Chapter 4. The Aminoglycoside Antibiotics

The aminoglycoside antibiotics are widely used for the treatment of severe gram-negative infections such as pneumonia or bacteremia, often in combination with a β-lactam antibiotic. Aminoglycosides are also used for gram-positive infections such as infective endocarditis in combination with penicillins when antibiotic synergy is required for optimal killing. Aminoglycoside antibiotics available in the United States that are in common use include gentamicin, tobramycin, netilmicin, and amikacin.

Aminoglycoside antibiotics are bactericidal, and the drugs exhibit concentration-dependent bacterial killing.1 Antibiotics with concentration-dependent killing characteristically kill bacteria at a faster rate when drug concentrations are higher. Also, aminoglycosides have a concentration-dependent postantibiotic effect. The postantibiotic effect is the phenomenon of continued bacterial killing even though serum concentrations have fallen below the minimum inhibitory concentration (MIC). Because the postantibiotic effect is concentration-dependent for the aminoglycosides, higher drug concentrations lead to a longer postantibiotic effect. The mechanisms of action for aminoglycosides are binding to the 30S ribosomal subunit inhibiting protein synthesis and misreading of mRNA causing dysfunctional protein production.

The MIC for susceptible bacteria is higher for amikacin than it is for the other aminoglycosides. Because the pharmacokinetics is similar for all these drugs, higher doses of amikacin are needed to treat infections. The conventional method of dosing aminoglycoside antibiotics is to administer multiple daily doses (usually every 8 hours).2 In order to take advantage of concentration-dependent bacterial killing and the postantibiotic effect, extended-interval (usually the total daily dose given once per day) aminoglycoside administration is also a dosing option.3 Because of these two different methods of dosage administration, it is important to identify which is being used when discussing serum concentration monitoring.

Conventional Dosing

Aminoglycoside antibiotics are given as short-term (1/2–1 hour) infusions. If a 1-hour infusion is used, maximum end of infusion “peak” concentrations are measured when the infusion is completed (Figure 4-1). If a 1/2-hour infusion is used, serum concentrations exhibit a distribution phase so that drug in the blood and in the tissues are not yet in equilibrium. Because of this, a 1/2-hour waiting period is allowed for distribution to finish if a 1/2-hour infusion is used before peak concentrations are measured. Therapeutic steady-state peak concentrations for gentamicin, tobramycin, and netilmicin are generally 5–10 μg/mL for gram-negative infections. Infection sites with more susceptible bacteria, such as intra-abdominal infections usually can be treated with steady-state peak concentrations at the lower end of this range (typically 5–7 μg/mL). Infection sites that are difficult to penetrate and with bacteria that have higher MIC values, such as pseudomonal pneumonia usually require steady-state peak concentrations in the higher end of the range (typically 8–10 μg/mL). When gentamicin, tobramycin, or netilmicin are used synergistically with penicillins or other antibiotics for the treatment of gram-positive infections such as infective endocarditis steady-state peak concentrations of 3–5 μg/mL are often times adequate. Therapeutic peak concentrations for amikacin are 15–30 μg/mL.

Exceeding peak steady-state concentrations of 12–14 μg/mL for gentamicin, tobramycin, or netilmicin or 35–40 μg/mL for amikacin when using conventional dosing leads to an increased risk of ototoxicity.4 The types of ototoxicity that aminoglycosides cause are auditory and vestibular, and the damage is permanent. Aminoglycosides accumulate in the lymph of the inner ear causing ongoing damage to cochlear or vestibular sensory cells.1 Auditory ototoxicity usually is first noted at high frequencies (>4000 Hz) and is difficult to detect using clinical means. Audiometry is required to detect high-tone hearing loss and is seldom done in patient care areas. Older patients may have lost the ability to hear in this range for other reasons. If aminoglycoside treatment is not discontinued in individuals with high-frequency auditory ototoxicity, hearing loss will progress to lower frequencies. As a result, aminoglycoside-induced hearing losses are not usually detected until the patient is unable to detect sounds in the conversational frequency zone (<4000 Hz). Often, the first sign of auditory ototoxicity is tinnitus. Vestibular ototoxicity results in the loss of balance. Again, this type of ototoxicity is difficult to detect because many patients treated with aminoglycosides are bed-bound. Besides loss of equilibrium, headache, ataxia, nausea, vomiting, nystagmus, and vertigo can all be signs of vestibular ototoxicity. Although this version of ototoxicity is also permanent, patients can often compensate using visual cues, such as use of the horizon, to maintain balance and avoid ataxia. In some studies, predose (“trough”) steady-state concentrations have been found to be related to ototoxicity.5,6 However, peak steady-state concentrations have also been elevated in these patients which clouds the relationship between serum concentrations and this type of drug-induced adverse effect.

Trough steady-state concentrations (predose or minimum concentrations usually obtained within 30 minutes of the next dose) above 2–3 μg/mL for tobramycin, gentamicin, or netilmicin or 10 μg/mL for amikacin predispose patients to an increased risk of nephrotoxicity.7,8 Aminoglycoside antibiotics accumulate in the proximal tubular cells of the kidney, decrease the ability of the kidney to concentrate urine, and, ultimately, decrease glomerular filtration.9–11 Nephrotoxicity due to aminoglycoside therapy is unlikely to occur before 3–5 days of therapy with proper dosing of the antibiotic. Because many patients receiving aminoglycosides are critically ill, other sources of nephrotoxicity, such as hypotension or other nephrotoxic drug therapy, should be ruled out before a diagnosis of aminoglycoside renal damage is made in a patient. Unlike ototoxicity, aminoglycoside-induced nephrotoxicity is usually reversible with little, if any, residual damage if the antibiotic is withdrawn soon after renal function tests change. With proper patient monitoring, mild renal dysfunction resulting in serum creatinine increases of 0.5–2 mg/dL may be the only result of aminoglycoside nephrotoxicity. However, if the patient develops renal failure, the cost of maintaining the patient on dialysis until kidney function returns can exceed $50,000–$100,000 and, if the patient is critically ill, may contribute to his or her death. In some investigations, peak concentrations have been related to nephrotoxicity.12 However, trough concentrations have also been high in these patients, which obscure the relationship between serum concentrations and nephrotoxicity.

Keeping peak and trough concentrations within the suggested ranges does not completely avoid nephrotoxicity and ototoxicity in patients, but, hopefully, decreases the likelihood that patients will experience these serious adverse effects.13 Also, even though serum concentrations are controlled within the suggested ranges, duration of therapy exceeding 14 days, large total cumulative doses, and concurrent therapy with other nephrotoxic drugs such as vancomycin can predispose patients to these side effects of the aminoglycoside antibiotics.14–17

Extended-Interval Dosing

Because aminoglycoside antibiotics exhibit concentration-dependent bacterial killing and the postantibiotic effect is longer with higher concentrations, investigators began studying the possibility of giving a higher dose of aminoglycoside once daily.3,18,19 Generally, these studies have shown comparable microbiologic and clinical cure rates for many infections and about the same rate of nephrotoxicity (~5–10%) as with conventional dosing. Auditory ototoxicity has not been monitored using audiometry in most of these investigations, but loss of hearing in the conversational range as well as signs and symptoms of vestibular toxicity have usually been assessed and found to be similar to aminoglycoside therapy dosed conventionally. Based on this data, clinicians have begun using extended-interval dosing in selected patients. For Pseudomonas aeruginosa infections where the organism has an expected MIC ≈ 2 μg/mL, peak concentrations between 20 and 30 μg/mL and trough concentrations <1 μg/mL have been suggested.3 At the present time, there is not a consensus on how to approach concentration monitoring using this mode of administration.20–26 Some clinicians measure steady-state peak and trough concentrations while others measure two steady-state postdose concentrations or a single steady-state postdose concentration.27

Because of the extremely high peak concentrations obtained during extended-interval dosing of aminoglycosides, it can be difficult to understand why increased toxicity is not seen in patients. The hypothesized reason is that both nephrotoxicity and ototoxicity are due to accumulation of aminoglycoside in the relevant tissue. Because the dosage interval is prolonged in extended-interval administration, aminoglycoside concentrations are low for a long period of time and may allow for diffusion of drug out of tissue and into the blood which avoids drug accumulation in the ear and kidney. Also, some of the uptake mechanisms into the ear and kidney may be saturable, so that high peak serum concentrations of aminoglycosides may not result in high renal or ear tissue concentrations.

Since large doses of aminoglycoside are given as a single dose with this mode of administration, two additional adverse effects become of concern. Because of the manufacturing process used to produce aminoglycoside antibiotics, very low residual amounts of gram-negative endotoxin are sometimes present in the commercial product. Reports of infusion-related hypotension in patients receiving extended-interval aminoglycosides during the late 1990s have been attributed to the amount of toxin administered at one time.28,29 Acute neuromuscular blockade, usually associated with concurrent administration of anesthetics or neuromuscular blockers, is also a possible adverse effect of aminoglycosides associated with high drug concentrations. Because of the high peak concentrations achieved using extended-interval dosing, surgical and intensive care patients should be monitored for this possible adverse effect.

Differential Toxicity among Aminoglycosides

Studies are available that attempt to determine nephrotoxicity differences among antibiotics. Gentamicin accumulates to a greater extent in kidney tissue when compared to tobramycin.11,13,16 Because doses of amikacin are larger than for gentamicin and tobramycin, amikacin in renal accumulation must be adjusted for dosage differences.9,13 When this is done, amikacin accumulation patterns are similar to gentamicin. Based on these accumulation profiles and associated clinical data and other trials, some clinicians believe that tobramycin is less nephrotoxic than gentamicin or amikacin.30 There are less conclusive data for netilmicin. Other clinical trials that compare the nephrotoxicity potential of gentamicin and tobramycin indicate that the two drugs are similar in this area.31,32 Generally, gentamicin is the most widely used aminoglycoside, followed by tobramycin and netilmicin. This usage pattern is due, in part, to the fact that gentamicin was the first aminoglycoside available generically and was much less expensive than the other drugs for a number of years. Amikacin is usually reserved for use in infections where the organism is resistant to other aminoglycosides.

Clinicians should always consult the patient’s chart to confirm that antibiotic therapy is appropriate for current microbiologic cultures and sensitivities. Also, it should be confirmed that the patient is receiving other appropriate concurrent antibiotic therapy, such as β-lactam or anaerobic agents, when necessary to treat the infection. Patients with severe infections usually have elevated white blood cell counts and body temperatures. Measurement of serial white blood cell counts and body temperatures are useful to determine the efficacy of antibiotic therapy. A white blood cell count with a differential will identify the types of white blood cells that are elevated. A large number of neutrophils and immature neutrophils, clinically known as a “shift to the left,” can also be observed in patients with severe bacterial infections. Favorable response to antibiotic treatment is usually indicated by high white blood cell counts decreasing toward the normal range, the trend of body temperatures (plotted as body temperature vs. time, also known as the “fever curve”) approaching normal, and any specific infection site tests or procedures resolving. For instance, in pneumonia patients the chest x-ray should be resolving, in patients with an intraabdominal infection abdominal pain and tenderness should be decreasing, or in patients with a wound infection the wound should be less inflamed with less purulent discharge. Clinicians should also be aware that immunocompromised patients with a bacterial infection may not be able to mount a fever or elevated white blood cell count.

Aminoglycoside steady-state peak and trough serum concentrations should be measured in 3–5 estimated half-lives when the drug is given using conventional dosage approaches. Methods to estimate this parameter are given in the initial dose calculation portion of this chapter. Since prolongation of the dosage interval is often used in patients with decreased elimination, a useful clinical rule is to measure serum concentrations after the third dose. If this approach is used, the dosage interval is increased in tandem with the increase in half-life so that 3–5 half-lives have elapsed by the time the third dose is administered. Additionally, the third dose typically occurs 1–3 days after dosing has commenced and this is a good time to assess clinical efficacy of the treatment also. Steady-state serum concentrations, in conjunction with clinical response, are used to adjust the antibiotic dose, if necessary. Methods to adjust aminoglycoside doses using serum concentrations are discussed later in this chapter. If the dosage is adjusted, aminoglycoside elimination changes or laboratory and clinical monitoring indicate that the infection is not resolving or worsening, clinicians should consider rechecking steady-state drug concentrations.

When extended-interval aminoglycoside therapy is used, several different monitoring techniques can be used.27 Some clinicians measure steady-state peak and trough concentrations while others measure two steady-state postdose concentrations. Other approaches include obtaining only a steady-state trough concentration, or measuring a single aminoglycoside serum concentration 6–14 hours after a dose and using a dosage nomogram to adjust the dosage interval (please see dosing section later in chapter for details).

Serial monitoring of serum creatinine concentrations should be used to detect nephrotoxicity. Ideally, a baseline serum creatinine concentration is obtained before aminoglycoside therapy is initiated and three times weekly during treatment. An increasing serum creatinine test on two or more consecutive measurement occasions indicates that more intensive monitoring of serum creatinine values, such as daily, is needed. If serum creatinine measurements increase more than 0.5 mg/dL over the baseline value (or >25–30% over baseline for serum creatinine values >2 mg/dL) and other causes of declining renal function have been ruled out (other nephrotoxic drugs or agents, hypotension, etc.), alternatives to aminoglycoside therapy or, if that option is not possible, intensive aminoglycoside serum concentration monitoring should be initiated to ensure that excessive amounts of aminoglycoside do not accumulate in the patient. In the clinical setting, audiometry is rarely used to detect ototoxicity because it is difficult to accomplish in severely ill patients. Instead, clinical signs and symptoms of auditory (decreased hearing acuity in the conversational range, feeling of fullness or pressure in the ears, tinnitus) or vestibular (loss of equilibrium, headache, nausea, vomiting, vertigo, nystagmus, ataxia) ototoxicity are monitored at the same time intervals as serum creatinine determination.

The aminoglycosides are eliminated almost completely (≥90%) unchanged in the urine primarily by glomerular filtration (Table 4-1).10,13,16 These antibiotics are usually given by short-term (1/2–1 hour) intermittent intravenous infusions, although they can be given intramuscularly. When aminoglycosides are given intramuscularly they exhibit very good bioavailability of ~100% and are rapidly absorbed with maximal concentrations occurring about 1 hour after injection. Exceptions to this situation are patients who are hypotensive or obese. Hypotensive patients shunt blood flow away from peripheral tissues, such as muscle, to provide maximal blood flow to internal organs. As a result, intramuscularly administered drugs may be malabsorbed in hypotensive patients, such as those with gram-negative sepsis. Care must be taken with obese individuals to use a long enough needle to penetrate subcutaneous fat and enter muscle tissue when administering aminoglycoside antibiotics. Drug injected into poorly perfused fatty tissue will likely be malabsorbed. Oral bioavailability is poor (<10%) so systemic infections cannot be treated by this route of administration. Plasma protein binding is low (<10%).

Manufacture recommended doses for conventional dosing in patients with normal renal function are 3–5 mg/kg/d for gentamicin and tobramycin, 4–6 mg/kg/d for netilmicin, and 15 mg/kg/d for amikacin. These amounts are divided into three equal daily doses for gentamicin, tobramycin, or netilmicin, or two or three equal daily doses for amikacin. Extended-interval doses obtained from the literature for patients with normal renal function are 4–7 mg/kg/d for gentamicin, tobramycin, or netilmicin and 11–20 mg/kg/d for amikacin.3,19–26,33–38

Nonobese adults with normal renal function (creatinine clearance >80 mL/min, Table 4-1) have an average aminoglycoside half-life of 2 hours (range: 1.5–3 hours), and the average aminoglycoside volume of distribution is 0.26 L/kg (range: 0.2–0.3 L/kg) in this population.39–42 The volume of distribution is similar to extracellular fluid content of the body, and fluid balance will be an important factor when estimating the aminoglycoside volume of distribution for a patient. Patients who have been febrile due to their infections for 24 hours or more may be significantly dehydrated and have lower volumes of distribution until rehydrated.

Because aminoglycosides are eliminated primarily by glomerular filtration, renal dysfunction is the most important disease state that affects aminoglycoside pharmacokinetics.43,44 The elimination rate constant decreases in proportion to creatinine clearance because of the decline in drug clearance (Figure 4-2).45,46 This relationship between renal function and aminoglycoside elimination will form the basis for initial dosage computation later in this chapter. Because the kidney is the organ responsible for maintaining fluid and electrolyte balance in the body, patients with renal failure are sometimes overhydrated. Body weight can be an effective way to detect overhydration in a patient. If the usual weight of the patient is 70 kg when they are in normal fluid balance, known as the patient’s “dry weight,” and the patient is currently 75 kg with signs and symptoms of overhydration (pedal edema, extended neck veins, etc.), the additional 5 kg of weight could be considered extra fluid and added to the estimated volume of distribution for the patient. Since 1 L of water weighs 1 kilogram, the estimated volume of distribution for this patient would be 18.2 L using the patient’s dry weight (V = 0.26 L/kg · 70 kg = 18.2 L) plus 5 L to account for the additional 5 kg of extra fluid yielding a total volume of distribution equal to 23.2 L (V = 18.2 L + 5 L = 23.2 L). Care would be needed to alter the estimated volume of distribution toward normal as the excess fluid was lost and the patient’s weight returned to its usual value.

A major body burn (>40% body surface area) can cause large changes in aminoglycoside pharmacokinetics.47–49 Forty-eight to seventy-two hours after a major burn, the basal metabolic rate of the patient increases to facilitate tissue repair. Because of the increase in basal metabolic rate, glomerular filtration rate increases which increases aminoglycoside clearance. Because of the increase in drug clearance, the average half-life for aminoglycosides in burn patients is ~1.5 hours. If the patient is in normal fluid balance, the average volume of distribution will be the same as in normal adults (0.26 L/kg). However, since the skin is the organ which prevents fluid evaporation from the body and the integrity of the skin has been violated by thermal injury, these patients can be dehydrated, especially if they have had a fever for more than 24 hours. The result is a lower volume of distribution for aminoglycosides. Alternatively, some burn patients may be overhydrated due to vigorous fluid therapy used to treat hypotension. This will result in a larger than expected aminoglycoside volume of distribution. Unfortunately, there is no precise way to correct for fluid balance in these patients. Frequent use of aminoglycoside serum concentrations are used to guide therapy in this population.

Concurrent therapy with some penicillins can increase aminoglycoside clearance by chemically inactivating both the penicillin and aminoglycoside via formation of a covalent bond between the two antibiotic molecules.50–54 Penicillin G, ampicillin, nafcillin, carbenicillin, and ticarcillin are the penicillins most likely to cause this interaction. Piperacillin and mezlocillin, as well as cephalosporins, do not inactivate aminoglycosides to an appreciable extent. This in vivo interaction is most likely to occur in patients with poor renal function (creatinine clearance <30 mL/min) so that the elimination of both the aminoglycoside and penicillin is slower. Under these conditions, serum concentrations of both antibiotics will be higher for a longer period of time and facilitate the inactivation process. In patients with renal failure receiving an aminoglycoside alone, the addition of one of the interacting penicillins can decrease the aminoglycoside half-life from ~50 hours when given alone to ~12 hours when given in combination and result in a dosage increase for the aminoglycoside. Another place where this interaction is important to note is when patients are receiving concurrent therapy with one of the interacting penicillins and an aminoglycoside antibiotic, and serum concentration monitoring of the aminoglycoside is planned. When a blood sample is obtained for measurement of the aminoglycoside serum concentration, penicillin contained in the blood collection tube can continue to inactivate aminoglycoside. This will lead to a spuriously low aminoglycoside concentration results which can lead to dosing adjustment errors. For example, a peak gentamicin serum concentration is obtained in a patient receiving concurrent gentamicin and penicillin G therapy. When the blood sample was drawn from the patient, the gentamicin concentration was 8 μg/mL. By the time the sample is processed by the lab, 6 hours expire because of transportation and other factors. Because of this, penicillin G inactivated aminoglycoside molecules, and the concentration of gentamicin decreased to 4 μg/mL. The lab measured this concentration and reported it to the clinicians caring for the patient. Because the desired peak concentration was 8 μg/mL, the dose of gentamicin was doubled so that the reported peak concentration of 4 μg/mL would increase to the target concentration. Of course, since the actual peak concentration was 8 μg/mL in the patient all along, the new peak concentration resulting from the dosage increase would be 16 μg/mL. In order to prevent this in vitro inactivation interaction in patients receiving concurrent penicillin and aminoglycoside treatment when the drug assay will not be run for longer than 1–2 hours after specimen collection, blood samples should have the serum separated using centrifugation. The serum is removed and placed in a separate tube, then frozen to prevent the chemical reaction from occurring. Alternatively, a small amount of β-lactamase (<5% of total blood volume to prevent sample dilution) can be added to break the β-lactam bond of the penicillin and avoid inactivation of the aminoglycoside antibiotic.

Aminoglycosides are relatively polar molecules with good water solubility. Because of this, they do not enter adipose cells to any significant extent. However, in patients who weigh more that 30% over their ideal body weight, the volume of distribution for aminoglycosides increases because of the additional extracellular fluid contained in adipose tissue (Figure 4-3).55–57 The reason that aminoglycoside volume of distribution is affected by this relatively small amount of additional extracellular fluid in adipose tissue is because the baseline volume of distribution for these drugs is relatively small to begin with (0.26 L/kg or ~18 L for a 70-kg person). For other water-soluble drugs with larger volumes of distribution, the additional extracellular fluid contained in adipose tissue may not be a significant factor. Adipose tissue contains ~40% of the extracellular fluid that is present in lean tissue. To compensate for the increased extracellular fluid of adipose tissue and the greater volume of distribution found in obese patients (>30% over ideal body weight), the following formula can be used to estimate aminoglycoside volume of distribution (V in Liter) for initial dosing purposes: V = 0.26 · [IBW + 0.4(TBW – IBW)], where IBW is ideal body weight and TBW is the patient’s actual total body weight. In morbidly obese (>90% above ideal body weight) patients with normal serum creatinine concentrations, the clearance of aminoglycoside antibiotics is also increased.55–57 The reason for the increased drug clearance is larger kidneys which result in larger creatinine clearance rates. Because both volume of distribution and clearance simultaneously change in obese patients to about the same extent, the aminoglycoside half-life value is appropriate for the patient’s renal function [t1/2 = (0.693 · V)/Cl].

Cystic fibrosis is a disease state that affects exocrine glands. In the lung, the result is the production of thick, tenacious sputum that predisposes patients to pulmonary infections. Patients with cystic fibrosis have larger aminoglycoside volumes of distribution (0.35 L/kg) because body composition is altered.33–36,58–61 Generally cystic fibrosis patients have decreased adipose tissue and increased extracellular fluid due to disease-state-induced gastrointestinal malabsorption. These patients also have higher aminoglycoside clearance values due to increased glomerular filtration rates. Because clearance rates tend to increase more than volume of distribution values, the average aminoglycoside half-life is typically shorter in patients with cystic fibrosis (t1/2 = 1.5 hours). Extended-interval dosing can be used to treat pulmonary exacerbations in cystic fibrosis patients.33,35,62,63 Aminoglycosides can also be administered via inhalation at a dose of 300 mg twice daily in a cyclic fashion (4 weeks on, 4 weeks off) for patients with cystic fibrosis.64

Liver disease patients with ascites have additional extracellular fluid due to accumulation of ascitic fluid.65–67 Since aminoglycosides pass into ascitic fluid, the volume of distribution is increased in these patients. The approach to estimating an initial volume of distribution is similar to that used in renal failure patients who are fluid overloaded. The weight of the patient when ascitic fluid is not present is known as the patient’s dry weight. If this value is not known and the patient is not obese, ideal body weight can be used as an estimate of the dry weight. A reasonable estimate of the volume of distribution (V in liter) for a patient with ascites, or who is overhydrated for other reasons, can be estimated using the following equation: V = (0.26 · DBW) + (TBW – DBW), where DBW is the patient’s dry body weight and TBW is the patient’s actual total body weight. Because of the large amount of variation in aminoglycoside volume of distribution for patients with ascites or overhydration, dosing should be guided by aminoglycoside serum concentrations. Also, as excess fluid is lost, clinicians should anticipate a decrease in the volume of distribution for these drugs.

Premature infants (gestational age ≤34 weeks) have a larger amount of body water compared to adults.37,68–70 Aminoglycoside volume of distribution is larger (0.5–0.6 L/kg) because of this physiologic difference. Additionally, kidneys are not completely developed either so glomerular filtration and aminoglycoside clearance are decreased. A larger volume of distribution and lower clearance rate result in a prolonged average half-life equal to 6–10 hours. Full-term neonates (gestational age ~40 weeks) also have a larger volume of distribution (mean V = 0.4–0.5 L/kg) and lower aminoglycoside clearance resulting in longer half-life values (t1/2 = 4–5 hours). By about 6 months, the mean volume of distribution is still large (V = 0.3–0.4 L/kg), but kidney development is complete, aminoglycoside clearance increases, and half-life is shorter (t1/2 = 2–3 hours). These values remain relatively constant until about 2 years of age. At that time, aminoglycoside volume of distribution, clearance, and half-life gradually approach adult values at puberty (~12–14 years old). Initial doses for neonates are based on birth weight and age:71

Doses for infants and children are: amikacin 15–22.5 mg/kg/d IV or IM given every 8 hours, gentamicin or tobramycin 7.5 mg/kg/d IV or IM given every 8 hours.72 Extended-interval aminoglycoside dosing can be conducted in pediatric patients.73 After initial doses are started, steady-state aminoglycoside serum concentrations are used to individualize doses for either conventional or extended-interval dosing.

Hemodialysis efficiently removes aminoglycoside antibiotics from the body.74–78 Gentamicin, tobramycin, netilmicin, and amikacin are relatively small molecules that are water soluble and have a small volume of distribution and low plasma protein binding. All of these characteristics lead to very good hemodialysis removal. The average aminoglycoside half-life in a renal failure patient is 50 hours. During hemodialysis with a “low-flux” artificial kidney, half-life decreases to 4 hours and results in about 50% of the drug being removed during a typical dialysis period (3–4 hours). Similarly, hemodialysis performed with a “high-flux” filter decreases aminoglycoside half-life to 2 hours.79 If the patient is properly hydrated, the volume of distribution for aminoglycosides is 0.26 L/kg. Hemodialysis procedures, such as ultrafiltration, can be used to assist in the maintenance of proper fluid balance in patients. Because kidneys provide fluid and electrolyte balance, it is not unusual for patients with renal failure receiving hemodialysis to be over- or underhydrated. As previously discussed in the renal failure section in the above paragraphs, body weight is an effective way to assess hydration status and can be used to adjust initial volume of distribution estimates.

Peritoneal dialysis is much less efficient in removing aminoglycosides from the body.80–82 Peritoneal dialysis will decrease the half-life of aminoglycosides in a renal failure patient from about 50 hours to about 36 hours during the dialysis procedure. If the patient is receiving peritoneal dialysis on a chronic, ongoing basis, such as continuous ambulatory peritoneal dialysis (CAPD), aminoglycoside half-life will be shorter all of the time because of the additional dialysis clearance. Patients receiving continuous ambulatory peritoneal dialysis sometimes develop peritonitis which can be treated by adding aminoglycoside (or other) antibiotics to the peritoneal dialysis fluid. While about one-half of the intraperitoneal aminoglycoside dose is systemically absorbed during a 5–6 hours dwell time, if a patient with peritonitis develops secondary bacteremia, it may be necessary to use parenteral antibiotics to cure the infection.80–82 Peritonitis causes inflammation of the peritoneal membrane, which facilitates absorption of aminoglycoside administered via dialysis fluid and elimination of aminoglycoside present in the body.

Continuous hemofiltration consists of a family of techniques that provides removal of toxic metabolic substances in patients with acute renal failure.83 A large amount of variability exists in aminoglycoside removal depending on the type of hemofiltration used in a patient. Average sieving coefficients for the aminoglycoside antibiotics are:84,85 gentamicin 0.81, tobramycin 0.90, netilmicin 0.93, amikacin 0.95. Because continuous arteriovenous hemofiltration (CAVH) provides an average creatinine clearance of ~30 mL/min, this value is typically used to initiate therapy in patients, then aminoglycoside serum concentration monitoring is used to individualize dosing early in therapy.86

Most important drug interactions are pharmacodynamic, and not pharmacokinetic, in nature. Vancomycin,14,17,87 amphotericin B,17 cyclosporin,88 and furosemide12,16,17 enhance the nephrotoxicity potential of the aminoglycosides. Each of these agents can cause nephrotoxicity when administered alone. When these drugs are administered concurrently with an aminoglycoside, serum creatinine concentrations should be monitored on a daily basis. Additionally, serum concentrations of vancomycin or cyclosporin, as well as the aminoglycoside, should be measured. Loop diuretics,89,90 including furosemide, bumetanide, and ethacrynic acid, can cause ototoxicity, and an increased incidence of this adverse effect has been reported when aminoglycosides have been coadministered. If aminoglycoside antibiotics are administered with loop diuretics, clinical signs and symptoms of ototoxicity (auditory: decreased hearing acuity in the conversational range, feeling of fullness or pressure in the ears, tinnitus; vestibular: loss of equilibrium, headache, nausea, vomiting, nystagmus, vertigo, ataxia) should be monitored daily.

Aminoglycosides have intrinsic nondepolarizing neuromuscular blocking activity and may prolong the effects of neuromuscular blocking agents such as succinylcholine.91 Surgical and intensive care patients receiving neuromuscular blockers and aminoglycosides should be monitored for this potential adverse effect. As previously discussed, penicillins (primarily penicillin G, ampicillin, nafcillin, carbenicillin, ticarcillin) can inactivate aminoglycosides in vivo and in blood specimen tubes intended for the measurement of aminoglycoside serum concentrations.50–54 These two classes of drugs can also inactive each other in intravenous administration bags and syringes and should not be mixed together.

Several methods to initiate aminoglycoside therapy are available. The pharmacokinetic dosing method is the most flexible of the techniques. It allows for individualized target serum concentrations to be chosen for a patient, so it can be used for both conventional and extended-interval dosing. Also, each pharmacokinetic parameter can be customized to reflect specific disease states and conditions present in the patient. However, it is computationally intensive. The Hull and Sarubbi nomogram uses the dosing concepts in the pharmacokinetic dosing method. But, in order to simplify calculations, it makes simplifying assumptions: target concentration ranges consistent with conventional dosing only, fixed volume of distribution parameter in the normal range, limited dosage interval selection (no longer than 24 hours). Thus, it should be used only in patients who only have renal dysfunction and/or obesity as complicating factors and only when conventional dosing is to be used. The Hartford nomogram has similar strengths and weaknesses when compared to the Hull and Sarubbi nomogram, but is designed for use when extended-interval dosing is desired. This nomogram also incorporates a method to adjust aminoglycoside doses based on serum concentration feedback. Literature-based recommended dosing is a commonly used method to prescribe initial doses of aminoglycosides to pediatric patients. Doses are based on those that commonly produce steady-state concentrations within the therapeutic range, although there is a wide variation in the actual concentrations for a specific patient.

Pharmacokinetic Dosing Method

The goal of initial dosing of aminoglycosides is to compute the best dose possible for the patient given their set of disease states and conditions that influence aminoglycoside pharmacokinetics and the site and severity of the infection. In order to do this, pharmacokinetic parameters for the patient will be estimated using average parameters measured in other patients with similar disease state and condition profiles.

Elimination Rate Constant Estimate

Aminoglycosides are almost totally eliminated unchanged in the urine, and there is a good relationship between creatinine clearance and aminoglycoside elimination rate constant (Figure 4-2). This relationship allows the estimation of the aminoglycoside elimination rate constant for a patient which can be used to compute an initial dose of the antibiotic. Mathematically, the equation for the straight line shown in Figure 4-2 is: ke = 0.00293(CrCl) + 0.014, where ke is the aminoglycoside elimination rate constant in h–1 and CrCl is creatinine clearance in mL/min. A limitation in using elimination rate constant as the elimination parameter in this relationship is that it is a hybrid pharmacokinetic constant whose value can be influenced by either clearance or volume of distribution (ke = Cl/V). Because gentamicin, tobramycin, netilmicin, and amikacin have similar pharmacokinetic properties, the same elimination rate constant versus creatinine clearance relationship can be used for all of the antibiotics. For example, the estimated elimination rate constant for an individual with a creatinine clearance of 10 mL/min is 0.043 h–1 which yields an estimated half-life of 16 hours [ke = 0.00293(CrCl) + 0.014 = 0.00293 · (10 mL/min) + 0.014 = 0.043 h–1; t1/2 = 0.693/(0.043 h–1) = 16 h]. Taking the patient’s renal function into account when deriving initial doses of aminoglycoside antibiotics is the single most important characteristic to assess.

Volume of Distribution Estimate

The average volume of distribution for patients without disease states and conditions that change this parameter is 0.26 L/kg. Thus, for a nonobese 70-kg patient, the estimated volume of distribution would be 18.2 L (V = 0.26 L/kg · 70 kg = 18.2 L). If a patient weighs less than their ideal body weight, actual body weight is used to estimate volume of distribution. For patients whose weight is between their ideal body weight and 30% over ideal weight, actual body weight can be used to compute estimated volume of distribution, although some clinicians prefer to use ideal body weight for these individuals. In patients who are more than 30% above their ideal body weight, volume of distribution (V) estimates should include both ideal and actual total body weighs using the following equation: V = 0.26[IBW + 0.4(TBW – IBW)], where V is in L, IBW is ideal body weight in kilograms, and TBW is total body weight in kilograms. For an obese patient whose ideal body weight is 55 kg and total body weight is 95 kg, the estimated volume of distribution would be 18.5 L: V = 0.26[IBW + 0.4(TBW – IBW)] = 0.26[55 kg + 0.4(95 kg – 55 kg)] = 18.5 L. In patients who are overhydrated or have ascites, their dry body weight (weight without the extra fluid) can be used to provide an improved volume of distribution estimate (V in L) using the following formula: V = (0.26 · DBW) + (TBW – DBW), where DBW is the patient’s dry body weight and TBW is the patient’s actual total body weight. For example, a patient with a significant amount of ascitic fluid currently weighs 80 kg. It is known from previous clinic visits and history that the patient usually weighs 70 kg without the additional fluid. The estimated volume of distribution for this patient would be 28.2 L: V = (0.26 · DBW) + (TBW – DBW) = (0.26 · 70 kg) + (80 kg – 70 kg) = 28.2 L. Other disease states and conditions also influence aminoglycoside volume of distribution, and the values of this parameter given in Table 4-1 will be used when necessary. For instance, the average volume of distribution for cystic fibrosis patients is 0.35 L/kg. Therefore, the estimated volume of distribution for a 55-kg patient with cystic fibrosis is 19.3 L: V = 0.35 L/kg (55 kg) = 19.3 L.

Selection of Appropriate Pharmacokinetic Model and Equations

When given by intravenous injection over less than 1 hour, aminoglycosides follow a three-compartment pharmacokinetic model (Figure 4-4). After the end of infusion, serum concentrations drop rapidly because of distribution of drug from blood to tissues (α or distribution phase). If aminoglycosides are infused over 1 hour, the distribution phase is not usually observed. By about 1 hour after the beginning of the antibiotic infusion, drug concentrations decline more slowly, and the elimination rate constant for this segment of the concentration/ time curve is the one that varies with renal function (β or elimination phase). Finally, at very low serum concentrations not detected by aminoglycoside concentration assays in clinical use (≤0.5 μg/mL), drug that was tissue-bound to various organs (especially the kidney) is released from tissue-binding sites and eliminated (γ or tissue-release phase). While this model was instrumental in advancing current ideas regarding aminoglycoside tissue accumulation and nephrotoxicity, it cannot easily be used clinically because of its mathematical complexity.9–11,13,16 Because of this, the simpler one-compartment model is widely used and allows accurate dosage calculation.2,3,27,45,46,48,49,92

Intravenously administered aminoglycosides are given over 1/2–1 hour as intermittent continuous infusions. Since drug is eliminated during the infusion time (and any waiting time that is necessary to allow for distribution to finish), pharmacokinetic equations that take into account this loss are preferred in patients with good renal function. If this is not done, a large amount of drug may be eliminated during infusion and waiting periods, and the peak concentration will be miscalculated. Generally, infusion equations should be used if the patient has a creatinine clearance greater than 30 mL/min. For creatinine clearances of 30 mL/min or less, very little aminoglycoside is eliminated during infusion and waiting period times, and intravenous bolus equations accurately compute peak concentrations.93 Aminoglycoside steady-state peak (Cssmax) and trough (Cssmin) serum concentrations are chosen to treat the patient based on the type, site, and severity of infection as well as the infecting organism. Steady-state versions of one-compartment model intermittent intravenous infusion {Cssmax = [k0/(keV)][(1 – e–ket′) / (1 – e–keτ)] Cminss = Cmaxss–[ke(τ–t′)], where k0 is the infusion rate, ke is the elimination rate constant, V is the volume of distribution, t´ is the drug infusion time, and τ is the dosage interval} or intravenous bolus {Cssmax = (D/V)[e–ket/ (1 – e–keτ)], Cssmax = Cssmax e–keτ, where D is the antibiotic dose, V is the volume of distribution, ke is the elimination rate constant, t is the time Cssmax was measured, and τ is the dosage interval} equations are chosen based on the patient’s renal function to compute the required doses needed to achieve desired aminoglycoside concentrations. Note that intermittent intravenous infusion equations will work well regardless of the patient’s creatinine clearance. However, the intravenous bolus equations are easier to solve, save time, and are less likely to invoke a computational error.

Steady-State Concentration Selection

Aminoglycoside peak steady-state concentrations are selected based on site and severity of infection as well as the infecting organism. Severe infections, such as gram-negative pneumonia or septicemia, or infections with organisms that have a high minimum inhibitory concentration (MIC) such as Pseudomonas aeruginosa (typical MIC ≈ 2 μg/mL for gentamicin, tobramycin, or netilmicin) generally require peak steady-state serum concentrations of 8–10 μg/mL for gentamicin, tobramycin, or netilmicin or 25–30 μg/mL for amikacin when using conventional dosing. Moderate infections at sites that are easier to penetrate or with organisms that display lower MIC values, such as intraabdominal infections are usually treated with peak gentamicin, tobramycin, or netilmicin steady-state serum concentrations equal to 5–7 μg/mL or with amikacin peak steady-state serum concentrations equal to 15–25 μg/mL. When treating urinary tract infections due to susceptible organisms or using aminoglycosides for synergy in combination with penicillins or other antibiotics for the treatment of gram-positive infections such as infective endocarditis, steady-state peak concentrations of 3–5 μg/mL are usually adequate for gentamicin, tobramycin, or netilmicin; or 12–15 μg/mL for amikacin. Pyelonephritis is considered a soft-tissue infection, not a urinary tract infection, and requires higher peak steady-state concentrations to achieve a cure. Similar target peak steady-state concentrations for extended-interval aminoglycoside dosing are less established, although concentrations 20–30 μg/mL have been suggested for Pseudomonas aeruginosa and other serious infections including pulmonary exacerbations in cystic fibrosis patients. Desirable concentrations for steady-state trough concentrations are chosen based on avoidance of potential toxicity. For conventional dosing, steady-state trough concentrations should be maintained <2 μg/mL for tobramycin, gentamicin, and netilmicin or <5–7 μg/mL for amikacin. Using extended-interval dosing, steady-state trough concentrations should be <1 μg/mL for gentamicin, tobramycin, and netilmicin.

Dosage Computation

The equations given in Tables 4-2A, 4-2B, and 4-2C are used to compute aminoglycoside doses. One approach is to use different equations depending upon the renal function of the patient (intermittent intravenous infusion for creatinine clearances >30 mL/min, intravenous bolus for creatinine clearances ≤30 mL/min). Alternatively, intermittent intravenous infusion equations can be used for all patients regardless of renal function.

Example 1

JM is a 50-year-old, 70-kg (5 ft 10 in) male with gram-negative pneumonia. His current serum creatinine is 0.9 mg/dL, and it has been stable over the last 5 days since admission. Compute a gentamicin dose for this patient using conventional dosing.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• 3.Estimate volume of distribution (V).
• The patient has no disease states or conditions that would alter the volume of distribution from the normal value of 0.26 L/kg:

• 4.Choose desired steady-state serum concentrations.
• Gram-negative pneumonia patients treated with aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 8–10 μg/mL; steady-state trough (Cssmin) concentrations should be <2 μg/mL to avoid toxicity. Set Cssmax = 9 μg/mL and Cssmin = 1 μg/mL.
• 5.Use intermittent intravenous infusion equations to compute dose (Tables 4-2A, 4-2B, and 4-2C).
• Calculate required dosage interval (τ) using a 1-hour infusion:

• Dosage intervals should be rounded to clinically acceptable intervals of 8 hours, 12 hours, 18 hours, 24 hours, 36 hours, 48 hours, 72 hours, and multiples of 24 hours thereafter, whenever possible. In this case, the dosage interval would be rounded to 8 hours. Also, steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.

• Aminoglycoside doses should be rounded to the nearest 5–10 mg. This dose would be rounded to 170 mg. (Note: μg/mL = mg/L, and this concentration unit was substituted for Cssmax so that unnecessary unit conversion was not required.)
• The prescribed maintenance dose would be 170 mg every 8 hours.
• 6. Compute loading dose (LD), if needed.
• Loading doses should be considered for patients with creatinine clearance values below 60 mL/min. The administration of a loading dose in these patients will allow achievement of therapeutic peak concentrations quicker than if maintenance doses alone are given. However, since the pharmacokinetic parameters used to compute these initial doses are only estimated values and not actual values, the patient’s own parameters may be much different than the estimated constants and steady state will not be achieved until 3–5 half-lives have passed.

• As noted, this loading dose is only about 10% greater than the maintenance dose and wouldn’t be given to the patient. Since the expected half-life is 2.3 hours, the patient should be at steady state after the second dose is given.

Example 2

Same patient profile as in example 1, but serum creatinine is 3.5 mg/dL indicating renal impairment.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• 3.Estimate volume of distribution (V).
• The patient has no disease states or conditions that would alter the volume of distribution from the normal value of 0.26 L/kg:

• 4.Choose desired steady-state serum concentrations.
• Gram-negative pneumonia patients treated with aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 8–10 μg/mL; steady-state trough (Cssmin) concentrations should be <2 μg/mL to avoid toxicity. Set Cssmax = 9 μg/mL and Cssmin = 1 μg/mL.
• 5.Use intravenous bolus equations to compute dose (Tables 4-2A, 4-2B, and 4-2C).
• Calculate required dosage interval (τ):

• Dosage intervals should be rounded to clinically acceptable intervals of 8 hours, 12 hours, 18 hours, 24 hours, 36 hours, 48 hours, 72 hours, and multiples of 24 hours thereafter, whenever possible. In this case, the dosage interval would be rounded to 24 hours. Also, steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose (D) could be administered either way.

• Aminoglycoside doses should be rounded to the nearest 5–10 mg. This dose would be rounded to 145 mg. (Note: μg/mL = mg/L and this concentration unit was substituted for Cssmax so that unnecessary unit conversion was not required.)
• The prescribed maintenance dose would be 145 mg every 24 hours.
• Note: Although this dose is given once daily, it is not extended-interval dosing because desired serum concentrations are within the conventional range.
• 6. Compute loading dose (LD), if needed.
• Loading doses should be considered for patients with creatinine clearance values below 60 mL/min. The administration of a loading dose in these patients will allow achievement of therapeutic peak concentrations quicker than if maintenance doses alone are given. However, since the pharmacokinetic parameters used to compute these initial doses are only estimated values and not actual values, the patient’s own parameters may be much different from the estimated constants and steady state will not be achieved until 3–5 half-lives have passed.

• Round loading dose to 165 mg. It would be given as the first dose. The next dose would be a maintenance dose given a dosage interval away from the loading dose, in this case 24 hours later.

Example 3

ZW is a 35-year-old, 150-kg (5 ft 5 in) female with an intraabdominal infection. Her current serum creatinine is 1.1 mg/dL and is stable. Compute a tobramycin dose for this patient using conventional dosing.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is obese [IBWfemales (in kg) = 45 + 2.3(Ht – 60 in) = 45 + 2.3(65 – 60) = 57 kg]. The Salazar and Corcoran equation can be used to estimate creatinine clearance:

• Note: Height is converted from inches to meters: Ht = (65 in · 2.54 cm/in) / (100 cm/m) = 1.65 m.
• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• 3.Estimate volume of distribution (V).
• The patient is obese, so the volume of distribution would be estimated using the following formula:

• 4.Choose desired steady-state serum concentrations.
• Intraabdominal infection patients treated with aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 5–7 μg/mL; steady-state trough (Cssmin) concentrations should be <2 μg/mL to avoid toxicity. Set Cssmax = 6 μg/mL and Cssmin = 0.5 μg/mL.
• 5.Use intermittent intravenous infusion equations to compute dose (Tables 4-2A, 4-2B, and 4-2C).
• Calculate required dosage interval (τ) using a 1-hour infusion:

• Dosage intervals should be rounded to clinically acceptable intervals of 8 hours, 12 hours, 18 hours, 24 hours, 36 hours, 48 hours, 72 hours, and multiples of 24 hours thereafter, whenever possible. In this case, the dosage interval is 8 hours. Also, steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.

• Aminoglycoside doses should be rounded to the nearest 5–10 mg. This dose does not need to be rounded. (Note: μg/mL = mg/L and this concentration unit was substituted for Cssmax so that unnecessary unit conversion was not required.)
• The prescribed maintenance dose would be 165 mg every 8 hours.
• 6. Compute loading dose (LD), if needed.
• Loading doses should be considered for patients with creatinine clearance values below 60 mL/min. The administration of a loading dose in these patients will allow achievement of therapeutic peak concentrations quicker than if maintenance doses alone are given. However, since the pharmacokinetic parameters used to compute these initial doses are only estimated values and not actual values, the patient’s own parameters may be much different than the estimated constants and steady state will not be achieved until 3–5 half-lives have passed.

• As noted, this loading dose is about 10% greater than the maintenance dose and wouldn’t be given to the patient. Since the expected half-life is 1.9 hours, the patient should be at steady state after the second dose is given.

Example 4

JM is a 20-year-old, 76-kg (height = 5 ft 8 in) male with a gram-negative pneumonia. His current serum creatinine is 1.1 mg/dL and is stable. Compute a tobramycin dose for this patient using extended-interval dosing.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese {IBWmales = 50 + 2.3(Ht – 60 in) = 50 + 2.3(68 – 60) = 68 kg; % overweight = [100(76 kg – 68 kg)] / 68kg = 12%}. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• 3.Estimate volume of distribution (V).
• The patient has no disease states or conditions that would alter the volume of distribution from the normal value of 0.26 L/kg:

• 4.Choose desired steady-state serum concentrations.
• Gram-negative pneumonia patients treated with extended-interval aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 20–30 μg/mL; steady-state trough (Cssmin) concentrations should be <1 μg/mL to avoid toxicity. Set Cssmax = 30 μg/mL and Cssmin = 0.1 μg/mL.
• 5.Use intermittent intravenous infusion equations to compute dose (Tables 4-2A, 4-2B, and 4-2C).
• Calculate required dosage interval (τ) using a 1-hour infusion:

• Dosage intervals for extended-interval dosing should be rounded to clinically acceptable intervals of 24 hours, 36 hours, 48 hours, 60 hours, 72 hours, and multiples of 24 hours thereafter, whenever possible. Some clinicians prefer to avoid the use of extended-interval dosing beyond a dosage interval of 48 hours because serum concentrations can be below the MIC far beyond the time frame afforded by the postantibiotic effect. For these situations, they revert to conventional dosing for the patient. In this case, the patient’s dosage interval will be rounded to 24 hours. Because of this, the steady-state trough concentration would be expected to fall below 0.1 μg/mL. Also, steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.

• Aminoglycoside doses should be rounded to the nearest 10–50 mg for extended-interval dosing. This dose would be rounded to 700 mg. (Note: μg/mL = mg/L and this concentration unit was substituted for Cssmax so that unnecessary unit conversion was not required.)
• The prescribed maintenance dose would be 700 mg every 24 hours.
• 6. Compute loading dose (LD), if needed.
• Loading doses should be considered for patients with creatinine clearance values below 60 mL/min. The administration of a loading dose in these patients will allow achievement of therapeutic peak concentrations quicker than if maintenance doses alone are given. However, since the pharmacokinetic parameters used to compute these initial doses are only estimated values and not actual values, the patient’s own parameters may be much different from the estimated constants and steady state will not be achieved until 3–5 half-lives have passed.

• As noted, this loading dose is about 10% greater than the maintenance dose and wouldn’t be given to the patient. Since the expected half-life is 2 hours, the patient should be at steady state after the first dose is given.

Example 5

JM is an 80-year-old, 80-kg (5 ft 8 in) male with Streptococcus viridans endocarditis. His current serum creatinine is 1.5 mg/dL, and it has been stable. Ampicillin and gentamicin will be used to treat the infection. Compute a gentamicin dose for this patient using conventional dosing.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese {IBWmales = 50 + 2.3(Ht – 60 in) = 50 + 2.3(68 – 60) = 68 kg; % overweight = [100(80 kg – 68 kg)] / 68 kg = 18%}. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• 3.Estimate volume of distribution (V).
• The patient has no disease states or conditions that would alter the volume of distribution from the normal value of 0.26 L/kg:

• 4.Choose desired steady-state serum concentrations.
• S. viridans endocarditis patients treated with aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 3–5 μg/mL; steady-state trough (Cssmin) concentrations should be <2 μg/mL to avoid toxicity. Set Cssmax = 4 μg/mL and Cssmin = 1 μg/mL.
• 5.Use intermittent intravenous infusion equations to compute dose (Tables 4-2A, 4-2B, and 4-2C).
• Calculate required dosage interval (τ) using a 1-hour infusion:

• Dosage intervals should be rounded to clinically acceptable intervals of 8 hours, 12 hours, 18 hours, 24 hours, 36 hours, 48 hours, 72 hours, and multiples of 24 hours thereafter, whenever possible. In this case, the dosage interval would be rounded to 12 hours. Also, steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.

• Aminoglycoside doses should be rounded to the nearest 5–10 mg. This dose would be rounded to 70 mg. (Note: μg/mL = mg/L and this concentration unit was substituted for Cssmax so that unnecessary unit conversion was not required.)
• The prescribed maintenance dose would be 70 mg every 12 hours.
• Because the patient is receiving concurrent treatment with ampicillin, care would be taken to avoid in vitro inactivation in blood sample tubes intended for the determination of aminoglycoside serum concentrations.
• 6. Compute loading dose (LD), if needed.
• Loading doses should be considered for patients with creatinine clearance values below 60 mL/min. The administration of a loading dose in these patients will allow achievement of therapeutic peak concentrations quicker than if maintenance doses alone are given. However, since the pharmacokinetic parameters used to compute these initial doses are only estimated values and not actual values, the patient’s own parameters may be much different from the estimated constants and steady state will not be achieved until 3–5 half-lives have passed.

• The loading dose would be given as the first dose. The next dose would be a maintenance dose given a dosage interval away from the loading dose, in this case 12 hours later.

Example 6

Same patient profile as in example 2, but extended-interval dosing is used.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• 3.Estimate volume of distribution (V).
• The patient has no disease states or conditions that would alter the volume of distribution from the normal value of 0.26 L/kg:

• 4.Choose desired steady-state serum concentrations.
• Gram-negative pneumonia patients treated with aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) >20 μg/mL; steady-state trough (Cssmin) concentrations should be <1 μg/mL to avoid toxicity. Set Cssmax = 20 μg/mL and Cssmin = 0.5 μg/mL.
• 5.Use intermittent intravenous infusion equations to compute dose (Tables 4-2A, 4-2B, and 4-2C).
• Calculate required dosage interval (τ):

• Dosage intervals for extended-interval dosing should be rounded to clinically acceptable intervals of 24 hours, 36 hours, 48 hours, 60 hours, 72 hours, and multiples of 24 hours thereafter, whenever possible. Some clinicians prefer to avoid the use of extended-interval dosing beyond a dosage interval of 48 hours because serum concentrations can be below the MIC far beyond the time frame afforded by the postantibiotic effect. For these situations, they revert to conventional dosing for the patient. In this case, the dosage interval would be rounded to 48 hours. Also, steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose (D) could be administered either way.

• For extended-interval dosing, aminoglycoside doses should be rounded to the nearest 10–50 mg. This dose would be rounded to 350 mg. (Note: μg/mL = mg/L and this concentration unit was substituted for Cssmax so that unnecessary unit conversion was not required.)
• The prescribed maintenance dose would be 350 mg every 48 hours.
• 6. Compute loading dose (LD), if needed.
• Loading doses should be considered for patients with creatinine clearance values below 60 mL/min. The administration of a loading dose in these patients will allow achievement of therapeutic peak concentrations quicker than if maintenance doses alone are given. However, since the pharmacokinetic parameters used to compute these initial doses are only estimated values and not actual values, the patient’s own parameters may be much different from the estimated constants and steady state will not be achieved until 3–5 half-lives have passed.

• As noted, this loading dose is about 10% greater than the maintenance dose and wouldn’t be given to the patient. Since the expected half-life is 8 hours, the patient should be at steady state after the first dose is given.

Example 7

DQ is a 20-year-old, 61-kg (height = 5 ft 8 in) male with a pulmonary exacerbation due to cystic fibrosis. His current serum creatinine is 0.7 mg/dL and is stable. Compute a tobramycin dose for this patient using extended-interval dosing.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• 3.Estimate volume of distribution (V).
• The patient has cystic fibrosis, so the volume of distribution equals 0.35 L/kg:

• 4.Choose desired steady-state serum concentrations.
• Cystic fibrosis patients treated with extended-interval aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 20–30 μg/mL; steady-state trough (Cssmin) concentrations should be <1 μg/mL to avoid toxicity. Set Cssmax = 30 μg/mL and Cssmin = 0.01 μg/mL.
• 5.Use intermittent intravenous infusion equations to compute dose (Tables 4-2A, 4-2B, and 4-2C).
• Calculate required dosage interval (τ) using a 1-hour infusion:

• Dosage intervals for extended-interval dosing should be rounded to clinically acceptable intervals of 24 hours, 36 hours, 48 hours, 60 hours, 72 hours, and multiples of 24 hours thereafter, whenever possible. Some clinicians prefer to avoid the use of extended-interval dosing beyond a dosage interval of 48 hours because serum concentrations can be below the MIC far beyond the time frame afforded by the postantibiotic effect. For these situations, they revert to conventional dosing for the patient. In this case, the patient’s dosage interval will be rounded to 24 hours. Because of this, the steady-state trough concentration would be expected to fall below 0.01 μg/mL. Also, steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.

• Aminoglycoside doses should be rounded to the nearest 10–50 mg for extended-interval dosing. This dose would be rounded to 800 mg. (Note: μg/mL = mg/L and this concentration unit was substituted for Cssmax so that unnecessary unit conversion was not required.)
• The prescribed maintenance dose would be 800 mg every 24 hours.
• 6. Compute loading dose (LD), if needed.
• Loading doses should be considered for patients with creatinine clearance values below 60 mL/min. The administration of a loading dose in these patients will allow achievement of therapeutic peak concentrations quicker than if maintenance doses alone are given. However, since the pharmacokinetic parameters used to compute these initial doses are only estimated values and not actual values, the patient’s own parameters may be much different than the estimated constants and steady state will not be achieved until 3–5 half-lives have passed.

• As noted, this loading dose is about 10% greater than the maintenance dose and wouldn’t be given to the patient. Since the expected half-life is 1.6 hours, the patient should be at steady state after the first dose is given.

Hull and Sarubbi Nomogram Method

For patients who do not have disease states or conditions that alter volume of distribution, the only two patient-specific factors that change when using the pharmacokinetic dosing method is patient weight and creatinine clearance. Because of this, it is possible to make a simple nomogram to handle uncomplicated patients with a standard volume of distribution (Table 4-3). The Hull and Sarubbi aminoglycoside dosing nomogram is a quick and efficient way to apply pharmacokinetic dosing concepts without using complicated pharmacokinetic equations.45,46 With a simple modification, it can also be used for obese patients. If the patient is ≥30% above ideal body weight, an adjusted body weight (ABW) can be calculated and used as the weight factor [ABW (in kg) = IBW + 0.4(TBW – IBW), where IBW is ideal body weight in kilograms and TBW is actual total body weight in kilograms].55–57 As can be seen, this equation is derived from the computation for volume of distribution in obese patients. Also, the Salazar and Corcoran method of estimating creatinine clearance in obese patients should be used to compute renal function in these individuals.94–97

Steady-state peak concentrations are selected as discussed in the pharmacokinetic dosing method section and used to determine a loading dose from the nomogram (Table 4-3). Logically, lower loading doses produce lower expected peak concentrations, and higher loading doses result in higher expected peak concentrations. Once the loading dose is found the patient’s creatinine clearance is used to estimate the half-life, dosage interval, and maintenance dose (as a percent of the administered loading dose). The maintenance dose supplied by the nomogram is the percent of the loading dose that was eliminated during the different dosage interval time frames, and will, therefore, provide the same estimated peak concentration at steady state as that supplied by the loading dose. To illustrate how the nomogram is used, the same conventional-dosing patient examples utilized in the previous section will be repeated for this dosage approach using the same example number. Since the nomogram uses slightly different estimates for volume of distribution and elimination rate constant, some minor differences in suggested doses are expected. Because the cystic fibrosis example requires a different volume of distribution (0.35 L/kg), the Hull and Sarubbi nomogram cannot be used.

Example 1

JM is a 50-year-old, 70-kg (5 ft 10 in) male with gram-negative pneumonia. His current serum creatinine is 0.9 mg/dL, and it has been stable over the last 5 days since admission. Compute a gentamicin dose for this patient using conventional dosing.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2.Choose desired steady-state serum concentrations.
• Gram-negative pneumonia patients treated with aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 8–10 μg/mL.
• 3. Select loading dose (Table 4-3).
• A loading dose (LD) of 2 mg/kg will provide a peak concentration of 8–10 μg/mL.

• 4. Determine estimated half-life, maintenance dose, and dosage interval.
• From the nomogram the estimated half-life is 2–3 hours, the maintenance dose (MD) is 90% of the loading dose [MD = 0.90(140 mg) = 126 mg], and the dosage interval is 8 hours.
• Aminoglycoside doses should be rounded to the nearest 5–10 mg. Steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.
• The prescribed maintenance dose would be 125 mg every 8 hours.

Example 2

Same patient profile as in example 1, but serum creatinine is 3.5 mg/dL indicating renal impairment.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Choose desired steady-state serum concentrations.
• Gram-negative pneumonia patients treated with aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 8–10 μg/mL.
• 3. Select loading dose (Table 4-3).
• A loading dose (LD) of 2 mg/kg will provide a peak concentration of 8–10 μg/mL.

• 4. Determine estimated half-life, maintenance dose, and dosage interval.
• From the nomogram the estimated half-life is 9.9 hours, the maintenance dose (MD) is 81% of the loading dose [MD = 0.81(140 mg) = 113 mg], and the dosage interval is 24 hours. Note: Because of the Cmaxss and Cminss chosen for this patient, the 24-hour dosage interval was used.
• Aminoglycoside doses should be rounded to the nearest 5–10 mg. Steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.
• The prescribed maintenance dose would be 115 mg every 24 hours.

Example 3

ZW is a 35-year-old, 150-kg (5 ft 5 in) female with an intraabdominal infection. Her current serum creatinine is 1.1 mg/dL and is stable. Compute a tobramycin dose for this patient using conventional dosing.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is obese [IBWfemales (in kg) = 45 + 2.3(Ht – 60 in) = 45 + 2.3(65 – 60) = 57 kg]. The Salazar and Corcoran equation can be used to estimate creatinine clearance:

• Note: Height is converted from inches to meters: Ht = (65 in · 2.54 cm/in) / (100 cm/m) = 1.65 m.
• 2. Choose desired steady-state serum concentrations.
• Intraabdominal infection patients treated with aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 5–7 μg/mL.
• 3. Select loading dose (Table 4-3).
• A loading dose (LD) of 1.7 mg/kg will provide a peak concentration of 5–7 μg/mL.
• Because the patient is obese, adjusted body weight (ABW) will be used to compute the dose:

• 4. Determine estimated half-life, maintenance dose, and dosage interval.
• From the nomogram the estimated half-life is 2–3 hours, the maintenance dose (MD) is 90% of the loading dose [MD = 0.90(160 mg) = 144 mg], and the dosage interval is 8 hours.
• Aminoglycoside doses should be rounded to the nearest 5–10 mg. Steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.
• The prescribed maintenance dose would be 145 mg every 8 hours.

Example 5

JM is an 80-year-old, 80-kg (5 ft 8 in) male with S. viridans endocarditis. His current serum creatinine is 1.5 mg/dL, and it has been stable. Ampicillin and gentamicin will be used to treat the infection. Compute a gentamicin dose for this patient using conventional dosing.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese {IBWmales = 50 + 2.3(Ht – 60 in) = 50 + 2.3(68 – 60) = 68 kg; % overweight = [100(80 kg – 68 kg)] / 68 kg = 18%}. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2.Choose desired steady-state serum concentrations.
• S. viridans endocarditis patients treated with aminoglycoside antibiotics require steady- state peak concentrations (Cssmax) equal to 3–5 μg/mL.
• 3. Select loading dose (Table 4-3).
• A loading dose (LD) of 1.5 mg/kg will provide a peak concentration of 5–7 μg/mL. This is the lowest dose suggested by the nomogram and will be used in this example. However, some clinicians may substitute a loading dose of 1–1.2 mg/kg designed to produce a steady-state peak concentration equal to 3–4 μg/mL.

• 4. Determine estimated half-life, maintenance dose, and dosage interval.
• From the nomogram the estimated half-life is 6.5 hours, suggesting that a 12-dosage interval is appropriate. The maintenance dose (MD) is 72% of the loading dose [MD = 0.72(120 mg) = 86 mg or MD = 0.72(95 mg) = 68 mg], and the dosage interval is 12 hours.
• Aminoglycoside doses should be rounded to the nearest 5–10 mg. Steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.
• The prescribed maintenance dose would be 85 mg every 12 hours or 70 mg every 12 hours, depending on the loading dose chosen.
• Because the patient is receiving concurrent treatment with ampicillin, care would be taken to avoid in vitro inactivation in blood sample tubes intended for the determination of aminoglycoside serum concentrations.

Hartford Nomogram Method for Extended-Interval Dosing

Extended-interval dosing is now a mainstream method used to administer aminoglycoside antibiotics. Conventional dosing is still preferred for endocarditis patients because the aminoglycoside is usually used for antibiotic synergy. Extended-interval doses obtained from the literature for patients with normal renal function are 4–7 mg/kg/d for gentamicin, tobramycin, or netilmicin and 11–20 mg/kg/d for amikacin.3,19–26,33–38 The most widely used extended-interval aminoglycoside dosage nomogram for patients with renal dysfunction is the Hartford nomogram which uses a 7-mg/kg dose (Table 4-4).3 Because the nomogram is essentially the concentration-time graph for gentamicin after a single dose of 7 mg/kg, it cannot be used for other dosage rates. The initial dose is 7 mg/kg of gentamicin (although it has not been tested with netilmicin, because of the pharmacokinetic similarity among the antibiotics it should be possible to use this aminoglycoside as well). The dosage interval is set according to the patient’s creatinine clearance (Table 4-4).

The Hartford nomogram includes a method to adjust doses based on gentamicin serum concentrations. This portion of the nomogram contains average serum concentration/time lines for gentamicin in patients with creatinine clearances of 60 mL/min, 40 mL/min, and 20 mL/min. A gentamicin serum concentration is measured 6–14 hours after the first dose is given, and this concentration/time point is plotted on the graph (Table 4-4). The suggested dosage interval is indicated by which zone the serum concentration/time point falls in. To illustrate how the nomogram is used, the same patient examples utilized in the pharmacokinetic dosing section will be repeated for this dosage approach using the same example number. Because the cystic fibrosis example requires a different volume of distribution (0.35 L/kg) and extended-interval dosing has not been adequately tested in patients with endocarditis, the Hartford nomogram should not be used in these situations.

Example 1

JM is a 50-year-old, 70-kg (5 ft 10 in) male with gram-negative pneumonia. His current serum creatinine is 0.9 mg/dL, and it has been stable over the last 5 days since admission. Compute a gentamicin dose for this patient using extended-interval dosing.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Compute initial dose and dosage interval (Table 4-4).
• A dose (D) of 7 mg/kg will provide a peak concentration >20 μg/mL.

• Dosage interval would be 24 hours using the nomogram. Extended-interval aminoglycoside doses should be rounded to the nearest 10–50 mg.
• The prescribed maintenance dose would be 500 mg every 24 hours.
• 3. Determine dosage interval using serum concentration monitoring.
• A gentamicin serum concentration measured 10 hours after the dose equals 3 μg/mL. Based on the nomogram, a dosage interval of 24 hours is the correct value and does not need to be altered.

Example 2

Same patient profile as in example 1, but serum creatinine is 3.5 mg/dL indicating renal impairment.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Compute initial dose and dosage interval (Table 4-4).
• A dose (D) of 7 mg/kg will provide a peak concentration >20 μg/mL.

• Dosage interval would be 48 hours using the nomogram. Extended-interval aminoglycoside doses should be rounded to the nearest 10–50 mg.
• The prescribed maintenance dose would be 500 mg every 48 hours.
• 3. Determine dosage interval using serum concentration monitoring.
• A gentamicin serum concentration measured 13 hours after the dose equals 9 μg/mL. Based on the nomogram, a dosage interval of 48 hours is too short and serial concentrations should be monitored. When the gentamicin serum concentration is <1 μg/mL, the next dose can be given. Based on the patient’s estimated elimination rate constant [ke = 0.00293(CrCl) + 0.014 = 0.00293(25 mL/min) + 0.014 = 0.087 h–1; t1/2 = 0.693/ke = 0.693 / 0.087 h–1 = 8 h], it will take approximately 3–4 half-lives or about an additional 24–32 hours after the gentamicin serum concentration for the value to drop below 1 μg/mL.
• Some clinicians prefer to avoid the use of extended-interval dosing beyond a dosage interval of 48 hours because serum aminoglycoside concentrations can be below the MIC far beyond the time frame afforded by the postantibiotic effect. For these situations, they revert to conventional dosing for the patient.

Example 3

ZW is a 35-year-old, 150-kg (5 ft 5 in) female with an intraabdominal infection. Her current serum creatinine is 1.1 mg/dL and is stable. Compute a tobramycin dose for this patient using extended-interval dosing.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is obese [IBWfemales (in kg) = 45 + 2.3(Ht – 60 in) = 45 + 2.3(65 – 60) = 57 kg]. The Salazar and Corcoran equation can be used to estimate creatinine clearance:

• Note: Height is converted from inches to meters: Ht = (65 in · 2.54 cm/in) / (100 cm/m) = 1.65 m.
• 2. Compute initial dose and dosage interval (Table 4-4).
• A dose (D) of 7 mg/kg will provide a peak concentration >20 μg/mL. Because the patient is obese, adjusted body weight (ABW) will be used to compute the dose: ABW = IBW + 0.4(TBW – IBW) = 57 kg + 0.4(150 kg – 57 kg) = 94 kg.

• Dosage interval would be 24 hours using the nomogram. Extended-interval aminoglycoside doses should be rounded to the nearest 10–50 mg.
• The prescribed maintenance dose would be 650 mg every 24 hours.
• 3. Determine dosage interval using serum concentration monitoring.
• A gentamicin serum concentration measured 8 hours after the dose equals 4 μg/mL. Based on the nomogram, a dosage interval of 24 hours is the correct value and does not need to be altered.
• Assuming linear pharmacokinetics, clinicians have begun to use the Hartford nomogram for doses other than 7 mg/kg. Because this approach has not been formally evaluated, extreme care should be exercised when using this approach. For example, if the clinical situation warrants it, a dose of 5 mg/kg could be administered to a patient, the initial dosage intervals suggested in the Hartford nomogram used, and a serum concentration measured to confirm the dosage interval. Assuming linear pharmacokinetics, the critical concentrations for changing dosage intervals on the Hartford nomogram graph would be decreased to 5/7 (the ratio of the 5 mg/kg dose administered to the 7 mg/kg dose suggested by the nomogram). Additionally, a similar nomogram for gentamicin or tobramycin doses of 5 mg/kg is also available.98,99

Literature-Based Recommended Dosing

Because of the large amount of variability in aminoglycoside pharmacokinetics, even when concurrent disease states and conditions are identified, many clinicians believe that the use of standard aminoglycoside doses for pediatric patients is warranted. The original computation of these doses was based on the pharmacokinetic dosing methods described in the previous section, and subsequently modified based on clinical experience. In general, the expected aminoglycoside steady-state serum concentrations used to compute these doses were similar to those for adults given conventional dosing. Suggested initial aminoglycoside doses for various pediatric patients are listed in the Effects of Disease States and Conditions on Aminoglycoside Pharmacokinetics and Dosing section. Doses for neonates that are below 10 mg are usually rounded to the nearest tenth of a milligram. If serum creatinine values are available, estimated creatinine clearance can be computed using equations that are specific for pediatric patients [age 0–1 year, CrClest (in mL/min/ 1.73 m2) = (0.45 · Ht) / SCr; age 1–20 years, CrClest (in mL/min/1.73 m2) = (0.55 · Ht) / SCr, where Ht is in cm and SCr is in mg/dL].100

Example 1

MM is a 3-day-old, 1015-g male with suspected neonatal sepsis. His serum creatinine has not been measured, but it is assumed that it is typical for his age and weight. Compute an initial gentamicin dose for this patient.

• 1. Compute initial dose and dosage interval.
• Often, serum creatinine measurements are not available for initial dosage computation in neonates. The dosage recommendations for this population assume typical renal function, so it is important to verify that the assumption is valid.
• From the pediatrics dosage recommendations given in earlier in the chapter, a patient in this age and weight category should receive gentamicin 2.5 mg/kg every 18–24 hours. Because the patient is in the lower end of the age range, it is likely he has lower renal function due to poor organ maturation. Based on this information, the longer dosage interval will be chosen. (Note: Grams will be converted to kilograms before the computation is made.)

• The prescribed dose would be 2.5 mg every 24 hours.

Because of pharmacokinetic variability among patients, it is likely that doses computed using patient population characteristics will not always produce aminoglycoside serum concentrations that are expected. Because of this, aminoglycoside serum concentrations are measured in many patients to ensure that therapeutic, nontoxic levels are present. However, not all patients may require serum concentration monitoring. For example, if it is expected that only a limited number of doses will be administered as is the case for surgical prophylaxis or an appropriate dose for the renal function and concurrent disease states of the patient is prescribed (e.g., 1 mg/kg every 8 hours for 3–5 days in a patient with a creatinine clearance of 80–120 mL/min for antibiotic synergy in the treatment of methicillin-sensitive Staphylococcus aureus aortic or mitral valve endocarditis), aminoglycoside serum concentration monitoring may not be necessary. Whether or not aminoglycoside concentrations are measured, important patient parameters (fever curves, white blood cell counts, serum creatinine concentrations, etc.) should be followed to confirm that the patient is responding to treatment and not developing adverse drug reactions.

When aminoglycoside serum concentrations are measured in patients and a dosage change is necessary, clinicians should seek to use the simplest, most straightforward method available to determine a dose that will provide safe and effective treatment. In most cases, a simple dosage ratio can be used to change aminoglycoside doses since these antibiotics follow linear pharmacokinetics. Sometimes, it is not possible to simply change the dose, and the dosage interval must also be changed to achieve desired serum concentrations. In this case, it may be possible to use pharmacokinetic concepts to alter the aminoglycoside dose that the patient needs. In some situations, it may be necessary to compute the aminoglycoside pharmacokinetic parameters for the patient using the Sawchuk-Zaske method and utilize these to calculate the best drug dose. Some clinicians advocate using individualized area under the concentration-time curve determinations to individualize aminoglycoside doses. Finally, computerized methods that incorporate expected population pharmacokinetic characteristics (Bayesian pharmacokinetic computer programs) can be used in difficult cases where renal function is changing, serum concentrations are obtained at suboptimal times, or the patient was not at steady state when serum concentrations were measured.

Linear Pharmacokinetics Method

Because aminoglycoside antibiotics follow linear, dose-proportional pharmacokinetics, steady-state serum concentrations change in proportion to dose according to the following equation: Dnew / Css,new = Dold / Css,old or Dnew = (Css,new / Css,old)Dold, where D is the dose, Css is the steady-state peak or trough concentration, old indicates the dose that produced the steady-state concentration that the patient is currently receiving, and new denotes the dose necessary to produce the desired steady-state concentration. The advantages of this method are that it is quick and simple. The disadvantages are steady-state concentrations are required, and it may not be possible to attain desired serum concentrations by only changing the dose.

Example 1

JM is a 50-year-old, 70-kg (5 ft 10 in) male with gram-negative pneumonia. His current serum creatinine is 0.9 mg/dL, and it has been stable over the last 5 days since admission. A gentamicin dose of 170 mg every 8 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 9 μg/mL and 1 μg/mL, respectively. After the third dose, steady-state peak and trough concentrations were measured and were 12 μg/mL and 1.4 μg/mL, respectively. Calculate a new gentamicin dose that would provide a steady-state peak of 9 μg/mL.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• Because the patient has been receiving gentamicin for more that 3–5 estimated half-lives, it is likely that the measured serum concentrations are steady-state values.
• 3. Compute new dose to achieve desired serum concentration.
• Using linear pharmacokinetics, the new dose to attain the desired concentration should be proportional to the old dose that produced the measured concentration:

• The new suggested dose would be 130 mg every 8 hours to be started at next scheduled dosing time.
• 4. Check steady-state trough concentration for new dosage regimen.
• Using linear pharmacokinetics, the new steady-state concentration can be estimated and should be proportional to the old dose that produced the measured concentration:

• This steady-state trough concentration should be safe and effective for the infection that is being treated.

Example 2

ZW is a 35-year-old, 150-kg (5 ft 5 in) female with an intraabdominal infection. Her current serum creatinine is 1.1 mg/dL and is stable. A tobramycin dose of 165 mg every 8 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 6 μg/mL and 0.5 μg/mL, respectively. After the fifth dose, steady-state peak and trough concentrations were measured and were 4 μg/mL and <0.5 μg/mL (e.g., below assay limits), respectively. Calculate a new tobramycin dose that would provide a steady-state peak of 6 μg/mL.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is obese [IBWfemales (in kg) = 45 + 2.3(Ht – 60) = 45 + 2.3(65 in – 60) = 57 kg]. The Salazar and Corcoran equation can be used to estimate creatinine clearance:

• Note: Height is converted from inches to meters: Ht = (65 in · 2.54 cm/in) / (100 cm/m) = 1.65 m.
• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• Because the patient has been receiving tobramycin for more that 3–5 estimated half-lives, it is likely that the measured serum concentrations are steady-state values.
• 3. Compute new dose to achieve desired serum concentration.
• Using linear pharmacokinetics, the new dose to attain the desired concentration should be proportional to the old dose that produced the measured concentration:

• The new suggested dose would be 250 mg every 8 hours to be started at next scheduled dosing time.
• 4. Check steady-state trough concentration for new dosage regimen.
• Using linear pharmacokinetics, the new steady-state concentration can be estimated and should be proportional to the old dose that produced the measured concentration. However, in this situation the trough concentration is below assay limits and was reported as <0.5 μg/mL. Because of this, the maximum value that the steady-state trough could possibly be is 0.5 μg/mL, and this value can be used to compute a rough approximation of the expected concentration:

• Thus, the steady-state trough concentration should be no greater than 0.8 μg/mL. This steady-state trough concentration should be safe and effective for the infection that is being treated.

Example 3

QZ is a 50-year-old, 70-kg (5 ft 10 in) male with gram-negative pneumonia. His current serum creatinine is 0.9 mg/dL, and it has been stable over the last 3 days since admission. A gentamicin dose of 550 mg every 24 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 30 μg/mL and <1 μg/mL, respectively. After the third dose, steady-state peak and trough concentrations were measured and were 37 μg/mL and 1 μg/mL, respectively. Calculate a new gentamicin dose that would provide a steady-state peak of 30 μg/mL and a steady-state trough <1 μg/mL.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• Because the patient has been receiving gentamicin for more than 3–5 estimated half-lives, it is likely that the measured serum concentrations are steady-state values.
• 3. Compute new dose to achieve desired serum concentration.
• Using linear pharmacokinetics, the new dose to attain the desired concentration should be proportional to the old dose that produced the measured concentration:

• The new suggested dose would be 450 mg every 24 hours to be started at next scheduled dosing time.
• 4. Check steady-state trough concentration for new dosage regimen.
• Using linear pharmacokinetics, the new steady-state concentration can be estimated and should be proportional to the old dose that produced the measured concentration:

• This steady-state trough concentration should be safe and effective for the infection that is being treated.

Pharmacokinetic Concepts Method

As implied by the name, this technique derives alternate doses by estimating actual pharmacokinetic parameters or surrogates for pharmacokinetic parameters.101 It is a very useful way to calculate drug doses when the linear pharmacokinetic method is not sufficient because a dosage change that will produce a proportional change in steady-state peak and trough concentrations is not appropriate. The only requirement is a steady-state peak and trough aminoglycoside serum concentration pair obtained before and after a dose (Figure 4-5). The following steps are used to compute new aminoglycoside doses:

• 1.Draw a rough sketch of the serum log concentration/time curve by hand, keeping tract of the relative time between the serum concentrations (Figure 4-5).
• 2.Since the patient is at steady state, the trough concentration can be extrapolated to the next trough value time (Figure 4-5).
• 3.Draw the elimination curve between the steady-state peak concentration and the extrapolated trough concentration. Use this line to estimate half-life. For example, a patient receives a gentamicin dose of 80 mg given every 8 hours that produces a steady-state peak equal to 7 μg/mL and a steady-state trough equal to 3.2 μg/mL, and the dose is infused over 1/2 hour and the peak concentration is drawn 1/2 hour later (Figure 4-5). The time between the measured steady-state peak and the extrapolated trough concentration is 7 hours (the 8-hour dosage interval minus the 1-hour combined infusion and waiting time). The definition of half-life is the time needed for serum concentrations to decrease by half. Because the serum concentration declined by approximately half from the peak concentration to the trough concentration, the aminoglycoside half-life for this patient is approximately 7 hours. This information will be used to set the new dosage interval for the patient.
• 4. Determine the difference in concentration between the steady-state peak and trough concentrations. The difference in concentration will change proportionally with the dose size. In the current example, the patient is receiving a gentamicin dose equal to 80 mg every 8 hours which produced steady-state peak and trough concentrations of 7 μg/mL and 3.2 μg/mL, respectively. The difference between the peak and trough values is 3.8 μg/mL. The change in serum concentration is proportional to the dose, and this information will be used to set a new dose for the patient.
• 5.Choose new steady-state peak and trough concentrations. For the purposes of this example, the desired steady-state peak and trough concentrations will be approximately 7 μg/mL and 1 μg/mL, respectively.
• 6.Determine the new dosage interval for the desired concentrations. In this example, the patient currently has the desired peak concentration of 7 μg/mL. In 1 half-life, the serum concentration will decline to 3.5 μg/mL, in an additional half-life the gentamicin concentration will decrease to 1.8 μg/mL, and in 1 more half-life the concentration will decline to 0.9 μg/mL (Figure 4-6). Since the approximate half-life is 7 hours and 3 half-lives are required for serum concentrations to decrease from the desired peak concentration to the desired trough concentration, the dosage interval should be 21 hours (7 hours × 3 half-lives). This value would be rounded off to the clinically acceptable value of 24 hours, and the actual trough concentration would be expected to be slightly lower than 0.9 μg/mL.
• 7. Determine the new dose for the desired concentrations. The desired peak concentration is 7 μg/mL, and the expected trough concentration is 0.9 μg/mL. The change in concentration between these values is 6.1 μg/mL. It is known from measured serum concentrations that administration of 80 mg changes serum concentrations by 3.8 μg/mL and that the change in serum concentration between the peak and trough values is proportional to the size of the dose. Therefore, a simple ratio will be used to compute the required dose: Dnew = (ΔCnew / ΔCold)Dold, where Dnew and Dold are the new and old doses, respectively; ΔCnew is the change in concentration between the peak and trough for the new dose; and ΔCold is the change in concentration between the peak and trough for the old dose. (Note: This relationship is appropriate because doses are given into a fixed, constant volume of distribution; it is not because the drug follows linear pharmacokinetics so this method will work whether the agent follows nonlinear or linear pharmacokinetics.) For this example: Dnew = (6.1 μg/mL / 3.8 μg/mL) 80 mg = 128 mg, which would be rounded to 130 mg. Gentamicin 130 mg every 24 hours would be started 24 hours after the last dose of the previous dosage regimen.

Once this method is mastered, it can be used without the need for a calculator. The following are examples that use the Pharmacokinetic Concepts method to change aminoglycoside doses.

Example 1

JM is a 50-year-old, 70-kg (5 ft 10 in) male with gram-negative pneumonia. His current serum creatinine is 3.5 mg/dL, and it has been stable over the last 5 days since admission. A gentamicin dose of 115 mg every 24 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 8–10 μg/mL and <2 μg/mL, respectively. After the third dose, steady-state peak and trough concentrations were measured and were 12 μg/mL and 3.5 μg/mL, respectively. Calculate a new gentamicin dose that would provide a steady-state peak of 9 μg/mL and a trough of <2 μg/mL.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• Because the patient has been receiving gentamicin for more than 3–5 estimated half-lives, it is likely that the measured serum concentrations are steady-state values.
• 3.Use Pharmacokinetic Concepts method to compute a new dose.
• 1. Draw a rough sketch of the serum log concentration/time curve by hand, keeping tract of the relative time between the serum concentrations (Figure 4-7).
• 2. Since the patient is at steady state, the trough concentration can be extrapolated to the next trough value time (Figure 4-7).
• 3. Draw the elimination curve between the steady-state peak concentration and the extrapolated trough concentration. Use this line to estimate half-life. The patient is receiving a gentamicin dose of 115 mg given every 24 hours that produces a steady-state peak equal to 12 μg/mL and a steady-state trough equal to 3.5 μg/mL, and the dose is infused over 1/2 hour and the peak concentration is drawn 1/2 hour later (Figure 4-7). The time between the measured steady-state peak and the extrapolated trough concentration is 23 hours (the 24-hour dosage interval minus the 1-hour combined infusion and waiting time). The definition of half-life is the time needed for serum concentrations to decrease by half. It would take 1 half-life for the peak serum concentration to decline from 12 μg/mL to 6 μg/mL, and an additional half-life for the serum concentration to decrease from 6 μg/mL to 3 μg/mL. The concentration of 3 μg/mL is very close to the extrapolated trough value of 3.5 μg/mL. Therefore, 2 half-lives expired during the 23-hour time period between the peak concentration and extrapolated trough concentration, and the estimated half-life is 12 hours (23 hours / 2 half-lives = ~12 hours). This information will be used to set the new dosage interval for the patient.
• 4. Determine the difference in concentration between the steady-state peak and trough concentrations. The difference in concentration will change proportionally with the dose size. In the current example, the patient is receiving a gentamicin dose equal to 115 mg every 24 hours which produced steady-state peak and trough concentrations of 12 μg/mL and 3.5 μg/mL, respectively. The difference between the peak and trough values is 8.5 μg/mL. The change in serum concentration is proportional to the dose, and this information will be used to set a new dose for the patient.
• 5. Choose new steady-state peak and trough concentrations. For the purposes of this example, the desired steady-state peak and trough concentrations will be approximately 9 μg/mL and <2 μg/mL, respectively.
• 6. Determine the new dosage interval for the desired concentrations (Figure 4-8). Using the desired concentrations, it will take 1 half-life for the peak concentration of 9 μg/mL to decrease to 4.5 μg/mL, 1 more half-life for the serum concentration to decrease to 2.3 μg/mL, and an additional half-life for serum concentrations to decline to 1.2 μg/mL. Therefore, the dosage interval will need to be approximately 3 half-lives or 36 hours (12 hours × 3 half-lives = 36 hours). When a dosage interval such as 36 hours is used, care must be taken that the scheduled doses are actually administered as the drug will only be given every other day and sometimes this type of administration schedule is overlooked and doses are missed.
• 7. Determine the new dose for the desired concentrations (Figure 4-8). The desired peak concentration is 9 μg/mL, and the expected trough concentration is 1.2 μg/mL. The change in concentration between these values is 7.8 μg/mL. It is known from measured serum concentrations that administration of 115 mg changes serum concentrations by 8.5 μg/mL and that the change in serum concentration between the peak and trough values is proportional to the size of the dose. In this case: Dnew = (ΔCnew / ΔCold)Dold = (7.8 μg/mL / 8.5 μg/mL) 115 mg = 105 mg. Gentamicin 105 mg every 36 hours would be started 36 hours after the last dose of the previous dosage regimen.

Example 2

ZW is a 35-year-old, 150-kg (5 ft 5 in) female with an intraabdominal infection. Her current serum creatinine is 1.1 mg/dL and is stable. A tobramycin dose of 165 mg every 8 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 6 μg/mL and 0.5 μg/mL, respectively. After the fifth dose, steady-state peak and trough concentrations were measured and were 5 μg/mL and 2.6 μg/mL, respectively. Calculate a new tobramycin dose that would provide a steady-state peak of 6 μg/mL and a steady-state trough ≤1 μg/mL.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is obese [IBWfemales (in kg) = 45 + 2.3(Ht – 60 in) = 45 + 2.3(65 – 60) = 57 kg]. The Salazar and Corcoran equation can be used to estimate creatinine clearance:

• Note: Height is converted from inches to meters: Ht = (65 in · 2.54 cm/in) / (100 cm/m) = 1.65 m.
• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the tobramycin elimination rate for this patient:

• Because the patient has been receiving tobramycin for more that 3–5 estimated half-lives, it is likely that the measured serum concentrations are steady-state values.
• 3.Use Pharmacokinetic Concepts method to compute a new dose.
• A. Draw a rough sketch of the serum log concentration/time curve by hand, keeping tract of the relative time between the serum concentrations (Figure 4-9).
• B. Since the patient is at steady state, the trough concentration can be extrapolated to the next trough value time (Figure 4-9).
• C. Draw the elimination curve between the steady-state peak concentration and the extrapolated trough concentration. Use this line to estimate half-life. The patient is receiving a tobramycin dose of 165 mg given every 8 hours that produces a steady state peak equal to 5 μg/mL and a steady-state trough equal to 2.6 μg/mL, and the dose is infused over 1/2 hour and the peak concentration is drawn 1/2 hour later (Figure 4-9). The time between the measured steady-state peak and the extrapolated trough concentration is 7 hours (the 8-hour dosage interval minus the 1-hour combined infusion and waiting time). The definition of half-life is the time needed for serum concentrations to decrease by half. It would take 1 half-life for the peak serum concentration to decline from 5 μg/mL to 2.5 μg/mL. The concentration of 2.6 μg/mL is very close to the extrapolated trough value of 2.5 μg/mL. Therefore, 1 half-life expired during the 7-hour time period between the peak concentration and extrapolated trough concentration, and the estimated half-life is 7 hours. This information will be used to set the new dosage interval for the patient.
• D.Determine the difference in concentration between the steady-state peak and trough concentrations. The difference in concentration will change proportionally with the dose size. In the current example the patient is receiving a tobramycin dose equal to 165 mg every 8 hours which produced steady-state peak and trough concentrations of 5 μg/mL and 2.6 μg/mL, respectively. The difference between the peak and trough values is 2.4 μg/mL. The change in serum concentration is proportional to the dose, and this information will be used to set a new dose for the patient.
• E. Choose new steady-state peak and trough concentrations. For the purposes of this example, the desired steady-state peak and trough concentrations will be approximately 6 μg/mL and ≤1 μg/mL, respectively.
• F. Determine the new dosage interval for the desired concentrations. Using the desired concentrations, it will take 1 half-life for the peak concentration of 6 μg/mL to decrease to 3 μg/mL, 1 more half-life for the serum concentration to decrease to 1.5 μg/mL, and an additional half-life for serum concentrations to decline to 0.8 μg/mL. Therefore, the dosage interval will need to be approximately 3 half-lives or 21 hours (7 hours × 3 half-lives = 21 hours) which would be rounded to 24 hours.
• G. Determine the new dose for the desired concentrations. The desired peak concentration is 6 μg/mL, and the expected trough concentration is 0.8 μg/mL. The change in concentration between these values is 5.2 μg/mL. It is known from measured serum concentrations that administration of 165 mg changes serum concentrations by 2.4 μg/mL and that the change in serum concentration between the peak and trough values is proportional to the size of the dose. In this case: Dnew = (ΔCnew / ΔCold)Dold = (5.2 μg/mL / 2.4 μg/mL) 165 mg = 358 mg, rounded to 360 mg. Tobramycin 360 mg every 24 hours would be started 24 hours after the last dose of the previous dosage regimen.

Sawchuk-Zaske Method

The Sawchuk-Zaske method of adjusting aminoglycoside doses was among the first techniques available to change doses using serum concentrations.2,47–49,92 It allows the computation of an individual’s own, unique pharmacokinetic constants and uses those to calculate a dose to achieve desired aminoglycoside concentrations. The standard Sawchuk-Zaske method conducts a small pharmacokinetic experiment using 3–4 aminoglycoside serum concentrations obtained during a dosage interval and does not require steady-state conditions. The modified Sawchuk-Zaske methods assume that steady state has been achieved and require only a pair of steady-state concentrations obtained during a dosage interval. The Sawchuk-Zaske method has also been successfully used to dose vancomycin and theophylline.

Standard Sawchuk-Zaske Method

The standard version of the Sawchuk-Zaske method does not require steady-state concentrations. A trough aminoglycoside concentration is obtained before a dose, a peak aminoglycoside concentration is obtained after the dose is infused (immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion), and 1–2 additional postdose serum aminoglycoside concentrations are obtained (Figure 4-10). Ideally, the 1–2 postdose concentrations should be obtained at least 1 estimated half-life from each other to minimize the influence of assay error. The postdose serum concentrations are used to calculate the aminoglycoside elimination rate constant and half-life (Figure 4-10). The half-life can be computed by graphing the postdose concentrations on semilogarithmic paper, drawing the best straight line through the data points, and determining the time needed for serum concentrations to decline by one-half. Once the half-life is known, the elimination rate constant (ke) can be computed: ke = 0.693/t1/2. Alternatively, the elimination rate constant can be directly calculated using the postdose serum concentrations [ke = (ln C1 – ln C2) / Δt, where C1 and C2 are postdose serum concentrations and Δt is the time that expired between the times that C1 and C2 were obtained), and the half-life can be computed using the elimination rate constant (t1/2 = 0.693 / ke). The volume of distribution (V) is calculated using the following equation

where D is the aminoglycoside dose, t′ is the infusion time, ke is the elimination rate constant, Cmax is the peak concentration and Cmin is the trough concentration. The elimination rate constant and volume of distribution measured in this fashion are the patient’s own, unique aminoglycoside pharmacokinetic constants and can be used in one-compartment model intravenous infusion equations to compute the required dose to achieve any desired serum concentration.

Steady-State Sawchuk-Zaske Method: Peak/Trough Version

If a steady-state peak and trough aminoglycoside concentration pair is available for a patient, the Sawchuk-Zaske method can be used to compute patient pharmacokinetic parameters and aminoglycoside doses (Figure 4-11). Since the patient is at steady state, the measured trough concentration obtained before the dose was given can be extrapolated to the next dosage time and used to compute the aminoglycoside elimination rate constant [ke = (ln Cssmax – ln Cssmin)/τ – t′, where Cssmax and Cssmin are the steady-state peak and trough serum concentrations and t′ and τ are the infusion time and dosage interval], and the half-life can be computed using the elimination rate constant (t1/2 = 0.693 / ke). The volume of distribution (V) is calculated using the following equation:

where D is the aminoglycoside dose, t′ is the infusion time, ke is the elimination rate constant, Cssmax is the steady-state peak concentration, and Cssmin is the steady-state trough concentration. The elimination rate constant and volume of distribution measured in this way are the patient’s own, unique aminoglycoside pharmacokinetic constants and can be used in one-compartment model intravenous infusion equations to compute the required dose to achieve any desired serum concentration. The dosage calculations are similar to those done in the initial dosage section of this chapter, except that the patient’s real pharmacokinetic parameters are used in the equations instead of population pharmacokinetic estimates.

Steady-State Sawchuk-Zaske Method: Two Postdose Concentrations Version

Sometimes, steady-state trough concentrations will be below the assay limit or it is not possible to measure a predose concentration. Trough concentrations that are too low to accurately measure occur commonly during therapy with extended-interval aminoglycoside dosing. In these cases, it may be preferable to measure two postdose steady-state concentrations and use these to compute values that can be used in the Sawchuk-Zaske method (Figure 4-12).

The two postdose steady-state concentrations should be drawn at least one estimated half-life apart in order to minimize the effect of assay error on the calculations. While one of the two steady-state concentrations can be a peak concentration, it is not a requirement. During extended-interval dosing, some patients may have longer distribution phases so many clinicians suggest that the first postdose be obtained several hours after the completion of the infusion for this method of administration. The second postdose concentration should be drawn early enough in the dosage interval so that it is not below assay limits (typically no later than 14–16 hours postdose during extended-interval or 4–6-hours postdose during conventional dosing for patients with CrCl > 60 mL/min).

Once the concentrations are obtained, they are extrapolated either mathematically or graphically (Figure 4-12) to determine peak and trough values. The elimination rate constant is calculated using the measured concentrations: ke = (ln C1 – ln C2)/Δt, where C1 and C2 are the first and second steady-state postdose concentrations and Δt is the time that expired between the two concentrations. If one of the concentrations is a peak concentration, it is unnecessary to extrapolate it, and only the trough concentration needs to be computed. However, if neither concentration is a peak, both steady-state peak and trough concentrations need to be calculated: Cssmax = C1 / (e–ket), where C1 is the first measured steady-state concentration, ke is the elimination rate constant, and t is the time between C1 and Cssmax; Cssmin = C2e–ket, where C2 is the second measured steady-state concentration, ke is the elimination rate constant, and t is the time between C2 and Cssmin.

The volume of distribution (V) is calculated using the following equation:

where D is the aminoglycoside dose, t′ is the infusion time, ke is the elimination rate constant, Cssmax is the steady-state peak concentration, and Cssmin is the steady-state trough concentration. The elimination rate constant and volume of distribution measured in this fashion are the patient’s own, unique aminoglycoside pharmacokinetic constants and can be used in one-compartment model intravenous infusion equations to compute the required dose to achieve any desired serum concentration. The dosage calculations are similar to those done in the initial dosage section of this chapter, except that the patient’s real pharmacokinetic parameters are used in the equations instead of population pharmacokinetic estimates.

To illustrate the similarities and differences between the Pharmacokinetic Concepts and the Sawchuk-Zaske methods, some of the same cases used in the previous section will be used as examples here.

Example 1

JH is a 24-year-old, 70-kg (6 ft 0 in) male with gram-negative pneumonia. His current serum creatinine is 1.0 mg/dL, and it has been stable over the last 7 days since admission. An amikacin dose of 400 mg every 8 hours was prescribed. After the third dose, the following amikacin serum concentrations were obtained:

Medication administration sheets were checked, and the previous dose was given 2 hours early (2200 H the previous day). Because of this, it is known that the patient is not at steady state. Calculate a new amikacin dose that would provide a steady-state peak of 28 μg/mL and a trough between 3 μg/mL.

Use Sawchuk-Zaske method to compute a new dose.

• 1.Plot serum concentration/time data (Figure 4-13). Because serum concentrations decrease in a straight line, use any two postdose concentrations to compute the patient’s elimination rate constant and half-life.

• 2.Compute the patient’s volume of distribution.

• 3.Choose new steady-state peak and trough concentrations. For the purposes of this example, the desired steady-state peak and trough concentrations will be 28 μg/mL and 3 μg/mL, respectively.
• 4. Determine the new dosage interval for the desired concentrations. As in the initial dosage section of this chapter, the dosage interval (τ) is computed using the following equation using a 1-hour infusion time (t′):

• 5.Determine the new dose for the desired concentrations. The dose is computed using the one-compartment model intravenous infusion equation used in the initial dosing section of this chapter:

A dose of amikacin 500 mg every 8 hours would be prescribed to begin 8 hours after the last dose of the previous regimen.

Example 2

JM is a 50-year-old, 70-kg (5 ft 10 in) male with gram-negative pneumonia. His current serum creatinine is 3.5 mg/dL, and it has been stable over the last 5 days since admission. A gentamicin dose of 115 mg every 24 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 8–10 μg/mL and <2 μg/mL, respectively. After the third dose, steady-state peak and trough concentrations were measured and were 12 μg/mL and 3.5 μg/mL, respectively. Calculate a new gentamicin dose that would provide a steady-state peak of 9 μg/mL and a trough <2 μg/mL.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• Because the patient has been receiving gentamicin for more that 3–5 estimated half-lives, it is likely that the measured serum concentrations are steady-state values.
• 3.Use Steady-state Sawchuk-Zaske method to compute a new dose.
• 1. Compute the patient’s elimination rate constant and half-life. (Note: For infusion times less than 1 hour, t′ is considered to be the sum of the infusion and waiting times.)

• 2. Compute the patient’s volume of distribution.

• 3. Choose new steady-state peak and trough concentrations. For the purposes of this example, the desired steady-state peak and trough concentrations will be approximately 9 μg/mL and 1.5 μg/mL, respectively.
• 4. Determine the new dosage interval for the desired concentrations. As in the initial dosage section of this chapter, the dosage interval (τ) is computed using the following equation using a 1-hour infusion time (t′):

• 5. Determine the new dose for the desired concentrations. The dose is computed using the one-compartment model intravenous infusion equation used in the initial dosing section of this chapter:

• A dose of gentamicin 100 mg every 36 hours would be prescribed to begin 36 hours after the last dose of the previous regimen. This dose is very similar to that derived for the patient using the Pharmacokinetic Concepts method (105 mg every 36 hours).

Example 3

ZW is a 35–year-old, 150-kg (5 ft 5 in) female with an intraabdominal infection. Her current serum creatinine is 1.1 mg/dL and is stable. A tobramycin dose of 165 mg every 8 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 6 μg/mL and 0.5 μg/mL, respectively. After the fifth dose, steady-state peak and trough concentrations were measured and were 5 μg/mL and 2.6 μg/mL, respectively. Calculate a new tobramycin dose that would provide a steady-state peak of 6 μg/mL and a steady-state trough ≤ 1 μg/mL.

• 1. Estimate creatinine clearance.
• This patient has a stable serum creatinine and is obese (IBWfemales (in kg) = 45 + 2.3(Ht – 60 in) = 45 + 2.3(65 – 60) = 57 kg). The Salazar and Corcoran equation can be used to estimate creatinine clearance:

• Note: Height is converted from inches to meters: Ht = (65 in · 2.54 cm/in) / (100 cm/m) = 1.65 m.
• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the tobramycin elimination rate for this patient:

• Because the patient has been receiving tobramycin for more that 3–5 estimated half-lives, it is likely that the measured serum concentrations are steady-state values.
• 3.Use Steady-state Sawchuk-Zaske method to compute a new dose.
  • 1. Compute the patient’s elimination rate constant and half-life. (Note: For infusion times less than 1 hour, t′ is considered to be the sum of the infusion and waiting times.)
  • 
  • 2. Compute the patient’s volume of distribution.
  • 
  • 3. Choose new steady-state peak and trough concentrations. For the purposes of this example, the desired steady-state peak and trough concentrations will be 6 μg/mL and 0.8 μg/mL, respectively.
  • 4. Determine the new dosage interval for the desired concentrations. As in the initial dosage section of this chapter, the dosage interval (τ) is computed using the following equation using a 1-hour infusion time (t′):
  • 
  • 5. Determine the new dose for the desired concentrations. The dose is computed using the one-compartment model intravenous infusion equation used in the initial dosing section of this chapter:
  • 

• A dose of gentamicin 335 mg every 24 hours would be prescribed to begin 24 hours after the last dose of the previous regimen. This dose is very similar to that derived for the patient using the Pharmacokinetic Concepts method (360 mg every 24 hours).

Example 4

PL is a 52-year-old, 67-kg (5 ft 6 in) female with neutropenia and gram-negative sepsis. Her current serum creatinine is 1.5 mg/dL, and it has been stable over the last 5 days. A gentamicin dose of 110 mg every 12 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 8–10 μg/mL and <2 μg/mL, respectively. After the third dose, steady-state concentrations were measured and were 3.8 μg/mL 1 hour after the end of a 1-hour infusion and 1.6 μg/mL 4 hours after the first concentration. Calculate a new gentamicin dose that would provide a steady-state peak of 9 μg/mL and a trough <2 μg/mL.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• Because the patient has been receiving gentamicin for more that 3–5 estimated half-lives, it is likely that the measured serum concentrations are steady-state values.
• 3.Use Steady-state Sawchuk-Zaske method to compute a new dose.
• 1. Compute the patient’s actual elimination rate constant and half-life. (Note: For infusion times less than 1 hour, t′ is considered to be the sum of the infusion and waiting times.)

• 2. Extrapolate measured concentrations to steady-state peak and trough values.

• 3. Compute the patient’s volume of distribution.

• 4. Choose new steady-state peak and trough concentrations. For the purposes of this example, the desired steady-state peak and trough concentrations will be 9 μg/mL and 1.5 μg/mL, respectively.
• 5. Determine the new dosage interval for the desired concentrations. As in the initial dosage section of this chapter, the dosage interval (τ) is computed using the following equation using a 1-hour infusion time (t′):

• 6. Determine the new dose for the desired concentrations. The dose is computed using the one-compartment model intravenous infusion equation used in the initial dosing section of this chapter:

• A dose of gentamicin 185 mg every 8 hours would be prescribed to begin approximately 8 hours after the last dose of the current regimen.

Example 5

KE is a 67-year-old, 81-kg (5 ft 11 in) male with a hepatic abcess. His current serum creatinine is 1.9 mg/dL, and it has been stable over the last 5 days. A gentamicin dose of 400 mg every 24 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 20 μg/mL and <1 μg/mL, respectively. After the third dose, steady-state concentrations were measured and were 17.5 μg/mL 2 hours after the end of a 1-hour infusion and 4.8 μg/mL 16 hours after the end of infusion. Calculate a new gentamicin dose that would provide a steady-state peak of 20 μg/mL and a trough <1 μg/mL.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• Because the patient has been receiving gentamicin for more that 3–5 estimated half-lives, it is likely that the measured serum concentrations are steady-state values.
• 3. Use Steady-state Sawchuk-Zaske method to compute a new dose.
• 1. Compute the patient’s actual elimination rate constant and half-life. (Note: For infusion times less than 1 hour, t′ is considered to be the sum of the infusion and waiting times.)

• 2. Extrapolate measured concentrations to steady-state peak and trough values.

• 3. Compute the patient’s volume of distribution

• 4. Choose new steady-state peak and trough concentrations. For the purposes of this example, the desired steady-state peak and trough concentrations will be 20 μg/mL and 0.5 μg/mL, respectively.
• 5. Determine the new dosage interval for the desired concentrations. As in the initial dosage section of this chapter, the dosage interval (τ) is computed using the following equation using a 1-hour infusion time (t′):

• 6. Determine the new dose for the desired concentrations. The dose is computed using the one-compartment model intravenous infusion equation used in the initial dosing section of this chapter:

• A dose of gentamicin 400 mg every 36 hours would be prescribed to begin approximately 12 hours after the last dose of the current regimen.

Area Under the Curve Method

Area under the concentration-time curve (AUC) is the best measurement of total exposure to a drug, and some clinicians recommend adjustment of aminoglycoside doses so that target steady-state AUC values are achieved instead of altering doses to attain target steady state peak and trough concentrations. Most often, the AUC method is used with extended-interval aminoglycoside dosing. Different therapeutic AUC values have been suggested by various investigations studying this dosing method. A target AUC equal to 70–120 (mg · h)/L for gentamicin or tobramycin will be used in examples and problems for this section (approximately: 5 mg/kg ≈ 72 (mg · h)/L, 6 mg/kg ≈ 86 (mg · h)/L, and 7 mg/kg ≈ 101 (mg · h)/L for patients with normal renal function).21,23,102–104 Steady-state peak and trough concentrations should also be evaluated when a dosage change is made to assure they are in the appropriate range.

To make use of this approach, the patient is started on an appropriate dose of extended-interval gentamicin or tobramycin. Typical doses of 5–7 mg/kg/d are used as an initial dose, with the dosage interval determined by renal function.3,99 After steady state has been achieved, two postdose serum concentrations are drawn. The two concentrations should be drawn at least one estimated half-life apart in order to minimize the effect of assay error on the calculations. While one of the two steady-state concentrations can be a peak concentration, it is not a requirement. During extended-interval dosing, some patients may have longer distribution phases so many clinicians suggest that the first postdose concentration be obtained several hours after the completion of the infusion. The second postdose concentration should be drawn early enough in the dosage interval so that it is not below assay limits (typically no later than 14–16 hours post dose for patients with CrCl > 60 mL/min).

Once the concentrations are obtained, they are extrapolated either mathematically or graphically (Figure 4-14) to determine steady-state peak and trough values. The elimination rate constant is calculated using the measured concentrations: ke = (ln C1 – ln C2) / Δt, where C1 and C2 are the first and second steady-state postdose concentrations and Δt is the time that expired between the two concentrations. If one of the concentrations is a peak concentration, it is unnecessary to extrapolate it, and only the trough concentration needs to be computed. However, if neither concentration is a peak, both steady-state peak and trough concentrations need to be calculated: Cssmax = C1 / (e–ket), where C1 is the first measured steady-state concentration, ke is the elimination rate constant, and t is the time between C1 and Cssmax; Cssmin = C2e–ket, where C2 is the second measured steady-state concentration, ke is the elimination rate constant, and t is the time between C2 and Cssmin.

The steady-state area under the concentration-time curve during the dosage interval (AUCss) is computed using the following equation:21,23,102–104

The dose is adjusted to attain the target AUCss using linear pharmacokinetics: Dnew = (AUCss,new / AUCss,old)Dold, where Dnew denotes the new computed dose and Dold the original dose, and AUCss,new and AUCss,old are the new target AUCss and the old original AUCss, respectively. Once the new dose has been determined, Cssmax and Cssmin should be calculated to ensure their values are also appropriate for the infection that is being treated: Css,new = (Dnew / Dold)Css,old, where Dnew denotes the new computed dose and Dold the original dose, and Css,new and Css,old are the new target Css and the old original Css, respectively. This calculation is repeated separately for both Cssmax and Cssmin.

Example 1

KE is a 23-year-old, 59-kg (5 ft 4 in) female with salpingitis. Her current serum creatinine is 0.6 mg/dL, and it has been stable over the last 3 days. A gentamicin dose of 250 mg every 24 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 25 μg/mL and <1 μg/mL, respectively. After the third dose, steady-state concentrations were measured and equaled 9.6 μg/mL 2 hours after the end of a 1-hour infusion and 2.6 μg/mL 6 hours after the end of infusion. Calculate a new gentamicin dose that would provide a steady-state AUC of 81 (mg · h)/L.

• 1.Estimate creatinine clearance.
• This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:

• 2. Estimate elimination rate constant (ke) and half-life (t1/2).
• The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:

• Because the patient has been receiving gentamicin for more that 3–5 estimated half-lives, it is likely that the measured serum concentrations are steady-state values.
• 3. Use Steady-state AUC method to compute a new dose.
• 1. Compute the patient’s actual elimination rate constant and half-life. (Note: For infusion times less than 1 hour, t′ is considered to be the sum of the infusion and waiting times.)

• 2. Extrapolate measured concentrations to steady-state peak and trough values.

• 3. Compute the patient’s AUCss (Note: mg/L = μg/mL and this substitution was made to aid the calculation).

• 4. Choose new target AUCss. For the purposes of this example, a desired steady state of AUC of 81 (mg · h)/L was chosen.
• 5. Determine the new dose for the desired AUCss.

• 6. Determine the new steady-state peak and trough concentrations.

• These steady-state peak and trough concentrations are acceptable for the infection being treated and the new prescribed dose would be 350 mg every 24 hours.

Computer programs are available that can assist in the computation of pharmacokinetic parameters for patients.105–109 The most reliable computer programs use a nonlinear regression algorithm that incorporates components of Bayes’ theorem. Nonlinear regression is a statistical technique that uses an iterative process to compute the best pharmacokinetic parameters for a concentration/time data set. Briefly, the patient’s drug dosage schedule and serum concentrations are input into the computer. The computer program has a pharmacokinetic equation preprogrammed for the drug and administration method (oral, intravenous bolus, intravenous infusion, etc.). Typically, a one-compartment model is used, although some programs allow the user to choose among several different equations. Using population estimates based on demographic information for the patient (age, weight, gender, renal function, etc.) supplied by the user, the computer program then computes estimated serum concentrations at each time there are actual serum concentrations. Kinetic parameters are then changed by the computer program, and a new set of estimated serum concentrations are computed. The pharmacokinetic parameters that generated the estimated serum concentrations closest to the actual values are remembered by the computer program, and the process is repeated until the set of pharmacokinetic parameters that result in estimated serum concentrations that are statistically closest to the actual serum concentrations are generated. These pharmacokinetic parameters can then be used to compute improved dosing schedules for patients. Bayes’ theorem is used in the computer algorithm to balance the results of the computations between values based solely on the patient’s serum drug concentrations and those based only on patient population parameters. Results from studies that compare various methods of dosage adjustment have consistently found that these types of computer dosing programs perform at least as well as experienced clinical pharmacokineticists and clinicians and better than inexperienced clinicians.

Some clinicians use Bayesian pharmacokinetic computer programs exclusively to alter drug doses based on serum concentrations. An advantage of this approach is that consistent dosage recommendations are made when several different practitioners are involved in therapeutic drug monitoring programs. However, since simpler dosing methods work just as well for patients with stable pharmacokinetic parameters and steady-state drug concentrations, many clinicians reserve the use of computer programs for more difficult situations. Those situations include serum concentrations that are not at steady state, serum concentrations not obtained at the specific times needed to employ simpler methods, and unstable pharmacokinetic parameters. When only a limited number of aminoglycoside concentrations are available, Bayesian pharmacokinetic computer programs can be used to compute a complete patient pharmacokinetic profile that includes clearance, volume of distribution, and half-life. Many Bayesian pharmacokinetic computer programs are available to users, and most should provide answers similar to the one used in the following examples. The program used to solve problems in this book is DrugCalc written by Dr. Dennis Mungall.110

Example 1

JM is a 50-year-old, 70-kg (5 ft 10 in) male with gram-negative pneumonia. His current serum creatinine is 0.9 mg/dL, and it has been stable over the last 5 days since admission. A gentamicin dose of 170 mg every 8 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 9 μg/mL and 1 μg/mL, respectively. After the third dose, steady-state peak and trough concentrations were measured and were 12 μg/mL and 1.4 μg/mL, respectively. Calculate a new gentamicin dose that would provide a steady-state peak of 9 μg/mL and steady-state trough of 1 μg/mL.

• 1.Enter patient’s demographic, drug dosing, and serum concentration/time data into the computer program.
• 2.Compute pharmacokinetic parameters for the patient using Bayesian pharmacokinetic computer program.
• The pharmacokinetic parameters computed by the program are a volume of distribution of 13.5 L, a half-life equal to 2.1 h, and an elimination rate constant of 0.326 h–1.
• 3. Compute dose required to achieve desired aminoglycoside serum concentrations.
• The one-compartment model intravenous infusion equations used by the program to compute doses indicates that a dose of 135 mg every 8 hours will produce a steady-state peak concentration of 9.2 μg/mL and a steady-state trough concentration of 0.9 μg/mL. Using the simpler linear pharmacokinetics method previously described in the chapter, a similar dose of 140 mg every 8 hours was computed.

Example 2

JM is a 50-year-old, 70-kg (5 ft 10 in) male with gram-negative pneumonia. His current serum creatinine is 3.5 mg/dL, and it has been stable over the last 5 days since admission. A gentamicin dose of 115 mg every 24 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 8–10 μg/mL and <2 μg/mL, respectively. After the third dose, steady-state peak and trough concentrations were measured and were 12 μg/mL and 3.5 μg/mL, respectively. Calculate a new gentamicin dose that would provide a steady-state peak of 9 μg/mL and a steady-state trough equal to 1.5 μg/mL.

• 1.Enter patient’s demographic, drug dosing, and serum concentration/time data into the computer program.
• 2.Compute pharmacokinetic parameters for the patient using Bayesian pharmacokinetic computer program.
• The pharmacokinetic parameters computed by the program are a volume of distribution of 14.6 L, a half-life equal to 14.7 h, and an elimination rate constant of 0.047 h–1. These values are slightly different than those computed using the Steady-state Sawchuk-Zaske method (V = 12.9 L, t1/2 = 12.8 h, ke = 0.054 h–1) because the patient probably was not at steady state when the serum concentrations were drawn.
• 3. Compute dose required to achieve desired aminoglycoside serum concentrations.
• The one-compartment model intravenous infusion equations used by the program to compute doses indicates that a dose of 110 mg every 36 hours will produce a steady-state peak concentration of 9 μg/mL and a steady-state trough concentration of 1.7 μg/mL. Using the Steady-state Sawchuk-Zaske and Pharmacokinetic Concepts methods previously described in the chapter, similar doses of 100 mg every 36 hours and 105 mg every 36 hours, respectively, were computed.

Example 3

JH is a 24-year-old, 70-kg (6 ft 0 in) male with gram-negative pneumonia. His current serum creatinine is 1.0 mg/dL, and it has been stable over the last 7 days since admission. An amikacin dose of 400 mg every 8 hours was prescribed. After the third dose, the following amikacin serum concentrations were obtained:

Medication administration sheets were checked, and the previous dose was given 2 hours early (2200 H the previous day). Because of this, it is known that the patient is not at steady state. Calculate a new amikacin dose that would provide a steady-state peak of 28 μg/mL and a trough between 3–5 μg/mL.

• 1.Enter patient’s demographic, drug dosing, and serum concentration/time data into the computer program.
• 2.Compute pharmacokinetic parameters for the patient using Bayesian pharmacokinetic computer program.
• The pharmacokinetic parameters computed by the program are a volume of distribution of 17.1 L, a half-life equal to 2.4 h, and an elimination rate constant of 0.292 h–1. These values are similar to those computed using the Sawchuk-Zaske method (V = 17.0 L, t1/2 = 2.2 h, ke = 0.311 h–1).
• 3. Compute dose required to achieve desired aminoglycoside serum concentrations.
• The one-compartment model intravenous infusion equations used by the program to compute doses indicates that a dose of 500 mg every 8 hours will produce a steady-state peak concentration of 28 μg/mL and a steady-state trough concentration of 3.6 μg/mL. Using the Sawchuk-Zaske method previously described in this chapter, the identical dose of 500 mg every 8 hours was computed.

Initial dose and dosage adjustment techniques using serum concentrations can be used in any combination as long as the limitations of each method are observed. Some dosing approaches link together logically when considered according to their basic approaches or philosophies. Dosage strategies that follow similar pathways are given in Tables 4-5A and 4-5B.

Hemodialysis Dosing

Aminoglycoside antibiotics are eliminated by dialysis, so renal failure patients receiving hemodialysis must have aminoglycoside dosage regimens that take dialysis clearance into account. Hemodialysis and other extracorporeal methods of drug removal are completely discussed in Chapter 3 (Computation of Initial Doses and Modification of Doses Using Drug Serum Concentrations section).

Example 1

A 62-year-old, 65-kg (5 ft 8 in) male who has chronic renal failure, and receives hemodialysis three times weekly with a low-flux dialysis filter. An initial dosage regimen for tobramycin needs to be computed for a patient to achieve peak concentrations of 6–7 mg/L and postdialysis concentrations 1–2 mg/L.

Initial Dosage Determination

• 1. Patients with renal failure are prone to having poor fluid balance because their kidneys are not able to provide this important function. Because of this, the patient should be assessed for overhydration (due to renal failure) or underhydration (due to renal failure and increased loss due to fever).
  • Weight is a good indication of fluid status, and this patient’s weight is less than his ideal weight [IBWmale = 50 kg + 2.3(Ht – 60 in) = 50 kg + 2.3(68 – 60) = 68 kg]. Other indications of state of hydration (skin turgor, etc.) indicate that the patient has normal fluid balance at this time. Because of this, the average volume of distribution for aminoglycoside antibiotics equal to 0.26 L/kg can be used.
• 2. A loading dose of tobramycin would be appropriate for this patient because the expected half-life is long (~50 h); administration of maintenance doses only might not result in therapeutic maximum concentrations for a considerable time period while drug accumulation is occurring. The loading dose is to be given after hemodialysis ends at 1300 H on Monday (hemodialysis conducted on Monday, Wednesday, and Friday from 0900 – 1300 H).
  • Because the patient is expected to have a long half-life compared to the infusion time of the drug (1/2 – 1 h), little drug will be eliminated during the infusion period, and IV bolus one-compartment model equations can be used. The loading dose for this patient would be based on the expected volume of distribution: V = 0.26 L/kg · 65 kg = 16.9 L; LD = Cmax · V = 6 mg/L · 16.9 L = 101 mg, rounded to 100 mg (LD is loading dose, Cmax is the maximum concentration after drug administration). This loading dose was given at 1400 H (Figure 4-15).
  • Until the next dialysis period at 0900 H on Wednesday, tobramycin is cleared only by the patient’s own body mechanisms. The expected elimination rate constant (ke) for a patient with a creatinine clearance of approximately zero is: ke (in h–1) = 0.00293 · CrCl + 0.014 = 0.00293 (0 mL/min) + 0.014 = 0.014 h–1. The expected concentration at 0900 H on Wednesday is: C = C0e–ket, where C is the concentration at t hours after the initial concentration of C0; C = (6 mg/L)e–(0.014 h–1)(43 h) = 3.3 mg/L.
• 3. While the patient is receiving hemodialysis, tobramycin is eliminated by the patient’s own mechanisms plus dialysis clearance. During hemodialysis with a low-flux filter, the average half-life for aminoglycosides is 4 hours. Because the patient is dialyzed for 4 hours, the tobramycin serum concentration should decrease by 1/2–1.7 mg/L, or using formal computations: ke = 0.693/(t1/2) = 0.693/4 h = 0.173 h–1; C = C0e–ket = (3.3 mg/L)e–(0.173 h–1)(4 h) = 1.7 mg/L.
  • At this time, a postdialysis replacement dose could be given to increase the maximum concentration to its original value of 6 mg/L: Replacement dose = (Cmax – Cbaseline)V = (6 mg/L – 1.7 mg/L)16.9 L = 73 mg, round to 75 mg (where Cmax is the maximum postdose concentration and Cbaseline is the predose concentration). The postdialysis replacement dose of 75 mg was administered at 1400 H on Wednesday. Because all time frames and pharmacokinetic parameters are the same for Monday to Wednesday and Wednesday to Friday, the postdialysis replacement dose on Friday at 1400 H would also be 75 mg.
  • However, more time elapses from Friday after drug administration to Monday before dialysis (67 hours), the next day for hemodialysis to be conducted in the patient and this needs to be accounted for: C = C0e–ket = (6 mg/L)e–(0.014 h–1)(67 h) = 2.3 mg/L. Again, a 4-hour hemodialysis period would decrease serum concentrations by 1/2 to 1.2 mg/L: C = C0e–ket = (2.3 mg/L)e–(0.173 h–1)(4 h) = 1.2 mg/L. At this time, a postdialysis replacement dose could be given to increase the maximum concentration to the original value of 6 mg/L: Replacement dose = (Cmax – Cbaseline)V = (6 mg/L – 1.2 mg/L)16.9 L = 81 mg, round to 80 mg (where Cmax is the maximum postdose concentration and Cbaseline is the predose concentration). The postdialysis replacement dose of 80 mg was administered at 1400 H on Monday.
  • Because all time frames and pharmacokinetic parameters will be the same in subsequent weeks, the following postdialysis replacement doses would be prescribed postdialysis at 1400: Wednesday and Friday 75 mg, Monday 80 mg. In this particular example, recommended daily doses are within 5 mg of each other, and if the clinician wished, the same postdialysis dose could be given on each day. However, this will not be true in every case.

Use of Aminoglycoside Serum Concentrations to Alter Dosages

• 1. Since the initial dosage scheme outlined for this patient used average, estimated pharmacokinetic parameters, it is likely that the patient has different pharmacokinetic characteristics. It is possible to measure the patient’s own unique pharmacokinetic parameters using four serum concentrations (Figure 4-15).
  • The intradialysis elimination rate constant can be determined by obtaining postdose (Cpostdose(1)) and predialysis (Cpredialysis) concentrations [ke = (Cpostdose(1) – Cpredialysis) / Δt, where Δt is the time between the two concentrations], the fraction of drug eliminated by dialysis can be computed using predialysis and postdialysis (Cpostdialysis) concentrations (fraction eliminated = [(Cpredialysis – Cpostdialysis) / Cpredialysis], and the volume of distribution can be calculated using postdialysis and postdose concentrations [V = D / (Cpostdose(2) – Cpredialysis)].
  • Note that if the drug has a postdialysis “rebound” in drug concentrations, postdialysis serum samples should be obtained after blood and tissue have had the opportunity to reequilibrate. In the case of aminoglycosides, postdialysis samples should be collected no sooner than 3–4 hours after the end of dialysis.
• 2. Once individualized pharmacokinetic parameters have been measured, they can be used in the same equations used to compute initial doses in the previous section in place of average, population pharmacokinetic parameters and used to calculate individualized doses for dialysis patients. It is also possible to use a mixture of measured and population- estimated pharmacokinetic parameters. For instance, a clinician may wish to measure the elimination rate constant or volume of distribution for a patient, but elect to use an average population estimate for fraction of drug removed by the artificial kidney.

The following problems are intended to emphasize the computation of initial and individualized doses using clinical pharmacokinetic techniques. Clinicians should always consult the patient’s chart to confirm that antibiotic therapy is appropriate for current microbiologic cultures and sensitivities. Also, it should be confirmed that the patient is receiving other appropriate concurrent antibiotic therapy, such as β-lactam or anaerobic agents, when necessary to treat the infection.

• 1. PQ is a 75-year-old, 62-kg (5 ft 9 in) male with gram-negative sepsis. His current serum creatinine is 1.3 mg/dL, and it has been stable since admission. Compute a gentamicin dose for this patient to provide a steady-state peak concentration of 8 μg/mL and a steady-state trough concentration of 1.5 μg/mL using conventional dosing.
• 2. Patient PQ (please see problem 1) was prescribed gentamicin 110 mg every 12 hours. Steady-state gentamicin concentrations were obtained before and after the fourth dose, and the peak concentration (obtained 1/2 hour after a 1/2-hour infusion of gentamicin) was 9.5 μg/mL while the trough concentration (obtained within 1/2 hour before dosage administration) was 3.0 μg/mL. Compute a revised gentamicin dose for this patient to provide a steady-state peak concentration of 8 μg/mL and a steady-state trough concentration of 1 μg/mL using conventional dosing.
• 3. ZW is a 35-year-old, 75-kg (5 ft 7 in) female with gram-negative pneumonia and chronic renal failure. Her current serum creatinine is 3.7 mg/dL, and it has been stable since admission. Compute a gentamicin dose for this patient to provide a steady-state peak concentration of 10 μg/mL and a steady-state trough concentration of 1.0 μg/mL using conventional dosing.
• 4. Patient ZW (please see problem 3) was prescribed gentamicin 120 mg every 24 hours. Steady-state gentamicin concentrations were obtained before and after the fourth dose, and the peak concentration (obtained 1/2 hour after a 1/2-hour infusion of gentamicin) was 7 μg/mL while the trough concentration (obtained within 1/2 hour before dosage administration) was <0.5 μg/mL. Compute a revised gentamicin dose for this patient to provide a steady-state peak concentration of 10 μg/mL and a steady-state trough concentration of <2 μg/mL using conventional dosing.
• 5. JK is a 55-year-old, 140-kg (5 ft 8 in) male with an intraabdominal infection secondary to a knife wound. His current serum creatinine is 0.9 mg/dL, and it has been stable since admission. Compute a gentamicin dose for this patient to provide a steady-state peak concentration of 6 μg/mL and a steady-state trough concentration of 0.5 μg/mL using conventional dosing.
• 6. Patient JK (please see problem 5) was prescribed gentamicin 120 mg every 8 hours. Steady-state gentamicin concentrations were obtained before and after the fourth dose, and the peak concentration (obtained 1/2 hour after a 1/2 hour infusion of gentamicin) was 5.9 μg/mL while the trough concentration (obtained within 1/2 hour before dosage administration) was 2.5 μg/mL. Compute a revised gentamicin dose for this patient to provide a steady-state peak concentration of 6 μg/mL and a steady-state trough concentration of <1 μg/mL using conventional dosing.
• 7. AF is a 45-year-old, 140-kg (5 ft 2 in) female with an S. viridans endocarditits. Her current serum creatinine is 2.4 mg/dL and is stable. Compute a tobramycin dose for this patient to provide a steady-state peak concentration of 4 μg/mL, and a steady-state trough concentration of 0.5 μg/mL using conventional dosing.
• 8. Patient AF (please see problem 7) was prescribed tobramycin 100 mg every 12 hours. Steady-state tobramycin concentrations were obtained before and after the fourth dose, and the peak concentration (obtained 1/2 hour after a 1/2-hour infusion of tobramycin) was 6.2 μg/mL while the trough concentration (obtained within 1/2 hour before dosage administration) was 1.5 μg/mL. Compute a revised tobramycin dose for this patient to provide a steady-state peak concentration of 4 μg/mL and a steady-state trough concentration of ≤1 μg/mL using conventional dosing.
• 9. FH is a 24-year-old, 60-kg (5 ft 7 in) male with cystic fibrosis and Pseudomonas aeruginosa cultured from a sputum culture. He was hospitalized due to worsening pulmonary function tests. His current serum creatinine is 0.7 mg/dL. Compute a tobramycin dose for this patient to provide a steady-state peak concentration of 10 μg/mL, and a steady-state trough concentration of <2 μg/mL using conventional dosing.
• 10. Patient FH (please see problem 9) was prescribed tobramycin 250 mg every 8 hours. Steady-state tobramycin concentrations were obtained before and after the fourth dose, and the peak concentration (obtained 1/2 hour after a 1/2-hour infusion of tobramycin) was 7.9 μg/mL while the trough concentration (obtained within 1/2 hour before dosage administration) was 1 μg/mL. Compute a revised tobramycin dose for this patient to provide a steady-state peak concentration of 10 μg/mL and a steady-state trough concentration of 1–2 μg/mL using conventional dosing.
• 11. TY is a 66-year-old, 65-kg (5 ft 5 in) female with a suspected tubo-ovarian abscess secondary to hysterectomy surgery. While in the hospital, she developed ascites due to preexisting liver cirrhosis and her current weight is 72 kg. Her current serum creatinine is 1.4 mg/dL. Compute a gentamicin dose for this patient to provide a steady-state peak concentration of 6 μg/mL, and a steady-state trough concentration of <2 μg/mL using conventional dosing.
• 12. Patient TY (please see problem 11) was prescribed gentamicin 120 mg every 12 hours. Steady-state gentamicin concentrations were obtained before and after the fourth dose, and the peak concentration (obtained 1/2 hour after a 1/2-hour infusion of gentamicin) was 4 μg/mL while the trough concentration (obtained within 1/2 hour before dosage administration) was 0.8 μg/mL. Compute a revised gentamicin dose for this patient to provide a steady-state peak concentration of 6 μg/mL and a steady-state trough concentration of 1 μg/mL using conventional dosing.
• 13. UQ is a 27-year-old, 85-kg (6 ft 2 in) male trauma patient with a gram-negative pneumonia and is currently on a respirator. He sustained multiple injuries secondary to a motor vehicle accident 2 weeks ago and lost a large amount of blood at the accident site. He developed acute renal failure due to prolonged hypotension and poor perfusion of his kidneys (current postdialysis serum creatinine is 5.3 mg/dL). He is currently receiving hemodialysis on Mondays, Wednesdays, and Fridays from 0800–1200 H using a low-flux dialysis filter. Recommend a gentamicin dosage regimen that will achieve peak concentrations of 8 μg/mL and postdialysis concentrations of ~2 μg/mL. The first dose of the regimen will be given immediately after hemodialysis is finished on Wednesday at 1200 H.
• 14. Patient UQ (please see problem 13) was prescribed gentamicin 180 mg loading dose and 130 mg after each dialysis. The following serum concentrations were obtained:

• Use these serum concentrations to compute the patient’s own pharmacokinetic parameters for gentamicin and a new dosage schedule that will achieve peak concentrations of 8 μg/mL and postdialysis concentrations of <2 μg/mL.
• 15. LS is a 67-year-old, 60-kg (5 ft 2 in) female with a serum creatinine equal to 1.8 mg/dL placed on tobramycin for a hospital acquired gram-negative pneumonia. The prescribed dose was tobramycin 80 mg every 8 hours (infused over 1 hour) and 2 doses have been given at 0800 and 1600 H. A trough concentration of 2.9 μg/mL was obtained at 1530 H (1/2 hour before the second dose) and a peak concentration of 5.2 μg/mL was obtained at 1705 H (5 minutes after infusion of the second dose). Compute the dose to give Cssmax = 8 μg/mL and Cssmin = 1.5 μg/mL.
• 16. KK is a 52-year-old, 87-kg (6 ft 2 in) male status post appendectomy who developed a fever, elevated white blood cell count, and abdominal pain 24 hours after surgery. His current serum creatinine is 1.4 mg/dL and stable. (A) Compute an initial extended-interval gentamicin dose for this patient. (B) Nine hours after the second dose of gentamicin 610 mg every 24 hours, a gentamicin serum concentration equal to 8.2 μg/mL is measured. Compute a revised gentamicin dose for this patient to provide steady-state peak concentrations above 20 μg/mL and steady-state trough concentrations below 1 μg/mL.
• 17. XS is a 45-year-old, 65-kg (5 ft 4 in) female bone marrow transplant recipient who develops a neutropenic fever. Her current serum creatinine is 1.1 mg/dL. She is administered tobramycin 5 mg/kg daily (325 mg) as part of her antibiotic therapy. A tobramycin serum concentration was obtained 5 hours after the first dose and equaled 19 μg/mL. Compute a revised tobramycin dose for this patient to provide steady-state peak concentrations above 25 μg/mL and steady-state trough concentrations below 1 μg/mL.
• 18. DT is a 3-day-old, 2050-g female with suspected neonatal sepsis. Her serum creatinine has not been measured, but it is assumed that it is typical for her age and weight. Compute an initial tobramycin dose for this patient.
• 19. Patient DT (please see preceding problem) was prescribed tobramycin 5 mg every 12 hours. Steady-state tobramycin concentrations were obtained, and the peak concentration (obtained 1/2 hour after a 1/2-hour infusion of tobramycin) was 4.5 μg/mL while the trough concentration (obtained within 1/2 hour before dosage administration) was 0.9 μg/mL. Compute a revised tobramycin dose for this patient to provide a steady-state peak concentration of 6 μg/mL and a steady-state trough concentration of 1.5 μg/mL using conventional dosing.
• 20. UL is a 7-year-old, 24-kg (3 ft 11 in) male with gram-negative sepsis. His serum creatinine is 0.5 mg/dL, and it has been stable for the last 2 days. Compute an initial gentamicin dose for this patient.
• 21. Patient UL (please see preceding problem) was prescribed gentamicin 60 mg every 8 hours and was expected to achieve steady-state peak and trough concentrations equal to 8 μg/mL and <2 μg/mL, respectively. Steady-state concentrations were measured and were 4.5 μg/mL 1 hour after the end of a 1-hour infusion and 1.5 μg/mL 4 hours after the end of infusion. Calculate a new gentamicin dose that would provide a steady-state peak of 9 μg/mL and a trough of 1 μg/mL.
• 22. RD is a 59-year-old, 79-kg (5 ft 11 in) male with a gram-negative pneumonia. His current serum creatinine is 1.5 mg/dL, and it has been stable over the last 3 days. A gentamicin dose of 450 mg every 24 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 30 μg/mL and <1 μg/mL, respectively. After the second dose, steady-state concentrations were measured and were 16.1 μg/mL 2 hours after the end of a 1-hour infusion and 2.5 μg/mL 16 hours after the end of infusion. Calculate a new gentamicin dose that would provide a steady-state peak of 30 μg/mL and a trough of <1 μg/mL.
• 23. KE is a 23-year-old, 67-kg (5 ft 8 in) male with peritonitis. His current serum creatinine is 0.8 mg/dL, and it has been stable over the last 3 days. A tobramycin dose of 350 mg every 24 hours was prescribed and expected to achieve steady-state peak and trough concentrations equal to 25 μg/mL and <1 μg/mL, respectively. After the second dose, steady-state concentrations were measured and equaled 9.6 μg/mL 2 hours after the end of a 1-hour infusion and 2.6 μg/mL 6 hours after the end of infusion. Calculate a new tobramycin dose that would provide a steady-state AUC of 81 (mg · h)/L.

• 1. Solution to problem 1. The initial gentamicin dose for patient PQ would be calculated as follows:
  • 1. Estimate creatinine clearance.
  • This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:
  • 
  • 2. Estimate elimination rate constant (ke) and half-life (t1/2).
  • The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:
  • 
  • 3. Estimate volume of distribution (V).
  • The patient has no disease states or conditions that would alter the volume of distribution from the normal value of 0.26 L/kg:
  • 
  • 4. Choose desired steady-state serum concentrations.
  • Gram-negative sepsis patients treated with aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 8–10 μg/mL; steady-state trough (Cssmin) concentrations should be <2 μg/mL to avoid toxicity. Set Cssmax = 8 μg/mL and Cssmin = 1.5 μg/mL.
  • 5. Use intermittent intravenous infusion equations to compute dose.
  • Calculate required dosage interval (τ) using a 1-hour infusion:
  • 
  • Dosage intervals should be rounded to clinically acceptable intervals of 8 hours, 12 hours, 18 hours, 24 hours, 36 hours, 48 hours, 72 hours, and multiples of 24 hours thereafter, whenever possible. In this case, the dosage interval would be rounded to 12 hours. Also, steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.
  • 
  • Aminoglycoside doses should be rounded to the nearest 5–10 mg. This dose would be rounded to 110 mg. (Note: μg/mL = mg/L and this concentration unit was substituted for Cssmax so that unnecessary unit conversion was not required.)
  • The prescribed maintenance dose would be 110 mg every 12 hours.
  • 6. Compute loading dose (LD), if needed.
  • Loading doses should be considered for patients with creatinine clearance values below 60 mL/min. The administration of a loading dose in these patients will allow achievement of therapeutic peak concentrations quicker than if maintenance doses alone are given. However, since the pharmacokinetic parameters used to compute these initial doses are only estimated values and not actual values, the patient’s own parameters may be much different than the estimated constants and steady state will not be achieved until 3–5 half-lives have passed.
  • 
  • The gentamicin dose computed using the Hull and Sarubbi nomogram would be:
  • 1. Estimate creatinine clearance.
  • This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:
  • 
  • 2. Choose desired steady-state serum concentrations.
  • Gram-negative sepsis patients treated with gentamicin require steady-state peak concentrations (Cssmax) equal to 8–10 μg/mL.
  • 3. Select loading dose (Table 4-3).
  • A loading dose (LD) of 2 mg/kg will provide a peak concentration of 8–10 μg/mL.
  • 
  • 4. Determine estimated half-life, maintenance dose, and dosage interval.
  • From the nomogram the estimated half-life is 6.5 hours (suggesting that a 12-hour dosage interval is appropriate), the maintenance dose (MD) is 72% of the loading dose [MD = 0.72(125 mg) = 90 mg], and the dosage interval is 12 hours.
  • Aminoglycoside doses should be rounded to the nearest 5–10 mg. Steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.
  • The prescribed maintenance dose would be 90 mg every 12 hours.

• 2. Solution to problem 2. The revised gentamicin dose for patient PQ using the Pharmacokinetic Concepts method would be calculated as follows:
  • 1. Draw a rough sketch of the serum log concentration/time curve by hand, keeping tract of the relative time between the serum concentrations (Figure 4-16).
  • 2. Since the patient is at steady state, the trough concentration can be extrapolated to the next trough value time (Figure 4-16).
  • 3. Draw the elimination curve between the steady-state peak concentration and the extrapolated trough concentration. Use this line to estimate half-life. The patient is receiving a gentamicin dose of 110 mg given every 12 hours that produces a steady-state peak equal to 9.5 μg/mL and a steady-state trough equal to 3.0 μg/mL, and the dose is infused over 1/2 hour and the peak concentration is drawn 1/2 hour later (Figure 4-16). The time between the measured steady-state peak and the extrapolated trough concentration is 11 hours (the 12-hour dosage interval minus the 1-hour combined infusion and waiting time). The definition of half-life is the time needed for serum concentrations to decrease by half. It would take 1 half-life for the peak serum concentration to decline from 9.5 μg/mL to 4.8 μg/mL, and an additional half-life for the serum concentration to decrease from 4.8 μg/mL to 2.4 μg/mL. The concentration of 3.0 μg/mL is close to, but slightly above, the extrapolated trough value of 2.4 μg/mL. Therefore, 1.75 half-lives expired during the 12-hour time period between the peak concentration and extrapolated trough concentration, and the estimated half-life is 7 hours (11 hours / 1.75 half-lives = ~7 hours). This information will be used to set the new dosage interval for the patient.
  • 4. Determine the difference in concentration between the steady-state peak and trough concentrations. The difference in concentration will change proportionately with the dose size. In the current example, the patient is receiving a gentamicin dose equal to 110 mg every 12 hours which produced steady-state peak and trough concentrations of 9.5 μg/mL and 3 μg/mL, respectively. The difference between the peak and trough values is 6.5 μg/mL. The change in serum concentration is proportional to the dose, and this information will be used to set a new dose for the patient.
  • 5. Choose new steady-state peak and trough concentrations. For the purposes of this example, the desired steady-state peak and trough concentrations will be approximately 8 μg/mL and 1 μg/mL, respectively.
  • 6. Determine the new dosage interval for the desired concentrations. Using the desired concentrations, it will take 1 half-life for the peak concentration of 8 μg/mL to decrease to 4 μg/mL, 1 more half-life for the serum concentration to decrease to 2 μg/mL, and an additional half-life for serum concentrations to decline to 1 μg/mL. Therefore, the dosage interval will need to be approximately 3 half-lives or 21 hours (7 hours × 3 half-lives = 21 hours). The dosage interval would be rounded to the clinically acceptable value of 24 hours.
  • 7. Determine the new dose for the desired concentrations. The desired peak concentration is 8 μg/mL, and the expected trough concentration is 1 μg/mL. The change in concentration between these values is 7 μg/mL. It is known from measured serum concentrations that administration of 110 mg changes serum concentrations by 6.5 μg/mL and that the change in serum concentration between the peak and trough values is proportional to the size of the dose. In this case: Dnew = (ΔCnew / ΔCold)Dold = (7 μg/mL / 6.5 μg/mL) 110 mg = 118 mg, rounded to 120 mg. Gentamicin 120 mg every 24 hours would be started 24 hours after the last dose of the previous dosage regimen.
  • The revised gentamicin dose for patient PQ using the Steady-state Sawchuk-Zaske method would be calculated as follows:
  • 1. Compute the patient’s elimination rate constant and half-life. (Note: For infusion times less than 1 hour, t′ is considered to be the sum of the infusion and waiting times.)
  • 
  • 2. Compute the patient’s volume of distribution.
  • 
  • 3. Choose new steady-state peak and trough concentrations. For the purposes of this example, the desired steady-state peak and trough concentrations will be approximately 8 μg/mL and 1 μg/mL, respectively.
  • 4. Determine the new dosage interval for the desired concentrations. As in the initial dosage section of this chapter, the dosage interval (τ) is computed using the following equation using a 1-hour infusion time (t′):
  • 
  • 5. Determine the new dose for the desired concentrations. The dose is computed using the one-compartment model intravenous infusion equation used in the initial dosing section of this chapter:
  • 
  • A dose of gentamicin 120 mg every 24 hours would be prescribed to begin 24 hours after the last dose of the previous regimen. This dose is identical to that derived for the patient using the Pharmacokinetic Concepts method (120 mg every 24 hours).

• 3. Solution to problem 3. The initial gentamicin dose for patient ZW would be calculated as follows:
  • 1. Estimate creatinine clearance.
  • This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:
  • 
  • 2. Estimate elimination rate constant (ke) and half-life (t1/2).
  • The elimination rate constant versus creatinine clearance relationship is used to estimate the gentamicin elimination rate for this patient:
  • 
  • 3. Estimate volume of distribution (V).
  • The patient has no disease states or conditions that would alter the volume of distribution from the normal value of 0.26 L/kg:
  • 
  • 4. Choose desired steady-state serum concentrations.
  • Gram-negative pneumonia patients treated with aminoglycoside antibiotics require steady-state peak concentrations (Cssmax) equal to 8–10 μg/mL; steady-state trough (Cssmin) concentrations should be <2 μg/mL to avoid toxicity. Set Cssmax = 10 μg/mL and Cssmin = 1 μg/mL.
  • 5. Use intermittent intravenous infusion equations to compute dose.
  • Calculate required dosage interval (τ) using a 1-hour infusion:
  • 
  • Dosage intervals should be rounded to clinically acceptable intervals of 8 hours, 12 hours, 18 hours, 24 hours, 36 hours, 48 hours, 72 hours, and multiples of 24 hours thereafter, whenever possible. In this case, the dosage interval would be rounded to 24 hours. Also, steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.
  • 
  • Aminoglycoside doses should be rounded to the nearest 5–10 mg. This dose would be rounded to 180 mg. (Note: μg/mL = mg/L and this concentration unit was substituted for Cssmax so that unnecessary unit conversion was not required.)
  • The prescribed maintenance dose would be 180 mg every 24 hours.
  • 6. Compute loading dose (LD), if needed.
  • Loading doses should be considered for patients with creatinine clearance values below 60 mL/min. The administration of a loading dose in these patients will allow achievement of therapeutic peak concentrations quicker than if maintenance doses alone are given. However, since the pharmacokinetic parameters used to compute these initial doses are only estimated values and not actual values, the patient’s own parameters may be much different than the estimated constants and steady state will not be achieved until 3–5 half-lives have passed.
  • 
  • The gentamicin dose computed using the Hull and Sarubbi nomogram would be:
  • 1. Estimate creatinine clearance.
  • This patient has a stable serum creatinine and is not obese. The Cockcroft-Gault equation can be used to estimate creatinine clearance:
  • 
  • 2. Choose desired steady-state serum concentrations.
  • Gram-negative sepsis patients treated with gentamicin require steady-state peak concentrations (Cssmax) equal to 8–10 μg/mL.
  • 3. Select loading dose (Table 4-3).
  • A loading dose (LD) of 2 mg/kg will provide a peak concentration of 8–10 μg/mL.
  • 
  • 4. Determine estimated half-life, maintenance dose, and dosage interval.
  • From the nomogram the estimated half-life is 9.9 hours (suggesting that a 24-hour dosage interval is appropriate), the maintenance dose (MD) is 81% of the loading dose [MD = 0.81(150 mg) = 122 mg], and the dosage interval is 24 hours. Note: 24-hour dosage interval chosen because longer time period is needed for concentration to drop from 10 μg/mL to 1 μg/mL.
  • Aminoglycoside doses should be rounded to the nearest 5–10 mg. Steady-state peak concentrations are similar if drawn immediately after a 1-hour infusion or 1/2 hour after a 1/2-hour infusion, so the dose could be administered either way.
  • The prescribed maintenance dose would be 120 mg every 24 hours. Note: 24-hour dosage interval chosen because longer time period needed for concentration to decline from 10 μg/mL to 1 μg/mL.