GFR <10 ml/min [83] Doses up to 1,500 mg/day IV have been used and are recognized as such in the package insert; continuous infusion may also be used

Amiloride (Midamor)௡

5–20 mg/day PO

2.5–10 mg/day PO

2.5–10 mg/day PO

Not effective

Not effective

Not effective

Bumetanide (Bumex)௡

1–2 mg IV every No change 8–12 hr or 1 mg bolus followed by 0.912 mg/hr continuous infusion ‫ ן‬12 hr

No change

8–10 mg PO No change or IV single dose; or an infusion of 12 mg over 12 hr yields maximal response Not effective Not effective

No change

Chlorothiazide (Diuril)௡

0.5–1 g IV every No change 12–24 hr

No change

Not effective

Ethacrynic acid (Edecrin)௡

50–100 g IV every 8 hr

50–100 mg IV every 8–12 hr

50–100 mg IV every 8–12 hr

Do not use

Do not use

Do not use

Furosemide (Lasix)௡

40–80 mg IV No change every 12 hr; higher doses may be required for desired urine output

No change

No change

Not effective

Not effective

DeBellis et al.: Drug Dosing in Renal Failure 291

Appendix: Continued
CrCl (30–50 ml/min) No change CrCl (10–30 ml/min) No change CrCl Hemodialysis (<10 ml/min) (HD) Not effective Not effective Continuous Hemoperfusion (CAVHD/ CVVHD) Not effective

Drug Hydrochlorothiazide (Hydrodiuril)௡

Normal Dose 25–50 mg PO bid for hypertension; 25–200 mg/day PO in 1–3 divided doses for edema

Notes Hydrochlorothiazide should not be used in patients with SCr >2.5 mg/dl

Isosorbide dinitrate (Isordil)௡ Metolazone (Zaroxylyn)௡

10–40 mg PO tid No change

No change

No change

Administer 10–40 mg PO tid Dose alteration dose after HD not necessary in renal failure [84] Not effective Not effective Although thiazides may not be effective in renal failure, metolazone has demonstrated efficacy with GFRs at 20 ml/ min

5–20 mg/day PO

No change

No change

No change

Nitroglycerin (Nitro Bid, Nitrostat, Nitrol, Nitrodur)௡ Spironolactone (Aldactone)௡

Many methods and routes of dosing 25–200 mg/day PO in 2–4 divided doses for edema; 50–100 mg/day for essential hypertension; 100 mg/day PO to start, increase to 200–400 mg/ day PO in 2–4 divided doses for ascites 10–20 mg IV/PO every 24 hr, doses may be titrated for clinical response 50–100 mg PO bid

No change

No change

No change

Guidelines not determined Not effective

No change

12.5–100 mg PO 12.5–100 mg PO Not effective every 24 hr every 24 hr

Not effective

Use should be avoided in patients with GFRs <10 ml/ min; risk of increased potassium

Torasemide (Demadex)௡

No change

No change

No change

Not effective

Not effective

Doses should not exceed 200 mg/day

Triamterene (Dyrenium)௡

No change

No change

Not effective

Not effective

Not effective

Avoid in patients with serum creatinine >2.5 mg/dl; risk of hyperkalemia Maximum dose should not exceed 40 mg; fixed doses of benazepril and amlodipine together should not be given to patients with SCr >3.0 mg/dl

Angiotensin converting enzyme (ACE) inhibitors Benazepril (Lotensin)௡ 10–40 mg PO every 24 hr 5–20 mg PO every 24 hr 5–20 mg PO every 24 hr 5–20 mg PO every 24 hr 5–20 mg PO every 24 hr 5–20 mg PO every 24 hr

Captopril (Capoten)௡

25–100 mg PO every 8 hr

18.75–75 mg PO 18.75–75 mg PO 12.5–50 mg every 12–18 hr every 12–18 hr PO every 24 hr

Supplement 18.75–75 mg PO 25–30% of every 12–18 hr dose after HD

292 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

Appendix: Continued
CrCl (30–50 ml/min) 2.5–7.5 mg PO every 12 hr CrCl (10–30 ml/min) Start at 2.5 mg PO every 24 hr and titrate up for BP control CrCl Hemodialysis (<10 ml/min) (HD) Start at 2.5 mg PO every 24 hr and titrate up for BP control Administer 25% of current dose in regimen after HD Continuous Hemoperfusion (CAVHD/ CVVHD) 2.5–7.5 mg PO every 12 hr

Drug Enalapril (Vasotec)௡

Normal Dose 5–10 mg PO every 12 hr

Notes Reduced initial doses are required in patients with CrCl <30 ml/min [85]

Enalaprilat (Vasotec)௡

1.25–5 mg IV every 6 hr

1.25–2.5 mg IV every 6 hr

0.625 mg IV ‫ן‬ 1; if after 1 hr there is an inadequate response, up to 1.25 mg may be given at 6-hr intervals 10–40 mg PO every 24 hr

0.625 mg IV 0.625 mg IV ‫ ן‬1; if after 1 every 6 hr hr there is an inadequate response, up to 1.25 mg may be given at 6-hr intervals 7.5–30 mg PO every 24 hr No change

0.625 mg IV ‫ן‬ 1; if after 1 hr there is an inadequate response, up to 1.25 mg may be given at 6-hr intervals 10–40 mg PO every 24 hr Hepatobiliary elimination, compensates for lack of renal elimination, dose adjustments are minimal Fixed dose combinations of lisinopril and hydrochlorothiazide should not be used where the GFR is less than 30 ml/min 2.5 mg in HD patients provides up to 90% inhibition of ACE 24 hr after dose [86]

Fosinopril (Monopril)௡

10–40 mg PO every 24 hr

10–40 mg PO every 24 hr

Lisinopril (Zestril)௡

10–40 mg PO every 24 hr

5–30 mg PO every 24 hr

5–30 mg PO every 24 hr

2.5–20 mg PO every 24 hr

Start at 2.5 mg 5–30 mg PO for initial every 24 hr dose; if patient has dosing regimen administer 20% patients dose after HD Start at 2.5 mg for initial dose; if patient has dosing regimen administer 25–35% of patients dose after HD 20% of the patients dose should be supplemented after HD 7.5–60 mg PO every 24 hr

Quinapril (Accupril)௡

10–80 mg PO every 24 hr

7.5–60 mg PO every 24 hr

7.5–60 mg PO every 24 hr

7.5–60 mg PO every 24 hr

Ramipril (Altace)௡

2.5–20 mg PO every 24 hr

1.25–15 mg PO every 24 hr

1.25–15 mg PO every 24 hr

1.25–10 mg PO every 24 hr

1.25–15 mg PO every 24 hr

2.5 mg three times a week after a 4-hr hemodialysis session was safe and effective in controlling BP in HD patients [87] Start with lower doses in patients with renal failure and slowly titrate Losartan is not better tolerated than ACE inhibitors in causing renal toxicity [88]

Angiotensin II receptor blockers Candesartan (Atacand)௡ 8–32 mg PO every 24 hr No change No change No data No change No change

Losartan (Cozaar)௡

25–100 mg PO every 12–24 hr

No change

No change

No change

No data

No change

Irbesartan (Avapro)௡

150–300 mg PO every 24 hr

No change

No change

No change

No change

No change

DeBellis et al.: Drug Dosing in Renal Failure 293

Appendix: Continued
CrCl (30–50 ml/min) No change CrCl (10–30 ml/min) No change CrCl Hemodialysis (<10 ml/min) (HD) No data No data Continuous Hemoperfusion (CAVHD/ CVVHD) No change

Drug Valsartan (Diovan)௡ Diazoxide (Proglycem)௡

Normal Dose 80–320 mg PO every 24 hr

Notes

Direct-acting vasodilators 50–100 mg IV No change every 5–15 min for severe hypertension, repeat every 4–24 hr to maintain blood pressure control; or 15–30 mg/min IV infusion to a total dose of 5 mg/kg or adequate blood pressure control is achieved 10–100 mg IV/ No change PO every 6–8 hr 2.5–40 mg PO every 12 hr No change No change No change No change No change

Hydralazine (Apresoline)௡ Minoxidil (Loniten)௡

No change

10–100 mg IV/PO every 12 hr No change

10–100 mg IV/PO every 12 hr No change

10–100 mg IV/ PO every 8–12 hr No change Smaller doses may be required in RF or HD; doses on HD days should be administered after HD or 8 hr prior to next HD if late in the day

No change

Nitroglycerin (see above) Nitroprusside (Nipride, Nitropress)௡

See above 0.25–10 ␮g/kg/ min; titrate for blood pressure control No change No change No change No change No change Thiocyanate, a toxic metabolite, accumulates in RF causing seizure and coma, thiocyanate is HD [89]; check thiocyanate levels every 24–48 hr with normal renal function and daily with impaired renal function Elimination half-life of acebutolol is unchanged, active metabolite (diacetolol) is prolonged, warranting dose adjustment In patients with CrCl <10 ml/ min, doses >25 mg should rarely be used

␤-adrenergic receptor blockers Acebutolol (Sectral)௡ 400–800 mg PO every 12–24 hr 200–400 mg PO every 12–24 hr 200–400 mg PO every 12–24 hr 100–200 mg PO every 12–24 hr 100–200 mg PO every 12–24 hr 200–400 mg PO every 12–24 hr

Atenolol (Tenormin)௡

50–100 mg PO every 24 hr

25–50 mg PO every 48 hr

25–50 mg PO every 48 hr

25–50 mg PO every 96 hr

25–50 mg PO every 96 hr, supplement by 12.5–25 mg post-HD

25–50 mg PO every 48 hr

294 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

Appendix: Continued
CrCl (30–50 ml/min) No change CrCl (10–30 ml/min) No change CrCl Hemodialysis (<10 ml/min) (HD) No change No data Continuous Hemoperfusion (CAVHD/ CVVHD) No change

Drug Carvedilol (Coreg)௡

Normal Dose 6.25–50 mg PO every 12 hr

Notes CHF dosing should begin at 3.125 mg every 12 hr and be increased to 6.25 mg as soon as tolerated and doubled every 2 weeks to maximum tolerated dose The half-life is 2 min, once drip is off; no lingering effects

Esmolol (Brevibloc)௡

50–200 ␮g/kg/ min; titrate for heart rate

No change

No change

No change

No change

No change

Metoprolol (Lopressor)௡

50–450 mg/day No change PO in two or three divided doses; or 5–15 mg IV every 6 hr titrated for heart rate 40–320 mg/day PO in single or divided doses 20–160 mg/day in single or divided doses; or use normal dose and change interval to every 24–36 hr No change

No change

No change

Supplement 50 mg PO after HD

No change

Nadolol (Corgard)௡

20–160 mg/day in single or divided doses; or use normal dose and change interval to every 24–48 hr No change

10–80 mg/ day in single or divided doses; or use normal dose and change interval to every 48 hr No change

Supplement 40 mg after HD

20–160 mg/day in single or divided doses; or use normal dose and change interval to every 24–48 hr No change

Alteration of the interval instead of dose is an option

Pindolol (Visken)௡

10–40 mg PO every 12 hr

No change

The half-life in renal failure increases by a factor of 1.4–1.5; dose adjustments are not deemed necessary [90]

Propranolol (Inderal)௡ Sotalol (Betapace)௡

80–320 mg PO every 6–12 hr

No change

No change Lengthen dosing interval to every 36–48 hr based on clinical response No change

No change Dose according to clinical response

No change Supplement 80 mg after HD

No change Lengthen dosing interval to every 36–48 hr based on clinical response No change Data suggests that HD patients need both a decrease in dose and increase in interval [91] A prolonged effect may be demonstrated in renal failure patients; although not specified a reduced dose or extended interval may be necessary [92] Administration is not contraindicated with renal failure, caution should be used

80–320 mg PO Lengthen every 12 hr; start dosing interval with 80 mg PO to every 24 hr every 12 hr

␣-Adrenergic receptor blockers Doxazosin (Cardura)௡ 1–16 mg PO every 24 hr No change No change No change

Phenoxybenzamine (Dibenzyline)௡

10–40 mg PO every 8–12 hr; start low and titrate for BP response

Guidelines not determined

Guidelines not determined

Guidelines not determined

Guidelines not determined

Guidelines not determined

DeBellis et al.: Drug Dosing in Renal Failure 295

Appendix: Continued
CrCl (30–50 ml/min) No change CrCl (10–30 ml/min) No change CrCl Hemodialysis (<10 ml/min) (HD) No change No change Continuous Hemoperfusion (CAVHD/ CVVHD) No change

Drug Phentolamine (Regitine)௡

Normal Dose 5–15 mg IV every 15 min in hypertensive crisis or pheochromocytoma 1–15 mg PO every 12 hr

Notes

Prazosin (Minipress)௡

No change

No change

No change

No change

No change

Patients with end-stage renal disease may respond to lower dosages [48] Doses >20 mg/ day do not provide further blood pressure reduction [93]

Terazosin (Hytrin)௡

1–20 mg/day PO

No change

No change

No change

No change

No change

Sympatholytics for blood pressure control Clonidine (Catapres)௡ 0.1–0.6 mg PO every 12 hr for hypertension; 0.1–0.2 mg can be given every hour for malignant hypertension to a maximum of 0.7 mg 10–100 mg PO every 24 hr, maximum dose 100 mg/day in two divided doses 250–500 mg PO every 8–12 hr, maximum dose PO 3 g/day; or 250–500 mg IV every 6 hr, maximum 1 g IV every 6 hr No change No change No change No change No change

Guanethidine (Ismelin)௡

No change

No change

10–100 mg PO every 36 hr

No data

10–100 mg/day PO every 24hr

Methyldopa (Aldomet)௡

250–500 mg PO/IV every 12 hr

250–500 mg PO/IV every 12 hr

250–500 mg PO/IV every 24 hr

Administer 250 mg after each HD

250–500 mg PO/IV every 12 hr

The half-life of methyldopa is significantly prolonged in patients with ESRD

Trimethaphan (Arofonad)௡

0.3–6 mg/min No change IV; titrate for BP control 2.5–10 mg PO every 24 hr 200–400 mg PO every 24 hr No change No change

No change

No change

No change

No change

Calcium channel blockers Amlodipine (Norvasc)௡ Bepridil (Vascor)௡ No change No change No change No change No change No change No change No change Bepridil does not appear to be effected by renal impairment, however, trials are ongoing

Diltiazem (Dilacor, Cardizem, Tiazac)௡

30–90 mg PO No change every 6–8 hr; for atrial fibrillation, administer IV at 20 mg ‫ ן‬1 (0.25 mg/kg), if no response in 15 min, give 25 mg ‫ ן‬1 (0.35 mg/ kg) then begin infusion at 5–15 mg/hr for rate control

No change

No change

No change

No change

296 Journal of Intensive Care Medicine Vol 15 No 6 November/December 2000

Appendix: Continued
CrCl (30–50 ml/min) No change CrCl (10–30 ml/min) No change CrCl Hemodialysis (<10 ml/min) (HD) No change No change Continuous Hemoperfusion (CAVHD/ CVVHD) No change

Drug Felodipine (Plendil)௡

Normal Dose 5–15 mg PO every 8–24 hr

Notes Dosing parameters in renal failure may vary with combination products with felodipine Patients with CrCl of 30–80 ml/min show a significant increase in area under the curve of isradipine, however, this effect returns to normal as renal function deteriorates Manufacturer recommends starting with lowest dose possible in patients with renal failure

Isradipine (Dynacirc)௡

1.25–10 mg/day bid

No change

No change

No change

No change

No change

Nicardipine (Cardene)௡

20–40 mg PO No change tid; or continuous infusion for difficult to control hypertension at 5–15 mg/hr, may increase dose by 2.5 mg/ hr every 5 min 10–30 mg/day PO tid 60 mg PO every 4 hr No change

No change

No change

No change

No change

Nifedipine (Adalat, Procardia)௡ Nimodipine (Nimotop)௡

No change

No change

No change

No change

No change

No change

No change

No change

No change

An increase in elimination halflife as well as an increase in area under the curve was noted in patients with renal failure, however, dosage adjustments are not necessary [94]

Verapamil (Calan, Isoptin, Verelan)௡

40–120 mg PO every 8 hr; IV dose for atrial arrhythmias is 5–10 mg, may repeat until maximum of 20 mg is reached; constant IV infusions run at 5–10 mg/hr 0.25 mg/kg IV bolus, then 0.125 ␮g/kg/ min ‫ ן‬12 hr 81–325 mg PO every 24 hr

No change

No change

20–60 mg PO every 8 hr; IV doses for antiarrhythmic effect, therefore no change

No supplementation necessary, however, observe doses for CrCl <10 ml/min

40–120 mg PO every 8 hr; IV dose for atrial arrhythmias is 5–10 mg, may repeat until maximum of 20 mg is reached

Antiplatelet drugs Abciximab (ReoPro)௡ No change No change No change No change, abciximab may not be cleared by dialysis Dose after dialysis No change, abciximab may not be cleared by dialysis No change Monitor platelets 4 hr into infusion

Aspirin

No change

No change

Use if benefits outweigh risks

May increase bleeding risk in uremic patients

DeBellis et al.: Drug Dosing in Renal Failure 297

Appendix: Continued
CrCl (30–50 ml/min) No change CrCl (10–30 ml/min) No change CrCl Hemodialysis (<10 ml/min) (HD) No change No guidelines determined, administer if benefits outweigh risks Continuous Hemoperfusion (CAVHD/ CVVHD) No guidelines determined, administer if benefits outweigh risks

Drug Clopidogrel (Plavix)௡

Normal Dose Loading doses of 300 mg ‫ ן‬1 may be given followed by 75 mg PO every 24 hr 50–100 mg PO every 6–8 hr Medical management: 180 ␮g/kg IV bolus, then 2 ␮g/kg/min. Catheterization lab: 180 ␮g/kg IV bolus, then 2 ␮g/kg/min IV bolus 10 min into the maintenance infusion 250 mg PO every 12 hr taken with food

Notes Monitor for signs and symptoms of TTP

Dipyradamole (Persantine)௡ Eptifibatide (Integrilin)௡

No change SCr Յ 2 mg/dl: 180 ␮g/kg bolus, then 2 ␮g/kg/min infusion. SCr 2–4 mg/dl: 135 ␮g/kg bolus, then 0.5 ␮g/kg/ min infusion

No change SCr Յ 2 mg/dl: 180 ␮g/kg bolus, then 2 ␮g/kg/min infusion. SCr 2–4 mg/dl: 135 ␮g/kg bolus, then 0.5 ␮g/kg/ min infusion

No change SCr Յ 2 mg/ dl: 180 ␮g/kg bolus, then 2 ␮g/kg/min infusion. SCr 2–4 mg/dl: 135 ␮g/kg bolus, then 0.5 ␮g/kg/ min infusion

No guidelines No guidelines determined determined No data available, eptifibatide may be cleared by dialysis No data available, eptifibatide may be cleared by dialysis If heparin is used, maintain aPTTs in the range of 50 to 70 sec; continue eptifibatide 12–24 hr postangioplasty unless specified otherwise

Ticlopidine (Ticlid)௡

No change, reduce dose or discontinue if hemorrhagic problems are encountered

No change, reduce dose or discontinue if hemorrhagic problems are encountered

No change, reduce dose or discontinue if hemorrhagic problems are encountered

200 mg/day has been used in uremic patients on chronic hemodialysis [98]; administer if benefits outweigh risks No data available, tirofiban is removed by HD

No change, reduce dose or discontinue if hemorrhagic problems are encountered

Monitor CBC every 2 weeks for the first 3 months, also monitor for signs and symptoms of TTP

Tirofiban (Aggrastat)௡

0.4 ␮g/kg/min No change IV bolus over 30 min followed by 0.1 ␮g/kg/min ‫ ן‬48–108 hr

0.2 ␮g/kg/min IV bolus over 30 min followed by 0.05 ␮g/kg/ min ‫ ן‬48–108 hr

0.2 ␮g/kg/ min IV bolus over 30 min followed by 0.05 ␮g/kg/ min ‫ן‬ 48–108 hr

No data available, tirofiban is removed by HD

If heparin is used, maintain aPTTs in the 50to 70-sec range, continue tirofiban 12–24 hr postangioplasty unless otherwise specified Maintain aPTT with heparin at 1.5–2 times control (50–70 sec) for 48 hr

Thrombolytic agents Alteplase (Activase)௡ >67 kg: 15 mg IV No change bolus, 50 mg over 30 min, then 35 mg over the next 60 min. Յ67 kg: 15 mg IV bolus, 0.75 mg/kg over 30 min (up to 50 mg), then 0.50 mg/kg over the next 60 min (up to 35 mg). IV bolus of 30 units infused over 2–5 min Two 10 U IV boluses, administered over 2 min, 30 min apart No change No change, administer if benefits outweigh risks No change, administer if benefits outweigh risks No guidelines determined, administer if benefits outweigh risks No guidelines determined, administer if benefits outweigh risks

Anistreplase (Eminase)௡ r-PA (Retaplase)௡

No change

No change

No guidelines No guidelines determined determined No guidelines determined, administer if benefits outweigh risks No guidelines