Purpose Results of a study comparing the performance of allometric versus consensus guideline–recommended vancomycin dosing in achieving initial trough concentrations within the desired range are reported.
Methods A retrospective study was conducted to compare selected outcomes with 2 vancomycin dosing methods: (1) dosing by total body weight, as recommended in current consensus guidelines, and (2) a new empirical vancomycin dosing strategy grounded in allometry (the study of the relationship between body size and physiology). The primary outcome was attainment of an initial vancomycin trough concentration within the target range (10–20 mg/L). Rates of nephrotoxicity associated with the 2 dosing methods were compared.
Results Allometric dosing resulted in 77% of the evaluated patient sample (n = 81) achieving vancomycin trough concentration targets at the initial measurement, as compared with a target attainment rate of 57% (n = 81) with guideline-recommended dosing (p = 0.0121); the rate of target attainment in obese patients was also improved with allometric dosing (73% versus 46%, p = 0.0327). Nephrotoxicity rates did not differ significantly between the 2 groups, but a lower rate was observed with allometric versus guideline-based dosing (1.2% versus 7.4%, p = 0.0584).
Conclusion In hospitalized adults, allometric vancomycin dosing achieved a higher frequency of initial vancomycin trough concentrations within the target range of 10–20 mg/L, compared with dosing as recommended by consensus guidelines. The difference between methods in the percentage of troughs within the target range was most pronounced in obese patients.
The aim of allometric vancomycin dosing is to optimize empirical therapy across the body weight continuum.
Approximately 3 out of every 4 patients in the evaluated sample of adult inpatients (n = 81) achieved an initial vancomycin trough concentration of 10–20 mg/L with allometric dosing.
Obese patients experienced similar results when allometric dosing was used, with 73% achieving an initial vancomycin trough of 10–20 mg/L.
Vancomycin is a glycopeptide antibiotic developed in the 1950s and is well known for its clinical utility in the treatment of gram-positive infections involving methicillin-resistant Staphylococcus aureus (MRSA).1 Over the years, vancomycin has been one of the most extensively studied antimicrobial agents, specifically regarding its dosing, monitoring, and potential to cause adverse effects.
In order to optimize empirical vancomycin therapy, a variety of dosing strategies have been utilized, ranging from the use of nomograms to population-based pharmacokinetics.2–4 In 2009, the American Society of Health-System Pharmacists, the Infectious Diseases Society of America, and the Society of Infectious Diseases Pharmacists published consensus guideline recommendations for vancomycin dosing and monitoring in adult patients.1 These guidelines recommend vancomycin doses of 15–20 mg/kg, with doses calculated using total body weight (TBW) and administered every 8–12 hours for most patients with normal renal function.
Unfortunately, dosing medications strictly by weight does not always result in optimal therapeutic effects.5 It can be particularly difficult to provide suitable doses for patients at the extremes of the weight continuum, as weight-based dosing may result in drug concentrations below or above the target range. In addition, the increasing population of obese patients makes optimal vancomycin dosing a challenge.6 Men and women in the United States average 88.9 and 75.5 kg in weight, respectively.7 Therefore, the average U.S. adult weighs substantially more than the classically held view of an average 70-kg adult.7,8
Due to the rising numbers of obese patients in the United States and worldwide, continual research has sought to characterize the effects of obesity on vancomycin therapy.9,10 An increase in body weight is accompanied by physiological changes that can affect vancomycin pharmacokinetics and pharmacodynamics.10 Increases in both adipose tissue and muscle mass associated with obesity contribute to vancomycin’s larger volume of distribution in obese versus nonobese patients. In addition, obesity is associated with increases in certain circulating proteins, which can result in altered free serum vancomycin concentrations. Increased cardiac output and blood volume result in increased blood flow and, consequently, increased vancomycin clearance in obese patients. Weight-based dosing assumes that drug clearance and distribution increase in proportion to body size. In reality, the relationship between TBW and the physiological variables determining clearance of most drugs is nonlinear.11 Thus, weight-based dosing typically leads to overestimation of vancomycin doses for obese patients.
Allometric vancomycin dosing aims to optimize empirical vancomycin therapy by improving the attainment of target drug concentrations. It is based on the discipline of allometry, which is the study of the relationship between body size and physiology.12 Allometric theory is often used to extrapolate drug doses across a variety of mammalian species; it utilizes a 2-variable mathematical power law to appropriately scale doses according to body size.5 In general, a large mammal always requires a lower milligram-per-kilogram dose than a small mammal.13 Due to the variability in body sizes among U.S. adults, allometric principles can theoretically be applied to vancomycin dosing with the goal of achieving isometric exposures of the drug across the weight continuum.
Monitoring of vancomycin therapy is necessary to ensure optimal patient outcomes. The aforementioned consensus guidelines suggest that the ratio of the 24-hour area under the concentration–time curve (AUC) to the minimum inhibitory concentration (MIC) of the targeted organism correlates with vancomycin effectiveness.1 Nevertheless, consensus guidelines recommend using trough vancomycin concentrations as a surrogate for the AUC because troughs are convenient to obtain and measure in clinical practice. Few studies have evaluated the relationship between serum vancomycin concentrations and patient outcomes; however, there is evidence of the potential for unintended effects resulting from concentrations outside the target range.14–20 Potential adverse outcomes include the development of bacterial resistance and nephrotoxicity resulting from concentrations below and above the target range, respectively.
Ultimately, the goal of vancomycin dosing and monitoring is to optimize patient outcomes while reducing the development of bacterial resistance and the risk of adverse events. The purpose of the study described here was to evaluate whether allometric vancomycin dosing increases the attainment of target vancomycin trough concentrations relative to target attainment with dosing based on current consensus guideline recommendations.
The trial was a retrospective pre– and post–protocol implementation study of adults treated with i.v. vancomycin from January 1 through June 30, 2013 (the control period), or from January 1 through June 30, 2014 (the intervention period). The study occurred at a 340-bed acute care regional referral center and was approved by the institutional review boards of all participating institutions.
The study population was identified using CareFusion software (CareFusion Corp., San Diego, CA), and cases were evaluated for study inclusion via review of electronic health records using Cerner Millennium, version 2012.01.26 (Cerner Corp., Kansas City, MO). Patients were included in the analysis if they were greater than 18 years of age, received i.v. vancomycin for at least 72 hours, and had at least 1 measured steady-state vancomycin trough concentration; patients were excluded if they had severe renal dysfunction (defined as an estimated baseline creatinine clearance [CLcr] of <25 mL/min or receipt of hemodialysis), were pregnant, or had a malignancy or cystic fibrosis or if the allometric dosing method was not used during the period January–June 2014.
Allometric dosing implementation. The hospital pharmacy department was responsible for vancomycin dosing and monitoring for all patients receiving vancomycin therapy during the study period. Allometric vancomycin dosing was formally introduced to the pharmacy department in January 2014. A hypothetical equation derived from the allometric 2-variable power law was used to calculate allometric vancomycin doses for patients in the intervention group.21 This allometric dosing equation can be expressed as follows: allometric dose (mg) = average dose (mg) × [TBW (kg)/average TBW (kg)]β, where β is an allometric exponent by which the dose is scaled according to the patient’s body size. In this equation, the average dose is the dose calculated for a patient of average weight per the current consensus guideline recommendations (i.e., a dose of 15 mg/kg).1,21 The average TBW for U.S. adults is approximately 80 kg; therefore, the average vancomycin dose for purposes of our study was 1,200 mg. The allometric vancomycin dosing equation employs a constant, theoretical allometric exponent of 0.5,21 which lies within the range of the most commonly accepted allometric exponent values (typically 0.4–0.75); furthermore, the use of this value simplifies the calculation and facilitates application of the equation in the clinical setting, since an exponent of 0.5 is equivalent to the mathematical square-root function. Thus, the allometric vancomycin dosing equation used in the study was as follows: allometric dose (mg) = 1,200 mg × [TBW (kg)/80 kg]0.5.
In this version of the equation, the only variable needed to calculate an allometric dose is the patient’s TBW. In general, compared with consensus-guided dosing, allometric dosing results in a slightly higher vancomycin dose for patients weighing less than 80 kg and a lower vancomycin dose for patients weighing more than 80 kg. A comparison of vancomycin doses calculated for patients of various body sizes using consensus guideline dosing and allometric dosing is illustrated in Figure 1.
Data collection and study definitions. Demographic and clinical data gathered from electronic health records included sex, age, race, weight, serum creatinine concentration (SCr), CLcr, comorbidities, admission to the intensive care unit (ICU), vancomycin treatment data (including indication, dose, dosing interval, duration of therapy, and initial trough concentration), and concomitant use of nephrotoxic agents. Patients in the control group received vancomycin doses consistent with consensus guideline recommendations (15–20 mg/kg, calculated using TBW). In the intervention group, vancomycin doses were determined using the allometric dosing equation discussed previously. Both guideline-consistent and allometric doses were rounded to the nearest 100 mg. No patients in the study received loading doses of vancomycin. Dosing intervals for both groups were determined on the basis of renal function, as estimated by CLcr. Patients with CLcr values of ≥75, 50–74, and 25–49 mL/min received vancomycin every 8, 12, or 24 hours, respectively, per institution protocol. The Cockcroft–Gault22 equation was used to calculate CLcr using TBW. For patients with a TBW of ≥120% of the ideal body weight (IBW), as calculated using the Devine23 equations, an adjusted body weight (ABW) was used to calculate the CLcr. The ABW was calculated as follows: ABW = 0.4 (TBW − IBW) + IBW. Patients 65 years of age or older may have falsely elevated estimated CLcr due to low SCr values reflecting reduced muscle mass24; to account for this, low SCr values were adjusted according to institution protocol. SCr values of <1 mg/dL were adjusted to 1 mg/dL in patients 65 years of age or older, and SCr values of <1.2 mg/dL were adjusted to 1.2 mg/dL in patients 80 years of age or older. Blood samples for measurement of trough vancomycin concentrations were obtained immediately prior to the fourth dose in patients receiving vancomycin every 8 or 12 hours and immediately prior to the third dose in patients receiving vancomycin every 24 hours.
Study outcomes. The primary outcome measure was the percentage of patients achieving an initial vancomycin trough concentration within the desired range of 10–20 mg/L. The percentages of patients with below- and above-target initial trough concentrations were also determined. As a secondary effectiveness outcome, we assessed the percentage of patients achieving an initial target trough concentration for each of the 4 main categories of body mass index (BMI)—underweight, normal weight, overweight, and obese—as classified by the World Health Organization.25 Additionally, the frequency of nephrotoxicity with each dosing method was evaluated as a secondary safety outcome. Nephrotoxicity was defined as an increase in SCr of 0.5 mg/dL or an increase from baseline of ≥50%, whichever was greater, on 2 consecutive measurements in the absence of an alternative explanation.1
Statistical analysis. Unpublished analyses of target trough concentration attainment using traditional vancomycin dosing at our institution (i.e., dosing based on consensus guideline recommendations, as described above) demonstrated that the initial vancomycin trough fell within the desired range in approximately 50–60% of patients. To provide a statistical power of 80%, 81 patients were needed in each group to detect a 20% improvement in the percentage of patients achieving an initial vancomycin trough concentration within the desired range. A comparison of baseline characteristics was performed using Student’s t test to analyze continuous data and the Kruskal–Wallis rank sum test to analyze categorical data. The primary outcome analysis was conducted using Fisher’s exact test. In addition, the difference in the rates of initial target trough attainment with the 2 dosing methods was assessed for each BMI category using Fisher’s exact test. The frequency of nephrotoxicity was evaluated using Fisher’s exact test, while similarity of concomitant nephrotoxic agent use was analyzed using the Kruskal–Wallis rank sum test. The a priori level of significance was 0.05 for 2-sided tests. All statistical tests were performed using SAS software, version 9.4 (SAS Institute Inc., Cary, NC).
Patient selection and demographics. A total of 2,039 patients were identified as having received vancomycin during the control (n = 934) and intervention (n = 1,105) periods, respectively. These cases were randomized and reviewed until 81 patients meeting the inclusion criteria for each study group were identified; 284 and 258 cases were evaluated for inclusion in the control and intervention groups, respectively. A comparison of patient characteristics at baseline is presented in Table 1. The groups were similar in respect to sex, age, race, renal function, comorbid conditions, and indications for vancomycin therapy. The mean weight did not differ significantly between the groups; however, the allometric dosing group had a higher percentage of patients with a BMI of 25.0–29.9 kg/m2 (p = 0.0443).
Vancomycin trough concentrations. Allometric dosing resulted in a higher frequency of initial trough concentrations within the target range (62 of 81 patients [77%]), as compared with consensus guideline dosing (46 of 81 patients [57%]; odds ratio [OR], 2.48; 95% confidence interval [CI], 1.26–4.88; p = 0.0121). The overall mean ± S.D. trough vancomycin concentrations were 15.0 ± 4.9 mg/L for the allometric dosing group and 15.7 ± 8.3 mg/L for the guideline dosing group (p = 0.5143). Initial trough concentrations were below the target range in 9 of 81 patients (11%) and 19 of 81 patients (23%) in the allometric dosing group and the guideline dosing group, respectively (p = 0.0601). In addition, 10 of 81 patients (12%) in the allometric dosing group had initial above-target trough concentrations, as compared with 16 of 81 patients (20%) in the guideline dosing group (p = 0.2844). Within the allometric dosing group, the frequencies of within-target vancomycin trough values did not differ significantly between men (37 of 45 patients [82%]) and women (25 of 36 patients [69%]) (p = 0.1976).
In obese patients (BMI of ≥30 kg/m2), allometric dosing resulted in a higher percentage of initial vancomycin trough concentrations within the desired range (24 of 33 patients [73%]) than consensus guideline dosing (19 of 41 patients [46%]; OR, 3.09; 95% CI, 1.16–8.24; p = 0.0327). For the same outcome, there was no significant difference between the dosing methods for any of the other 3 BMI categories.
Nephrotoxicity. The frequencies of nephrotoxicity did not differ significantly between the study groups (1% with allometric dosing versus 7% with guideline dosing; OR, 0.15; 95% CI, 0.02–1.33; p = 0.0584) (Table 2). The 2 groups were similar in respect to overall use of concomitant nephrotoxic agents; however, there was a higher percentage of patients taking angiotensin-converting enzyme inhibitors or angiotensin receptor blockers in the guideline dosing group versus the allometric dosing group.
Historically, vancomycin dosing has been challenging due to the drug’s complex pharmacokinetic and pharmacodynamic properties. Patients with altered physiological processes, such as obese patients, further complicate the ability to achieve target trough concentrations.5,26 Currently, vancomycin doses are typically selected only on the basis of body size, specifically in regard to weight; this assumes a proportionate relationship of body size to the physiological variables determining pharmacokinetics.5 In other words, the usual practice is to give a 120-kg patient twice the dose given to a 60-kg patient. However, vancomycin pharmacokinetics and pharmacodynamics become altered with increasing weight, and several analyses have demonstrated that obese patients usually require lower weight-based doses to achieve target vancomycin concentrations.3,7,26,27
In contrast to weight-based dosing, allometric vancomycin dosing involves the use of TBW values to extrapolate doses that more accurately correspond with the nonlinear relationships between body size and the physiological variables determining vancomycin pharmacokinetics.5,11,21 Many recent studies have sought to optimize vancomycin therapy by using alternative dosing strategies for patients in different weight (or BMI) categories; however, the threshold at which an alternative strategy should be used is still not clear.3,4,27,28 Because allometric vancomycin dosing involves the use of weight values to scale doses, the same dosing strategy (or allometric equation) could theoretically be used for all patients within the human weight distribution.
In our retrospective single-center trial, allometric vancomycin dosing, as compared with consensus guideline dosing, improved the overall attainment of initial trough concentrations within the target range of 10–20 mg/L. Because allometric dosing has the potential to optimize vancomycin therapy for patients of any weight, an evaluation of target trough concentration attainment for each of the primary BMI classifications was performed. We were unable to detect a difference between the 2 dosing methods for patients with a BMI of <30 kg/m2. This result may be related to underrepresentation by underweight, normal-weight, and overweight patients in the study, as obese patients constituted the majority of the population. Conversely, allometric dosing improved target trough concentration attainment for obese patients—a population for which vancomycin dosing has been especially challenging. Interestingly, target trough concentration attainment in the allometric dosing group did not differ significantly by sex. This finding seems to validate the use of a single allometric vancomycin dosing equation despite differences in average body weight between men and women.
Furthermore, compared with other dosing methods, allometric dosing resulted in a higher frequency of initial trough concentrations within the target range (77%). An observational study performed by Nunn et al.29 found that dosing regimens based on the current vancomycin consensus guidelines achieved an initial trough concentration of 10–20 mg/L only 33.8% of the time; in our study, that figure was 57%. Reynolds et al.3 retrospectively compared a revised dosing protocol for obese patients with a protocol based on consensus guideline recommendations and found that the revised dosing protocol resulted in 59% of patients reaching target trough concentrations of 10–20 mg/L, as compared with 36% of patients treated using the original dosing protocol (p = 0.006). In a prospective trial of a nomogram developed to improve the attainment of trough concentrations within the target range of 10–20 mg/L in both nonobese and obese patients, a target trough concentration was attained with the initial dosing regimen in 44% of patients, as compared with 33% of patients whose initial doses were calculated using traditional pharmacokinetic methods (p = 0.014).30 A recent study conducted by Hong et al.28 evaluated a 2-sample dosing strategy for obese patients through assessment of subsequent attainment of target trough concentrations when the initial trough value was outside the desired range. Empirical dosing in both study groups was performed using the same population-based pharmacokinetic equations. An analysis of initial trough concentrations found that the initial measurement was greater than 10 mg/L in 76% and 85.3% of patients in the preintervention and postintervention groups, respectively (p = 0.214). Moreover, initial trough concentrations were greater than 20 mg/L in 25.3% and 30.7% of patients in the preintervention and postintervention groups, respectively (p = 0.586). In our study, allometric dosing resulted in 89% of patients having an initial trough concentration of >10 mg/L; however, only 12% of patients in the allometric dosing group had an initial trough value of >20 mg/L. Based on our results, we believe allometric dosing has the potential to improve empirical vancomycin therapy beyond the improvements reported in previously published trials.
Since allometric dosing results in lower doses (on a milligram-per-kilogram basis) for overweight and obese patients than consensus guideline dosing, we hypothesized that allometric dosing would result in a lower frequency of nephrotoxicity. Unfortunately, our study was not powered to detect a statistical difference in rates of nephrotoxicity with the 2 dosing methods. The overall frequency of nephrotoxicity in our study (7 of 162 patients [4%]) was similar to the rate of 5–7% described in a study involving the relatively pure formulations of vancomycin currently used in clinical practice.19 However, the rate of nephrotoxicity that we observed with allometric dosing (1 of 81 patients [1%]) was much lower than rates reported in other studies.17,18 Due to the observed difference in nephrotoxicity rates between our study groups, we believe the potential safety benefits of allometric dosing warrant further evaluation.
Overall, the results of this study are applicable to a relatively diverse population including patients of various ages, body sizes, and degrees of renal function. Our study also included patients with various infection types and severities, with 45 of 162 patients (28%) in the overall study population spending at least 1 day in the ICU. However, our study was not free of limitations. These results cannot be applied to patients not represented in our study population, such as patients with a CLcr of <25 mL/min and pediatric, cancer, and severe trauma populations. Also, because of the heterogeneity of the indications for vancomycin therapy, clinical outcomes were not evaluated.
We also did not assess MIC data to determine if the achieved trough concentrations were likely to provide 24-hour AUC:MIC values of ≥400. This limitation was a consequence of the retrospective nature of our study and our institution’s lack of software needed to calculate AUC values from single trough concentration measurements. However, we recognize that both vancomycin effectiveness and toxicity are likely related to AUC, which suggests that some complicated MRSA infections may be adequately treated with vancomycin trough concentrations of <15 mg/L. A population pharmacokinetic modeling study performed by Neely et al.31 determined that 50–60% of adults who achieve a vancomycin AUC of ≥400 mg · hr/L are not expected to have vancomycin trough concentrations of >15 mg/L. In addition, a recent meta-analysis of clinical outcomes in patients with S. aureus bacteremia found that vancomycin trough concentrations of ≥15 mg/L were not associated with better clinical outcomes; however, 24-hour AUC:MIC values of ≥400 were associated with significant reductions in treatment failure, persistent bacteremia, and mortality.32 Despite the association between AUC and clinical outcomes, measurement of serum trough concentrations remains the most widely used method of assessing the likely effectiveness and safety of vancomycin therapy in clinical practice today. Well-designed prospective studies are needed to determine which therapeutic target (trough concentration versus AUC:MIC) results in better clinical outcomes.
Additionally, there may be opportunities to optimize the allometric dosing equation used in this study. Recently, the value of the allometric exponent β has been the subject of much debate. The optimal allometric exponent is thought to be patient-specific, since the value of β has been shown to change within the human weight distribution.33,34 Therefore, the use of a constant allometric exponent, as in our study, remains controversial. Additional research should be conducted to enhance the allometric equation employed in our study with the goal of further optimizing empirical vancomycin therapy. Lastly, we did not assess the duration of concomitant nephrotoxic agent use or the time to occurrence of nephrotoxicity. Performing these assessments in future research may help clarify the safety of allometric vancomycin dosing.
Future studies should prospectively evaluate allometric vancomycin dosing and should include types of patients not represented in our study. Due to the increasing prevalence of obesity within the U.S. population, allometric vancomycin dosing should be assessed within the various subcategories of obesity. As methods of AUC calculation, such as the Bayesian approach, become more practically incorporated into clinical practice, additional research should focus on the efficacy of allometric vancomycin dosing in terms of optimizing AUC:MIC values. An appropriately powered study to assess the risk of nephrotoxicity with allometric dosing should also be performed. Studies to validate the effectiveness of allometric vancomycin dosing in terms of clinical outcomes are also needed. Finally, allometric dosing should be explored as a method of optimizing the use of other antimicrobial agents.
In hospitalized adults, allometric vancomycin dosing achieved a higher frequency of initial vancomycin trough concentrations within the target range of 10–20 mg/L, compared with dosing as recommended by consensus guidelines. The difference between methods in the percentage of troughs within the target range was most pronounced in obese patients.
Manjunath P. Pai, Pharm.D., is acknowledged for using allometric theory to optimize the dosing of vancomycin and other antimicrobial agents.
This article will appear in the July 15, 2017, issue of AJHP.
The authors have declared no potential conflicts of interest.
- Copyright © 2017 by the American Society of Health-System Pharmacists, Inc. All rights reserved.