Antimicrobial Pharmacokinetics and Pharmacodynamics

Abstract

Background:

Antimicrobial medications are beneficial when used appropriately, but adverse effects and resistance sometimes limit therapy. These effects may be more problematic with inappropriate antimicrobial use. Consideration of the pharmacokinetic and pharmacodynamic properties of these medications can help optimize drug use.

Methods:

Review of the pertinent English-language literature.

Results:

The pharmacokinetic principles of absorption, distribution, metabolism, and elimination determine whether an appropriate dose of medication reaches the intended pathogen. The pharmacodynamic properties of antimicrobial medications define the relation between the drug concentration and its observed effect on the target pathogen. Improvements in clinical outcomes have been observed when antimicrobial agents are dosed optimally according to these properties. In surgical patients, substantial changes in the volume of distribution and elimination necessitate a clear understanding of these principles. Additionally, less adverse drug effects and antimicrobial resistance may occur with optimal use of these drugs.

Conclusion:

Selecting and dosing antimicrobial medications with consideration of pharmacokinetics and pharmacodynamics may improve patient outcomes and avoid adverse effects.

In the presence of infection, antimicrobial medications can convey substantial, even life-saving, benefits [1 –3]. Selection of the right medication at the right dose for the right duration is crucial to the success of therapy. Unfortunately, these drugs can be harmful, even when used correctly. Antimicrobial agents come with associated adverse effects, which can limit the practical application of therapy. Issues such as acute kidney injury, pancytopenia, and QTc prolongation sometimes render therapies intolerable. Additionally, resistance and Clostridium difficile infection can be associated with antimicrobial agent use [4 –7]. All of these untoward effects can occur with appropriate antimicrobial therapy, but they may be more problematic and common with inappropriate or unnecessary use.

Examples of inappropriate or unnecessary use of antimicrobial agents include: (1) Use of a drug when an infection is not present; (2) use of the wrong drug to treat a certain infection; (3) use of the wrong dose of a drug; or (4) use of a drug for a longer duration than necessary. The Infectious Diseases Society of America has noted that as much as 50% of antimicrobial agent use can be considered inappropriate [8]. This review focuses on optimizing drug and dose selection by consideration of pharmacokinetic and pharmacodynamic principles. By understanding these fundamental drug properties, clinicians can make more informed decisions when selecting antimicrobial agents, and they can dose these medications for a specific infection more effectively.

Pharmacokinetics

Clinical pharmacokinetics describes the complex process of absorption, distribution, metabolism, and elimination of drugs. By applying pharmacokinetics, a clinician can determine whether an appropriate dose of medication will reach the intended pathogen.

Absorption is the transfer of a drug from the site of administration to the systemic circulation. Although this term frequently refers to gastrointestinal absorption, other routes of absorption, such as transdermal and subcutaneous, are utilized sometimes.

Once the absorbed drug reaches the systemic circulation, it may remain in the blood stream or leave the vasculature and distribute into tissue. The ratio between the amounts of drug in the body and the serum concentration is referred to as the volume of distribution (V_D) [9]. A high V_D implies that the drug is distributed extensively to tissue (lipophilic), whereas a low V_D suggests the drug is concentrated in the plasma (mainly hydrophilic) [9].

It should be noted that drugs do not distribute uniformly in the body. Certain anatomic compartments, including bone, cerebrospinal fluid, and abscesses, are penetrated poorly by some antibiotics [10]. Additionally, distribution in any compartment is dictated by multiple factors, including the chemical properties of the drug, the characteristics of the tissues into which it distributes, and the blood flow resulting in various apparent V_D throughout a dosing cycle [10].

Pharmacokinetic parameters can be altered grossly in critically ill patients. Hydrophilic agents generally demonstrate a low V_D that is similar to that of extracellular water (0.1–0.6 L/kg) [11]. However, a larger V_D has been demonstrated with hydrophilic antibiotics, including the aminoglycosides, β-lactams, daptomycin, vancomycin, and colistin, in critically ill patients [12,13]. These antibiotics distribute primarily into the extracellular space, which is expanded in critically ill and surgical patients by large-volume resuscitation, infusion of blood products, positive-pressure ventilation, surgical procedures, and capillary leakage [9]. A larger V_D contributes to a lower antibiotic serum concentration that may lead to suboptimal bacterial killing and potential treatment failure [9,13]. Additionally, subtherapeutic concentrations may result in the development of antimicrobial resistance [13]. As a result, hydrophilic antimicrobial agents frequently require a loading dose to overcome the pharmacokinetic changes during critical illness, and this loading dose may exceed standard recommendations in non-critically ill patients [13].

Hydrophilicity and lipophilicity should be taken into consideration when deciding on a patient's dosing weight. Generally, doses of hydrophilic antibiotics are based on ideal or adjusted body weight, although this has not been shown with all antimicrobial agents [14]. Alternatively, lipophilic antimicrobial agents have greater tissue and intracellular penetration [13]. Lipophilic antimicrobial agents include fluoroquinolones, macrolides, and tigecycline. The V_D of these agents is unchanged by sepsis but is increased in obesity. For this reason, lipophilic medications usually require dosing based on total body weight [14].

Another factor affecting pharmacokinetics is altered protein binding. Albumin is the serum protein responsible for most of the drug–protein binding. Critically ill patients may have a reduced serum concentration of albumin, leading to a smaller amount of antibacterial drug bound to albumin, thereby increasing the unbound fraction of the drug [15]. This is important because the unbound fraction, or free drug, is the only drug available to exert pharmacologic effects. Additionally, the free drug is the only drug available for distribution and clearance. Therefore, hypoalbuminemia likely increases the V_D and clearance of a drug, leading to lower antibiotic exposure [15]. Examples of antibiotics that are moderately-to-highly protein bound and that would have an increased V_D in the setting of hypoalbuminemia include ceftriaxone, ertapenem, and daptomycin [16].

Metabolism is another pharmacokinetic parameter that should be considered when dosing an antimicrobial medication. The liver is perhaps the most notable contributor to drug metabolism, but other organs, such as the lungs and gastrointestinal tract wall, possess enzymes that metabolize drugs. The blood contains esterases, which cleave ester bonds in drug molecules. If organ, and particularly hepatic, dysfunction exists, the potential for altered metabolism should be considered when dosing medications.

Elimination is the final pharmacokinetic parameter to consider. Many antimicrobial agents are cleared from the body primarily by renal elimination. The kidney can excrete drugs and their metabolites by glomerular filtration or by proximal tubular secretion. Whereas hypoalbuminemia can result in increased clearance, other factors in enhanced renal elimination in critically ill patients include hyperdynamic conditions such as sepsis, increased ventricular preload following aggressive fluid administration, or the use of positive inotropes [13]. By contrast, renal elimination can be reduced in the setting of acute kidney injury. Renal dysfunction occurs frequently in critically ill patients, particularly in the context of sepsis [13]. Hydrophilic antibiotics are most affected by changes in renal elimination, whereas lipophilic antimicrobials are often cleared hepatically. Drugs also can be eliminated via biliary excretion or respiratory expiration. If antimicrobial agent doses are not adjusted for decreased elimination, drug concentrations may become supratherapeutic, which could result in adverse effects. With the potential for drug elimination to change in either direction, it is necessary to consider these factors to avoid under-dosing or over-dosing.

Pharmacodynamics

Pharmacodynamic parameters define the relation between the antimicrobial concentration and the observed effect on the target pathogen in the body. Factors that contribute to pharmacodynamics include the maximum drug concentration (C_max), the area under the elimination-curve (AUC), the minimum inhibitory concentration (MIC) of the organism, the duration of bactericidal effects, post-antibiotic effects (PAE), rate of time killing, and rate of development of resistance [17].

Antimicrobials exhibit their killing as a function of time dependence, concentration dependence, or a combination of time and concentration dependence, defined as the ratio AUC:MIC, where the MIC is defined as the minimum concentration of an antibacterial drug that prevents the visible growth of an organism, usually a 10-fold increase in the density of bacterial colony-forming units (CFU) [17]. The MIC is unique to a specific combination of micro-organism and antimicrobial drug.

Time-dependent killing

Time-dependent killing is maximized as the cumulative percentage of time that the free drug concentration exceeds the MIC of the organism. The prototypical examples of time-dependent antimicrobials are β-lactams, including penicillins, cephalosporins, and carbapenems. For penicillins and carbapenems, maximum killing occurs when the antibiotic concentration is maintained above the MIC for at least 40% to 50% of the dosing interval (T_>MIC), whereas cephalosporin concentrations must achieve 60% to 70% T_>MIC [16,17]. The percentage of time above the MIC is exceedingly important in the dosing of β-lactams, as these drugs exhibit little if any PAE and do not suppress bacterial growth after exposure, allowing microorganisms to resume growth rapidly once the drug concentration decreases below the MIC [17].

Adequate dosing of β-lactam antibiotics to achieve sustained concentrations above the MIC continues to be a challenge, especially in critically ill patients. The labeled doses for antimicrobials are obtained from studies in healthy volunteers and often do not account for pharmacokinetic and pharmacodynamic differences among critically ill patients, including increased V_D and augmented drug clearance. The prospective multinational Defining Antibiotic Levels in Intensive Care Patients (DALI) study of 384 patients found that among those who received intermittent infusions, 20% did not achieve “free drug concentration maintained above MIC of the known or suspected pathogen for at least 50% of the dosing interval” (50% fT_>MIC), whereas only 7% received a prolonged (at least 2 h) or continuous infusion [18]. Furthermore, patients who were treated for infection and did not achieve 50% fT_>MIC were significantly less likely to have a positive clinical outcome (odds ratio [OR] 0.68; 95% confidence interval [CI] 0.52–0.91; p=0.009]) [18].

Another multicenter randomized trial in 60 patients receiving continuous infusions compared with intermittent boluses of β-lactam antibiotics (ticarcillin-clavulanic acid, piperacillin-tazobactam, or meropenem) in severe sepsis found concentrations exceeded the MIC in 82% of patients in the continuous arm vs. 29% of patients in the intermittent arm (p=0.001). Clinical cure was more common in the group receiving continuous infusions than in those receiving intermittent infusions (70% vs. 43%, respectively; p=0.037), but the mortality rate was not significantly different (90% vs. 80%, respectively; p=0.47) [19].

The findings in these recent randomized trials have confirmed, to some extent, the findings of previous meta-analyses, including a number of smaller studies. One meta-analysis of randomized controlled trials and cohort studies of critically ill patients receiving continuous or extended infusions vs. traditional intermittent dosing of β-lactam antibiotics found a significant risk reduction in clinical failure rates with continuous/extended infusions (relative risk [RR] 0.60; 95% CI 0.41–0.87). The pooled results demonstrated a reduction in mortality risk nearing statistical significance (RR 0.83; 95% CI 0.69–1.00). However, when the mortality rate was analyzed by class of antibiotic, the only class to demonstrate a significantly reduced mortality rate was piperacillin-tazobactam (or piperacillin only) (RR 0.62; 95% CI 0.46–0.85) [20].

Similar mortality results were noted in a meta-analysis of continuous or extended infusion of piperacillin-tazobactam or carbapenems, including but not limited to critically ill patients, in which pooled results indicated a significantly reduced mortality rate with continuous or extended vs. short-term infusions (RR 0.59; 95% CI 0.41–0.83). This benefit again appeared to be driven by the benefit observed with piperacillin-tazobactam, as piperacillin-tazobactam (RR 0.55; 95% CI 0.34–0.89), but not carbapenems (RR 0.66; 95% CI 0.34–1.30), was associated with a lower mortality rate when analyzed by drug class. Interestingly, no significant benefit was noted with regard to clinical cure with piperacillin-tazobactam (RR 1.11; 95% CI 0.95–1.31) or carbapenems (RR 1.16; 95% CI 0.82–1.65) [21]. On the other hand, a Cochrane review of randomized trials published in 2013 found no difference in clinical cure or the all-cause mortality rate [22]. Additionally, differences were not observed in infection recurrence or super-infection post-therapy between patients receiving continuous vs. intermittent antibiotic dosing regimens. However, it should be noted that both time- and concentration-dependent antimicrobial drugs were included in this review, and definitions of clinical cure differed widely [22].

These data, taken together, suggest that a continuous- or extended-infusion piperacillin-tazobactam dosing regimen may improve clinical outcomes and decrease the mortality rate among critically ill patients with infections. The optimal dosing strategy of carbapenems and cephalosporins required to improve clinical outcomes and decrease the mortality rate remains to be determined and may include extended infusions of drug; however, further investigation is needed, specifically a large randomized trial.

Concentration-dependent killing

Concentration-dependent killing is maximized by free drug peak concentration (C_max) in a dosing interval divided by the MIC. The prototypical examples of concentration-dependent antimicrobial drugs are the aminoglycosides. The clinical efficacy of these drugs is related directly to the ratio of C_max:MIC, with higher C_max:MIC demonstrating improved outcomes and optimal outcomes in the treatment of gram-negative infections occurring with a ratio ≥10 [23].

Early studies of optimal aminoglycoside C_max:MIC ratios utilized the drugs as monotherapy for the treatment of gram-negative infections. This practice has been largely abandoned since the regulatory approval of extended-spectrum β-lactams and data demonstrating higher mortality rates with aminoglycoside monotherapy vs. β-lactam monotherapy or aminoglycoside/β-lactam combination therapy, the exception being patients treated for a urinary tract infection. Additionally, in the absence of neutropenia or pseudomonal infection, combination β-lactam/aminoglycoside therapy provides no additional benefit [24,25]. Conversely, a study of early combination antibiotic therapy with an aminoglycoside demonstrated a mortality advantage compared with monotherapy in the treatment of septic shock (36.3% vs. 29%, respectively; hazard ratio [HR] 0.77; 95% CI 0.67–0.88; p=0.0002). The beneficial effect was likely not attributable to expanded breadth of coverage, as the analysis included only culture-positive infection with organisms known to be susceptible to both antimicrobial drugs, but to the synergistic effect combination therapy may have had with regard to bacterial clearance [26]. There are few data to guide the choice of the optimal peak concentration (C_max) or C_max:MIC ratio required when aminoglycosides are used in combination with an extended-spectrum β-lactam or other antimicrobial drug targeting gram-negative infection.

In addition to concentration-dependent killing, aminoglycosides exhibit an extended PAE whereby suppression of microbial growth is maintained after drug concentrations fall below the MIC. Some studies have demonstrated a PAE for gram-negative organisms of as long as 10 h (for Pseudomonas aeruginosa) to >12 h (for Klebsiella pneumoniae) [24].

The use of aminoglycosides comes with risk. Nephrotoxicity secondary to accumulation of drug in the kidney, which may cause acute tubular necrosis, and ototoxicity secondary to penetration of drug into the endolymph, vestibular system, and cochlear tissue are of primary concern. The development of these adverse effects appears to be most closely related to total exposure (AUC) and may be mitigated by once-daily dosing for a short time (e.g., less than 7 d) [24]. It appears that optimal dosing regimens for aminoglycosides maximize the C_max/MIC ratio, allow adequate clearance to minimize toxicities and take advantage of the PAE; thus, a high-dose, extended-interval strategy may be ideal.

Concentration-dependent killing with time dependence

Concentration-dependent killing with time dependence occurs when the antimicrobial effect is defined by the AUC of free drug over a 24-h period divided by the MIC. Vancomycin, daptomycin, linezolid, and fluoroquinolones are examples of drugs that exhibit killing when AUC:MIC is maximized. Both linezolid and vancomycin have demonstrated greater clinical efficacy when AUC:MIC is maximized, particularly in the treatment of infections of the lower respiratory tract [27]. Clinical improvement and microbiologic eradication of Staphylococcus aureus were improved significantly when the vancomycin AUC:MIC ratio exceeded 400 [27,28]. Similarly, a study of ciprofloxacin for the treatment of Enterobacteriaceae blood stream infections demonstrated that an AUC₂₄/MIC ratio >250 was associated with a significantly greater cure rate (91.4% vs. 28.6%, respectively; p=0.001) [29].

Complex Antimicrobial Dosing in Surgical Patients and the Critically Ill

Antimicrobial dosing in the critically ill can be exceedingly challenging. Critically ill patients have alterations in a number of pharmacokinetic parameters, such as fluid balance, drug clearance, and organ function, including mechanisms of organ support. Fluid balance may change rapidly, often increasing quickly as a result of an inflammatory response, large-volume fluid resuscitation, or diuresis. Drug clearance is altered by changes in hemodynamics. Hyperdynamic cardiac output may increase drug clearance, whereas renal or hepatic impairment may decrease drug clearance. Additionally, renal replacement therapy or extracorporeal membrane oxygenation can increase or decrease the volume of distribution and augment clearance. The clinical status of critically ill patients often changes rapidly, and thus dosing must be evaluated and altered frequently [16].

Resistance: The Mutant Selection Window

The prevalence of drug-resistant pathogens is increasing. Although transmission of resistant pathogens can be minimized by infection control methods, recent studies have focused on strategies to control the precursor: The acquisition of resistance. The traditional dosing strategy to prevent mutations has been targeted at killing susceptible pathogens to reduce the pool from which new mutants arise. This approach addresses only bulk population susceptibility, which is measured by MIC, but does not consider the resistant subpopulation. As a result, resistance can emerge during the killing of susceptible cells. An alternative approach utilizes direct measurement of the resistant mutant subpopulation or the mutant prevention concentration (MPC). The MPC is estimated by placing 10¹⁰ CFU on agar (MIC determination uses 10⁴–10⁵ cells) with increasing drug concentrations. Large inocula ensure the presence of mutant subpopulations. The drug concentration at which there is no growth is the MPC; therefore, the MPC estimates resistant subpopulation susceptibility [30]. In other words, the MPC is the lowest drug concentration that prevents the growth of the least susceptible first-step resistant mutants [31]. The concentration range between the MIC of the organisms and the MPC is termed the mutant selection window (MSW). The MSW hypothesis suggests that drug concentrations should be kept above the MPC to restrict resistance selection.

The relation between drug exposure and the likelihood of resistance suppression has been analyzed in vitro utilizing pharmacokinetic and pharmacodynamic modeling, as well as in animal studies. Olofsson et al. found that clinical dosing regimens of fluoroquinolones were effective in eliminating susceptible Escherichia coli populations, but ciprofloxacin 750 mg twice daily was the only dose that prevented the selection of both of the E. coli mutants studied [31]. Similar to fluoroquinolones, β-lactam drugs have a threshold exposure that counter-selects resistance in P. aeruginosa. However, Tam et al. suggest that the threshold exposure necessary to suppress resistant subpopulations will differ by β-lactam subclasses because of different bactericidal profiles and various types of first-step mutations (e.g., AmpC, efflux pumps, etc.) [32]. Specifically, the propensity of meropenem to suppress P. aeruginosa resistance was studied. Resistant subpopulations were suppressed with a C_min/MIC≥6.2 or by adding tobramycin to meropenem (C_min/MIC=1.7) [33]. Likewise, the pharmacodynamics of daptomycin and vancomycin and their ability to prevent the selection of resistant S. aureus were analyzed in a study by Firsov et al. [34]. Their findings suggest comparable antistaphylococcal effects of clinically achievable AUC₂₄:MIC₉₀ values of daptomycin and vancomycin but slightly better prevention against the selection of resistant S. aureus by daptomycin [34]. The application of MPC-based pharmacokinetic/pharmacodynamic thresholds to patients requires standardized methodology for the measurement of MPC and development of databases in which MPC is established for many organisms. Optimal dosing to treat bacterial infections while preventing the emergence of resistance requires further study.

Conclusion

Understanding how to select and dose antimicrobial drugs appropriately is crucial to patient care, because appropriate use of these agents influences outcomes. The principles of pharmacokinetics and pharmacodynamics can inform clinician decision-making and ensure that effective agents are administered in sufficient doses to treat infection. Patient-specific factors, particularly for surgical and critically ill patients, should be taken into consideration when making these decisions in order to treat infection optimally, avoid adverse effects, and decrease antimicrobial resistance.

Footnotes

Author Disclosure Statement

No authors have no competing financial interests.

References

Leibovici

, Shraga

, Drucker

, et al. The benefit of appropriate empirical antibiotic treatment in patients with bloodstream infection. J Intern Med, 1998; 244:379–386.

Ibrahim

, Sherman

, Ward

, et al. The influence of inadequate antimicrobial treatment of bloodstream infections on patient outcomes in the ICU setting. Chest, 2000; 118:146–155.

Kumar

, Roberts

, Wood

, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med, 2006; 34:1589–1596.

Muller

, Mauny

, Bertin

, et al. Relationship between spread of methicillin-resistant Staphylococcus aureus and antimicrobial use in a French university hospital. Clin Infect Dis, 2003; 36:971–978.

Gottesman

, Carmeli

, Shitrit

, Chowers

. Impact of quinolone restriction on resistance patterns of Escherichia coli isolated from urine by culture in a community setting. Clin Infect Dis, 2009; 49:869–875.

Stevens

, Dumyati

, Fine

, et al. Cumulative antibiotic exposures over time and the risk of Clostridium difficile infection. Clin Infect Dis, 2011; 53:42–48.

McFarland

, Clarridge

, Beneda

, Raugi

. Fluoroquinolone use and risk factors for Clostridium difficile-associated disease within a Veterans Administration health care system. Clin Infect Dis, 2007; 45:1141–1151.

Dellit

, Owens

, McGowan

Jr , et al. Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis, 2007; 44:159–177.

Goncalves-Pereira

, Povoa

. Antibiotics in critically ill patients: A systematic review of the pharmacokinetics of beta-lactams. Crit Care, 2011; 15:R206.

10.

Fissell

. Antimicrobial dosing in acute renal replacement. Adv Chronic Kidney Dis, 2013; 20:85–93.

11.

Sinnollareddy

, Roberts

, Lipman

, Roberts

. Beta-lactam pharmacokinetics and pharmacodynamics in critically ill patients and strategies for dose optimization: A structured review. Clin Exp Pharmacol Physiol, 2012; 39:489–496.

12.

Udy

, Roberts

, Lipman

. Clinical implications of antibiotic pharmacokinetic principles in the critically ill. Intens Care Med, 2013; 39:2070–2082.

13.

Blot

, Lipman

, Roberts

. The influence of acute kidney injury on antimicrobial dosing in critically ill patients: Are dose reductions always necessary?. Diagn Microbiol Infect Dis, 2014; 79:77–84.

14.

Janson

, Thursky

. Dosing of antibiotics in obesity. Curr Opin Infect Dis, 2012; 25:634–649.

15.

Ulldemolins

, Roberts

, Rello

, et al. The effects of hypoalbuminaemia on optimizing antibacterial dosing in critically ill patients. Clin Pharmacokinet, 2011; 50:99–110.

16.

Roberts

, Abdul-Aziz

, Lipman

, et al. Individualized antibiotic dosing for patients who are critically ill: Challenges and potential solutions. Lancet Infect Dis, 2014; 14:498–509.

17.

Levison

, Levision

. Pharmacokinetics and pharmacodynamics of antibacterial agents. Infect Dis Clin North Am, 2009; 23:791–815; vii.

18.

Roberts

, Paul

, Akova

, et al. DALI: Defining antibiotic levels in intensive care unit patients: Are current beta-lactam antibiotic doses sufficient for critically ill patients?. Clin Infect Dis, 2014; 58:1072–1083.

19.

Dulhunty

, Roberts

, Davis

, et al. Continuous infusion of beta-lactam antibiotics in severe sepsis: A multicenter double-blind, randomized controlled trial. Clin Infect Dis, 2013; 56:236–244.

20.

Chant

, Leung

, Friedrich

. Optimal dosing of antibiotics in critically ill patients by using continuous/extended infusions: A systematic review and meta-analysis. Critical Care, 2013; 17:R:279.

21.

Falagas

, Tansarli

, Ikawa

, Vardakas

. Clinical outcomes with extended or continuous versus short-term intravenous infusion of carbapenems and piperacillin/tazobactam: A systematic review and meta-analysis. Clin Infect Dis, 2013; 56:272–282.

22.

Shiu

, Wang

, Tejani

, Wasdell

. Continuous versus intermittent infusions of antibiotics for the treatment of severe acute infections. Cochrane Database Syst Rev, 2013; 3:CD008481.

23.

Moore

, Lietman

, Smith

. Clinical response to aminoglycoside therapy: Importance of the ratio of peak concentration to minimal inhibitory concentration. J Infect Dis, 1987; 155:93–99.

24.

Craig

. Optimizing aminoglycoside use. Crit Care Clin, 2011; 27:107–121.

25.

Leibovici

, Paul

, Poznanski

, et al. Monotherapy versus beta-lactam-aminoglycoside combination treatment for gram-negative bacteremia: A prospective observational study. Antimicrob Agents Chemother, 1997; 41:1127–1133.

26.

Kumar

, Zarychanski

, Light

, et al. Early combination antibiotic therapy yields improved survival compared with monotherapy in septic shock: A propensity-matched analysis. Crit Care Med, 2010; 38:1773–1785.

27.

Housman

, Kuti

, Nicolau

. Optimizing antibiotic pharmacodynamics in hospital-acquired and ventilator-acquired bacterial pneumonia. Clin Chest Med, 2011; 32:439–450.

28.

Moise-Broder

, Forrest

, Birmingham

, Schentag

. Pharmacodynamics of vancomycin and other antimicrobials in patients with Staphylococcus aureus lower respiratory tract infections. Clin Pharmacokinet, 2004; 43:925–942.

29.

Zelenitsky

, Ariano

. Support for higher ciprofloxacin AUC₂₄/MIC targets in treating Enterobacteriaceae bloodstream infection. J Antimicrob Chemother, 2010; 65:1725–1732.

30.

Zhao

, Drlica

. A unified anti-mutant dosing strategy. J Antimicrob Chemother, 2008; 62:434–436.

31.

Olofsson

, Marcusson

, Stromback

, et al. Dose-related selection of fluoroquinolone-resistant Escherichia coli. J Antimicrob Chemother, 2007; 60:795–801.

32.

Tam

, Schilling

, Melnick

, Coyle

. Comparison of beta-lactams in counter-selecting resistance of Pseudomonas aeruginosa. Diagn Microbiol Infect Dis, 2005; 52:145–151.

33.

Tam

, Schilling

, Neshat

, et al. Optimization of meropenem minimum concentration/MIC ratio to suppress in vitro resistance of Pseudomonas aeruginosa. Antimicrob Agents Chemother, 2005; 49:4920–4927.

34.

Firsov

, Smirnova

, Lubenko

, et al. Testing the mutant selection window hypothesis with Staphylococcus aureus exposed to daptomycin and vancomycin in an in vitro dynamic model. J Antimicrob Chemother, 2006; 58:1185–1192.