OBJECTIVE

Vancomycin carries risks of treatment failure and emergent resistance with underexposure and renal toxicity with overexposure. Children with overweight or obesity may have altered pharmacokinetics. We aimed to examine how body weight metrics influence vancomycin serum concentrations and to evaluate alternative dosing strategies.

METHODS

This was a multicenter retrospective cohort study across 3 large, academic hospitals. Patients aged 2 to 18 years old who received ≥3 doses of intravenous vancomycin were included. Weight metrics included total body weight, adjusted body weight, ideal body weight, body surface area, and allometric weight. Outcomes included vancomycin concentration and ratios of area under the curve (AUC) to minimum inhibitory concentration (MIC). Regression analyses were used to examine which body-weight identifier predicted outcomes.

RESULTS

Of the 1099 children, 45% were girls, mean age was 9.0 (SD = 5.4) years, 14% had overweight, and 17% had obesity. Seventy-five percent of children had vancomycin concentrations in the subtherapeutic range by trough <10 µg/mL, and 63% had a ratio of AUC to MIC <400 μg-hr/mL. Three percent had a supratherapeutic initial trough >20 µg/mL or ratio of AUC to MIC >600 μg-hr/mL. Serum vancomycin concentrations were higher in children with overweight or obesity compared with children who were at a normal weight or underweight; the mean ratio of AUC to MIC also trended higher in the groups with overweight or obesity.

CONCLUSIONS

Most children received vancomycin regimens that produced suboptimal trough levels. Children with overweight or obesity experienced higher vancomycin trough levels than children of normal weight despite receiving lower total body weight dosing. Using the ratio of AUC to MIC was a better measure of drug exposure.

Childhood obesity, as a health condition, has far-reaching consequences for the US health care system and affects a significant proportion of hospitalized children as a comorbid condition.13  The prevalence of obesity among children aged 2 to 19 years was 18.5% in 2015–2016 by using established Centers for Disease Control and Prevention (CDC) guidelines on definitions of overweight and obesity based on BMI.4,5 Medication dosing in children is typically based on body weight. Children with obesity are more likely to be dosed inappropriately compared with children with normal weight.6  Children with obesity may have unexpected drug exposures because of differences in body composition, greater fat mass, higher total body water, and greater fat-free mass than children at normal weight. These differences between the populations may lead to changes in volume of distribution and clearance, affecting drug exposure.7 

Authors of a systematic review found scant evidence of pharmacokinetic studies in children with obesity for 21 different drugs.8  As obesity is becoming increasingly prevalent among hospitalized children, there is a need to evaluate dosing strategies of commonly prescribed medications in the inpatient setting.

There have been several relatively small studies in which researchers have examined vancomycin pharmacokinetic variations in children with obesity compared with children with a normal weight.915  The importance of these studies lies in the consequences of underexposure and concomitant treatment failure or of overexposure, resulting in nephrotoxicity. In most previous studies, researchers have found no difference in nephrotoxicity incidence or peak or trough concentrations in patients with and without obesity using total body weight (TBW) dosing, the most common way to dose vancomycin in the hospital setting. This may be a function of the sample size because larger studies (eg, Heble [2013], with N = 126) identify significantly higher trough concentrations in children with obesity.12  More recently, in another study in children (N = 174), researchers found significant differences in vancomycin pharmacokinetic parameters (volume of distribution and clearance) by adiposity.16 

We conducted a retrospective chart review to examine the relationship between child adiposity measures and vancomycin drug concentrations and ratio of area under the curve (AUC) to minimum inhibitory concentration (MIC) given the Infectious Disease Society of America 2020 guidelines recommend targeting the ratio of AUC to MIC.17  We aimed to evaluate alternative dosing strategies to TBW with other body-weight descriptors: BMI, ideal body weight (IBW), adjusted body weight (ABW), body surface area (BSA), and allometric weight (AW).

We conducted a retrospective review of pediatric patients (2–18 years of age) treated with intravenous (IV) vancomycin across 3 large, academic medical centers. We extracted data from the electronic health records of each institution by using similar protocols. For Texas, the data are from patients admitted from January 2012 to October 2016; for California, the data are from January 2012 to January 2017; and for Oregon, the data are from January 2009 to August 2018. The difference in dates primarily reflects variation in electronic record keeping over time.

Analysis was limited to patients 2 to 18 years of age with no chronic renal or acute or chronic hepatic disease, who received ≥3 doses of IV vancomycin and had complete weight and height data. We excluded the patients with the following International Classification of Diseases, Ninth Revision, and International Classification of Diseases, 10th Revision codes: 585, chronic kidney disease; N18, chronic kidney disease; 593.9, unspecified disorder of kidney and ureter; N28.9, disorder of kidney and ureter, unspecified; and K70–77, diseases of the liver. Patients for whom vancomycin was given via oral, ophthalmic, or continuous IV were removed from the data set.

Institutional review board approval for data abstraction from the medical record was obtained independently at each participating institution. Data were extracted with relevant identifiers: sex, height (centimeters) and weight (kilograms), age in months, length of stay (days), order start time, drug administration data, and drug concentration of vancomycin with timing. Patients who received <3 doses of vancomycin were excluded because they would not have valid trough concentrations. All 3 sites used TBW to guide dosing, and there was variation at all 3 sites in dosing frequency. At 2 sites (California and Texas), 2 independent reviewers performed individual data reviews to ensure that the blood sample was appropriately obtained ∼30 minutes before the fourth dose and that the patient had received 3 identical doses before the trough by using extracted drug administration data. At the Oregon site, drug administration data were not available in the data set. At that site, the accuracy of the sample was determined by using a calculated time window from the electronic order entry. The validity of this method was examined by using a sample chart review and resulted in inclusion of appropriately collected laboratory values but exclusion of a number of vancomycin troughs that were likely accurate but could not be confirmed. This resulted in a smaller contribution from the Oregon site in the overall data set. The majority of excluded patients were those who received <3 doses of vancomycin or did not have an appropriately timed trough level.

BMI was calculated by using the formula BMI equals weight in kilograms divided by height in meters squared. Epi Info 7 (CDC, Atlanta, GA) was used to compute BMI percentiles and z scores. Degree of adiposity was categorized as underweight (BMI < 5th percentile), normal weight (BMI in the 5th to 85th percentile), overweight (BMI in the 85th to 95th percentile) and obese (BMI > 95th percentile) by using CDC guidelines.18  The Mosteller equation for estimation of BSA was used.19  IBW was estimated with the formula20  IBW = 2.396 * e0. 01863 (height in centimeters). After IBW was calculated, ABW was estimated with the formula ABW = IBW + 0.4 (TBW−IBW).21  ABW is by definition higher than IBW, with the notion of accounting for some pharmacokinetic distribution into fat tissue. AW takes a similar approach; it was calculated by applying the three-quarters allometric scaling method as weight0.75. Vancomycin dose per TBW, IBW, ABW, BSA, and AW was calculated for each patient. Children who received a dose of TBW per kilogram >100 mg/kg per day were excluded from further analysis (n = 3). There were several maximum adult-dose recommendations at the time of the study; we used 3 g/day and examined the weight distribution of children receiving this dose or higher.

The vancomycin drug serum concentration data were examined, and levels >40 µg/mL were identified as outliers and were excluded (n = 18); we did examine whether these outliers were differentially distributed by weight, and they were not. We categorized the vancomycin drug serum concentration as <5 µg/mL (undetectable), 5 to 10 µg/mL (subtherapeutic), 10 to 15 µg/mL (therapeutic-uncomplicated infections), 15 to 20 µg/mL (therapeutic-complicated infections), or >20 µg/mL (supratherapeutic) for each patient. The final analytic sample included 1099 patients. Two centers provided creatinine data (n = 759).

The ratio of AUC to MIC (abbreviated as AUC) was calculated by using an assumed MIC of 1. The AUC was calculated for children for whom a creatinine level was obtained before the first dose of vancomycin. AUC was calculated as the total daily dose divided by vancomycin clearance. The clearance for each subject was estimated by using the equations from pharmacokinetic modeling (including age, weight, and serum creatinine). We used separate equations for children with overweight or obesity16 :

Cl (L/h) = 0.286 × weight (kg)∧0.75 × (0.4/Cr [mg/dL])∧0.290 × (ln[age]/8.3)∧0.755 and for children with underweight or normal weight22 :

Cl (L/h) = 0.248 × weight (kg)∧0.75 × (0.48/Cr [mg/dL])∧0.361 × (ln[age]/7.8)∧0.995

Using previously published studies,12,23  we estimated the sample size required to detect a significant difference using a type 1 error rate of 0.05, estimated mean vancomycin levels of 15 µg/mL in a combined group with overweight and obesity and 10 µg/mL in the normal weight group, with a SD of 5 µg/mL for a desired power of 0.80. Given the potential for publication biases in early studies, we estimated the sample size required if the mean difference was half of that observed in the previous studies (2.5 µg/mL), and we determined that we would need 126 subjects in a normal weight group and 126 in a group with overweight and obesity to detect this difference.

Descriptive statistics, including bivariate analyses and regression analyses, were performed by using SPSS Statistics version 24 (IBM SPSS Statistics, IBM Corporation). We used regression analyses to determine which body-weight identifier (TBW, BMI, IBW, BSA, and ABW) predicted ideal vancomycin trough concentrations most accurately, using predicted probabilities and best-fit metrics as outcomes. We examined variation in dosing regimens and patient characteristics between the different hospitals and over time. Given the variation, we included age, sex, and dosing frequency regimen in the final models and chose not to include a variable for hospital as a random intercept.

In the final analytic sample (N = 1099), the mean age was 9.0 years (SD = 5.4) and 490 (45%) were girls (Table 1). Of this sample, 162 (15%) had a BMI below the 5th percentile (underweight), 599 (55%) had a normal BMI (5th to 85th percentile), 153 (14%) had overweight (85th to 95th percentile), and 185 (17%) had obesity (>95th percentile). The Texas site provided valid data for 352 subjects, the California site 552 subjects, and the Oregon site 195. The Texas and Oregon sites had similar vancomycin prescription patterns in dose and frequency (most commonly 45–60 mg/kg per day divided every 6 hours by using TBW); the California site differed in the majority of dosing being every 8 hours (85% in California versus 41% for Oregon and 46% for Texas), driven by local practice and not associated with any clinical characteristics.

TABLE 1

Vancomycin Dosing and Covariates by Weight Status, Using Combined Hospital Data

Underweight, n = 162Normal wt, n = 599Overweight, n = 153Obese, n = 185All, n =1099P
Age, y, mean (SD) 7.7 (5.0) 8.6 (5.4) 10.3 (5.5) 10.5 (5.1) 9.0 (5.4) <.001 
Girls, n (%) 65 (40) 274 (45) 81 (53) 70 (38) 490 (45) .03 
Dosing regimen, n (%)      .03 
 every 6 h 45 (28) 183 (31) 41 (27) 39 (21) 308 (28)  
 every 8 h 110 (68) 379 (63) 95 (62) 127 (69) 711 (65)  
 every 12 h 7 (4) 37 (6) 17 (11) 19 (10) 80 (7)  
Vancomycin dose in mg/kg/d, mean (SD)       
 By TBW 50 (11)a 48 (11)a 45 (10)b 42 (12)c 47 (11) <.001 
 By IBW 41 (9)a 49 (12)b 57 (13)c 65 (19)d 52 (15) <.001 
 By ABW 44 (10)a 49 (11)b 51 (11)c 53 (14)c 49 (12) <.001 
 By BSA 1179 (325)a 1317 (351)b 1437 (340)c 1483 (378)c 1341 (363) <.001 
 By AW 103 (28)a 108 (27)a 111 (25)a 108 (27)a 107 (27) .05 
Vancomycin concentration, mean (SD) 7.0 (3.9)a 7.7 (4.8)a,b 8.6 (5.3)b,c 9.2 (5.3)c 8.0 (4.9) <.001 
Vancomycin concentration, categorical, n (%)      .003 
 <5 µg/mL 61 (38) 189 (32) 34 (22) 35 (19) 319 (29)  
 5–10 µg/mL 71 (44) 274 (46) 75 (49) 88 (48) 508 (46)  
 10–15 µg/mL 25 (15) 93 (16) 29 (19) 39 (21) 186 (17)  
 15–20 µg/mL 3 (2) 29 (5) 9 (6) 17 (9) 58 (5)  
 >20 µg/mL 2 (1) 14 (2) 6 (4) 6 (3) 28 (3)  
Underweight, n = 162Normal wt, n = 599Overweight, n = 153Obese, n = 185All, n =1099P
Age, y, mean (SD) 7.7 (5.0) 8.6 (5.4) 10.3 (5.5) 10.5 (5.1) 9.0 (5.4) <.001 
Girls, n (%) 65 (40) 274 (45) 81 (53) 70 (38) 490 (45) .03 
Dosing regimen, n (%)      .03 
 every 6 h 45 (28) 183 (31) 41 (27) 39 (21) 308 (28)  
 every 8 h 110 (68) 379 (63) 95 (62) 127 (69) 711 (65)  
 every 12 h 7 (4) 37 (6) 17 (11) 19 (10) 80 (7)  
Vancomycin dose in mg/kg/d, mean (SD)       
 By TBW 50 (11)a 48 (11)a 45 (10)b 42 (12)c 47 (11) <.001 
 By IBW 41 (9)a 49 (12)b 57 (13)c 65 (19)d 52 (15) <.001 
 By ABW 44 (10)a 49 (11)b 51 (11)c 53 (14)c 49 (12) <.001 
 By BSA 1179 (325)a 1317 (351)b 1437 (340)c 1483 (378)c 1341 (363) <.001 
 By AW 103 (28)a 108 (27)a 111 (25)a 108 (27)a 107 (27) .05 
Vancomycin concentration, mean (SD) 7.0 (3.9)a 7.7 (4.8)a,b 8.6 (5.3)b,c 9.2 (5.3)c 8.0 (4.9) <.001 
Vancomycin concentration, categorical, n (%)      .003 
 <5 µg/mL 61 (38) 189 (32) 34 (22) 35 (19) 319 (29)  
 5–10 µg/mL 71 (44) 274 (46) 75 (49) 88 (48) 508 (46)  
 10–15 µg/mL 25 (15) 93 (16) 29 (19) 39 (21) 186 (17)  
 15–20 µg/mL 3 (2) 29 (5) 9 (6) 17 (9) 58 (5)  
 >20 µg/mL 2 (1) 14 (2) 6 (4) 6 (3) 28 (3)  

Superscript letters indicate pairwise differences for continuous variables, with different letters indicating a significant difference at the P < .05 level.

We found minor but significant differences in dosing regimen by adiposity (Table 1). The majority of patients received dosing every 8 hours (n = 711, 65%). Examining the difference in vancomycin dose by TBW, children with overweight or obesity received a lower mean dose per kilogram than their normal weight or underweight peers (Table 1). When converted to ideal or ABW, children with overweight or obesity received a much higher per kilogram dose (Table 1). Of the 204 children who received ≥3 g/day, 80 (39%) had obesity, 44 (22%) had overweight, and 80 (39%) had a normal weight; 192 (94%) were adolescents.

The mean trough concentration was 19% and 31% higher in the group with obesity than in the normal and underweight weight groups, respectively (9.2 vs 7.7 and 7.0; P < .001) (Table 1). Notably, a significant minority (29%) of vancomycin trough concentrations obtained at the appropriate time were <5 µg/mL in this sample, with a larger proportion of these occurring in the normal weight (32%) or underweight (38%) groups. Additionally, 46% of all patients had vancomycin trough concentrations between 5 and 10 µg/mL. Hence, 75% of patients had initial vancomycin concentrations in the subtherapeutic range. Children with obesity had the highest proportion of troughs in the 15- to 20-µg/mL category (9%), and the number of children with an initial trough concentration of >20 µg/mL was low (3% overall, 1%–4% range) (Table 1).

We did a sensitivity analysis examining only children who received a dose of vancomycin between 55 and 65 mg/kg per day of TBW (n = 245), a small range around the currently recommended 60-mg/kg per day dosing. We found the same pattern of increasing vancomycin trough levels by weight category with children with obesity having a mean level of 12.0 (SD 6.4), those with overweight 10.2 (SD 5.4), normal weight 9.2 (SD 5.9), and underweight 8.3 (SD 3.9) (P = .03, analysis of variance).

The mean AUC likewise trended toward an increase with increasing weight. Children with obesity had a mean AUC of 409 μg-hr/mL (SD =108) versus 384 μg-hr/mL (SD =111) for children at a normal weight (Table 2). However, the group with obesity had a mean AUC level only 4% higher than that of the normal weight group. The difference in mean AUC difference only reached statistical significance in comparing those with obesity versus underweight. Children with obesity had a higher proportion (44%), with an AUC within 400 to 600 μg-hr/mL, the target range.

TABLE 2

Vancomycin Pharmacokinetics by Weight Status

AUC Data (2 Centers Only)Underweight, n = 109Normal wt, n = 408Overweight, n = 90Obese, n = 122All, n = 729P
AUC, mean (SD) 343 (94)a 384 (108)b 396 (119)b 409 (108)b 383 (109) <.001 
AUC categories, n (%)      <.001 
 <400 90 (83) 254 (63) 52 (58) 60 (49) 456 (63)  
 400–600 17 (16) 135 (33) 34 (38) 55 (45) 241 (33)  
 >600 2 (2) 16 (4) 4 (4) 7 (6) 29 (4)  
AUC Data (2 Centers Only)Underweight, n = 109Normal wt, n = 408Overweight, n = 90Obese, n = 122All, n = 729P
AUC, mean (SD) 343 (94)a 384 (108)b 396 (119)b 409 (108)b 383 (109) <.001 
AUC categories, n (%)      <.001 
 <400 90 (83) 254 (63) 52 (58) 60 (49) 456 (63)  
 400–600 17 (16) 135 (33) 34 (38) 55 (45) 241 (33)  
 >600 2 (2) 16 (4) 4 (4) 7 (6) 29 (4)  

Superscript letters indicate pairwise differences for continuous variables, with different letters indicating a significant difference at the P < .05 level.

We examined the association of vancomycin dose using each weight measure (Table 3) (TBW, ABW, IBW, BSA, and AW) with the outcomes of ratio of AUC to MIC and vancomycin trough concentrations. Each model was adjusted for age, sex, and dosing frequency.

TABLE 3

Multivariable Linear Regression Analysis of Predictors of Vancomycin Concentration (Continuous) or AUC and/or MIC

Model 1 (TBW)Model 2 (ABW)Model 3 (IBW)Model 4 (BSA)Model 5 (AW)
AUC/MIC      
 Normal wt      
  Wt metric variable, β (95% CI) 8.72 (8.12 to 9.31)* 8.10 (7.52 to 8.67)* 7.07 (6.50 to 7.65)* 0.32 (0.31 to 0.34)* 4.01 (3.80 to 4.22)* 
  R2 0.78 0.77 0.73 0.85 0.85 
 Overweight and obese      
  Wt metric variable, β (95% CI) 9.09 (8.04 to 10.14)* 7.07 (6.19 to 7.95)* 2.78 (2.12 to 3.44)* 0.27 (0.25 to 0.30)* 3.76 (3.43 to 4.08)* 
  R2 0.70 0.67 0.46 0.77 0.79 
 Underweight      
  Wt metric variable, β (95% CI) 7.96 (7.11 to 8.81)* 8.70 (7.67 to 9.73)* 8.87 (7.70 to 10.03)* 0.33 (0.30 to 0.37)* 3.81 (3.45 to 4.17)* 
  R2 0.78 0.75 0.71 0.82 0.82 
Vancomycin concentration      
 Normal wt      
  Wt metric variable, β (95% CI) 0.10 (0.06 to 0.14)* 0.11 (0.07 to 0.15)* 0.11 (0.07 to 0.14)* 0.00 (0.00 to 0.00)* 0.03 (0.02 to 0.05)* 
  R2 0.21 0.22 0.23 0.20 0.20 
 Overweight and obese      
  Wt metric variable, β (95% CI) 0.05 (−0.02 to 0.12) 0.06 (0.00 to 0.12)* 0.04 (0.01 to 0.08)* 0.00 (−0.00 to 0.00) 0.02 (−0.01 to 0.04) 
  R2 0.09 0.09 0.10 0.09 0.09 
 Underweight      
  Wt metric variable, β (95% CI) 0.07 (0.00 to 0.13)* 0.07 (−0.00 to 0.14) 0.07 (−0.01 to 0.14) 0.00 (0.00 to 0.01)* 0.03 (0.00 to 0.06)* 
  R2 0.16 0.16 0.14 0.15 0.16 
Model 1 (TBW)Model 2 (ABW)Model 3 (IBW)Model 4 (BSA)Model 5 (AW)
AUC/MIC      
 Normal wt      
  Wt metric variable, β (95% CI) 8.72 (8.12 to 9.31)* 8.10 (7.52 to 8.67)* 7.07 (6.50 to 7.65)* 0.32 (0.31 to 0.34)* 4.01 (3.80 to 4.22)* 
  R2 0.78 0.77 0.73 0.85 0.85 
 Overweight and obese      
  Wt metric variable, β (95% CI) 9.09 (8.04 to 10.14)* 7.07 (6.19 to 7.95)* 2.78 (2.12 to 3.44)* 0.27 (0.25 to 0.30)* 3.76 (3.43 to 4.08)* 
  R2 0.70 0.67 0.46 0.77 0.79 
 Underweight      
  Wt metric variable, β (95% CI) 7.96 (7.11 to 8.81)* 8.70 (7.67 to 9.73)* 8.87 (7.70 to 10.03)* 0.33 (0.30 to 0.37)* 3.81 (3.45 to 4.17)* 
  R2 0.78 0.75 0.71 0.82 0.82 
Vancomycin concentration      
 Normal wt      
  Wt metric variable, β (95% CI) 0.10 (0.06 to 0.14)* 0.11 (0.07 to 0.15)* 0.11 (0.07 to 0.14)* 0.00 (0.00 to 0.00)* 0.03 (0.02 to 0.05)* 
  R2 0.21 0.22 0.23 0.20 0.20 
 Overweight and obese      
  Wt metric variable, β (95% CI) 0.05 (−0.02 to 0.12) 0.06 (0.00 to 0.12)* 0.04 (0.01 to 0.08)* 0.00 (−0.00 to 0.00) 0.02 (−0.01 to 0.04) 
  R2 0.09 0.09 0.10 0.09 0.09 
 Underweight      
  Wt metric variable, β (95% CI) 0.07 (0.00 to 0.13)* 0.07 (−0.00 to 0.14) 0.07 (−0.01 to 0.14) 0.00 (0.00 to 0.01)* 0.03 (0.00 to 0.06)* 
  R2 0.16 0.16 0.14 0.15 0.16 

CI, confidence interval.

*

Significant at P < .05 in fully adjusted model for the main effect of the wt parameter on the outcome; all models adjusted for age, sex, and dosing frequency.

Regardless of weight group (underweight, normal weight, overweight, and obese), all of the above weight measures had a poor fit with vancomycin trough data. R2 ranged from 0.09 to 0.23. Despite the poor fit, there was a positive correlation between most weight measures and vancomycin trough levels. The correlation did not reach significance in TBW, BSA, and AW in the overweight and obese group and in ABW and IBW in the underweight group.

On the other hand, weight measures correlated well with AUC. All weight measures across all weight groups had a significant positive correlation. R2 mostly ranged from 0.67 to 0.85. The only measure with a fit below this range was IBW in the overweight and obese group, in which R2 was 0.46.

We also conducted multivariable logistic regression to examine the association between dose and whether the ratio of AUC to MIC was in the target range (400–600 μg-hr/mL). A separate model was created for each weight measure: TBW, ABW, IBW, BSA, and AW. All of the models significantly predicted the ratio of AUC to MIC in the target range after adjusting for age, sex, and dosing interval in children, regardless of weight (Fig 1). The model with IBW as the predictor for children with overweight or obesity had the lowest accuracy (71%).

FIGURE 1

Graphical representation of multivariable models revealing odds ratios for different weight metrics as independent variables and the outcome of vancomycin ratio of AUC to MIC within the target range of 400 to 600 μg-hr/mL.

FIGURE 1

Graphical representation of multivariable models revealing odds ratios for different weight metrics as independent variables and the outcome of vancomycin ratio of AUC to MIC within the target range of 400 to 600 μg-hr/mL.

Close modal

In this multicenter study using clinical data from 3 large, academic hospitals, we found significant differences in vancomycin trough levels by adiposity among children. Our findings are consistent with the larger studies completed to date, in which researchers found differences in serum vancomycin concentrations and pharmacokinetics by adiposity.12,16  Although 3 other studies with sample sizes of <100 overall found no significant differences in vancomycin trough levels,13,14,24,25  the current study had >1000 subjects and was powered to detect a small difference in vancomycin drug levels. Whether this small difference is clinically significant is unclear: vancomycin drug levels are not a clinical outcome and have not been adequately assessed as a surrogate for a clinical outcome.17  Nonetheless, our study reveals significant limitations in trough-based vancomycin dosing strategies, specifically, in that children with overweight or obesity have higher serum vancomycin troughs despite receiving lower TBW dosing and, conversely, that normal and underweight children have lower serum vancomycin troughs despite receiving higher TBW dosing. Furthermore, for all weight groups, none of the multiple weight measures tested had a good fit at predicting vancomycin trough levels. These results validate many pediatric hospitalists’ experience of frequently adjusting vancomycin dosing and rechecking levels to achieve goal trough levels in children of all weight groups.

In contrast, all of the weight metrics performed much better at predicting AUC. In fact, every weight metric performed well across every weight group, with the exception of IBW in the overweight and obese group. AUC is much better targeted by using weight-based dosing than are trough levels. In adults, robust population data sets have been developed that also incorporate creatinine levels and Bayesian modeling to allow even more accurate empirical AUC-, goal-based dosing. In children, these data sets and models are not yet available. However, using TBW or AW, AUC can be predicted with ∼80% accuracy across all weight groups.

These results validate recent recommendations for using ratio of AUC to MIC–based dosing in the 2020 Infectious Disease Society of America “Therapeutic Monitoring of Vancomycin for Serious Methicillin-Resistant Staphylococcus aureus Infections Guideline Vancomycin” guideline (jointly drafted with the American Society of Health-System Pharmacists, the Pediatric Infectious Diseases Society, and the Society of Infectious Diseases).17  Ratio of AUC to MIC is a measure of area under the total concentration time curve over a 24-hour period divided by the MIC of the bacteria. Assuming MIC is ≤1 μg/mL, as is typically the case for methicillin-resistant Staphylococcus aureus, AUC is recommended to be targeted to 400 to 600 μg-hr/mL. To achieve these levels, the guideline suggests dosing 60 to 80 mg/kg per day (using TBW), divided every 6 hours. Because of decreased vancomycin clearance in adolescents compared with younger children, 60 to 70 mg/kg per day is recommended for adolescents aged ≥12 years. The guideline notes that children with obesity have increased vancomycin levels compared with children without obesity for a given dose per kilogram of TBW, which is consistent with our results. The guideline recommends 60 mg/kg per day dosed by TBW in children with obesity. As with children without obesity, children with obesity aged <12 years have higher vancomycin clearance and may require higher dosing than adolescents. In addition, because of their larger total volume of distribution, children with obesity may benefit from a loading dose to achieve steady state faster. The 2020 guidelines suggest consideration of a 20-mg/kg loading dose on the basis of TBW. In our study, we did not investigate the effects of a loading dose, but we do support using TBW in empirical dosing for children with obesity.

In clinical practice, AUC can be accurately determined from 2 measurements of vancomycin levels (eg, 1 hour after completion of the first dose and before the second dose). Online calculators can be used to calculate AUC after entering the timing of the dose and levels. AUC can also be estimated by using 1 trough level and the patient’s creatinine clearance. Well-designed calculators are freely available online for adults and can be applied in children by using simple adaptations, such as manually entering the creatinine clearance. MIC can be assumed to be 1 for methicillin-resistant S aureus, or if known, the actual value can be used.

Our finding that the majority of patients, regardless of weight category, had subtherapeutic initial vancomycin exposure underscores the need for a reliable way to empirically dose vancomycin and measure its exposure. Until more robust calculators are widely available in pediatrics, calculating AUC from patient levels can be a somewhat tedious task. However, it is a relatively easy calculation that inpatient pharmacists should have no trouble performing. Furthermore, our results support the reliability of using TBW to predict AUC exposure, with good accuracy, in children with and without obesity alike. Although there is some adult data to suggest that AUC goal-based dosing leads to improved clinical outcomes, pediatric prospective data on clinical outcomes are lacking and would be a helpful next step.17 

A key limitation to this study is that this is a retrospective study. Prospective studies could allow for better ascertainment and definitions of clinical treatment failure. Secondly, this study included data from 3 academic medical institutions that differed in frequency of vancomycin dosing (ranging from every 6 hours to every 8 hours). However, we ensured that patients had received 3 identical doses before the trough by individual data reviews. We were also unable to assess nephrotoxicity or clinical treatment failure. It is unlikely that significant nephrotoxicity existed in this sample, given the general trend toward underexposure. Additionally, the results of this study are not generalizable to children with decreased renal function because they were not included.

Our study revealed a difference in vancomycin levels and ratio of AUC to MIC by adiposity. Children with overweight or obesity have higher serum vancomycin levels despite receiving lower TBW dosing, and conversely, normal and underweight children have lower serum vancomycin levels despite receiving higher TBW dosing. Modeling of alternative weight-based dosing strategies did not provide a better alternative to TBW dosing for predicting vancomycin trough levels: all were poor predictors of level. However, multiple weight metrics, including TBW, were reliable predictors of AUC in children of all weight groups. Optimal ratio of AUC to MIC in common pediatric infections remains largely unknown. Robust prospective trials in which researchers evaluate dosing regimens and clinical outcomes in children of all weight categories are needed for vancomycin and other commonly used inpatient drugs.

FUNDING: The National Institutes of Health provided funding to Dr Foster via grant K23 DK109199. The funder did not participate in the work. Funded by the National Institutes of Health (NIH).

Dr Khare assisted with the study design, designed the local data collection, collected data, and drafted and revised the manuscript; Dr Haag and Mr Kneese collected and analyzed data and drafted and revised portions of the manuscript; Drs Austin and Perlman provided input on the analyses and drafted and revised portions of the manuscript; Drs Azim and Orsi drafted and revised multiple versions of the manuscript; Dr Foster conceptualized the study, designed data collection, conducted data analyses, and drafted and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

1.
Withrow
D
,
Alter
DA
.
The economic burden of obesity worldwide: a systematic review of the direct costs of obesity
.
Obes Rev
.
2011
;
12
(
2
):
131
141
2.
Thorpe
KE
,
Florence
CS
,
Howard
DH
,
Joski
P
.
The impact of obesity on rising medical spending
.
Health Aff (Millwood)
.
2004
;(
Suppl Web Exclusives
):
W4-480
486
3.
Ebbeling
CB
,
Pawlak
DB
,
Ludwig
DS
.
Childhood obesity: public-health crisis, common sense cure
.
Lancet
.
2002
;
360
(
9331
):
473
482
4.
Hales
CM
,
Carroll
MD
,
Fryar
CD
,
Ogden
CL
.
Prevalence of obesity among adults and youth: United States, 2015-2016
.
NCHS Data Brief
.
2017
;(
288
):
1
8
5.
Hales
CM
,
Fryar
CD
,
Carroll
MD
,
Freedman
DS
,
Ogden
CL
.
Trends in obesity and severe obesity prevalence in US youth and adults by sex and age, 2007-2008 to 2015-2016
.
JAMA
.
2018
;
319
(
16
):
1723
1725
6.
Burke
CN
,
Voepel-Lewis
T
,
Wagner
D
, et al
.
A retrospective description of anesthetic medication dosing in overweight and obese children
.
Paediatr Anaesth
.
2014
;
24
(
8
):
857
862
7.
Wells
JC
,
Fewtrell
MS
,
Williams
JE
,
Haroun
D
,
Lawson
MS
,
Cole
TJ
.
Body composition in normal weight, overweight and obese children: matched case-control analyses of total and regional tissue masses, and body composition trends in relation to relative weight
.
Int J Obes
.
2006
;
30
(
10
):
1506
1513
8.
Harskamp-van Ginkel
MW
,
Hill
KD
,
Becker
KC
, et al
;
Best Pharmaceuticals for Children Act–Pediatric Trials Network Administrative Core Committee
.
Drug dosing and pharmacokinetics in children with obesity: a systematic review
.
JAMA Pediatr
.
2015
;
169
(
7
):
678
685
9.
Durham
SH
,
Simmons
ML
,
Mulherin
DW
,
Foland
JA
.
An evaluation of vancomycin dosing for complicated infections in pediatric patients
.
Hosp Pediatr
.
2015
;
5
(
5
):
276
281
10.
Camaione
L
,
Elliott
K
,
Mitchell-Van Steele
A
,
Lomaestro
B
,
Pai
MP
.
Vancomycin dosing in children and young adults: back to the drawing board
.
Pharmacotherapy
.
2013
;
33
(
12
):
1278
1287
11.
Matson
KL
,
Shaffer
CL
,
Beck
GL
,
Simonsen
KA
.
Assessment of initial serum vancomycin trough concentrations and their association with initial empirical weight-based vancomycin dosing and development of nephrotoxicity in children: a multicenter retrospective study
.
Pharmacotherapy
.
2015
;
35
(
3
):
337
343
12.
Heble
DE
 Jr
,
McPherson
C
,
Nelson
MP
,
Hunstad
DA
.
Vancomycin trough concentrations in overweight or obese pediatric patients
.
Pharmacotherapy
.
2013
;
33
(
12
):
1273
1277
13.
Nassar
L
,
Hadad
S
,
Gefen
A
, et al
.
Prospective evaluation of the dosing regimen of vancomycin in children of different weight categories
.
Curr Drug Saf
.
2012
;
7
(
5
):
375
381
14.
Moffett
BS
,
Kim
S
,
Edwards
MS
.
Vancomycin dosing in obese pediatric patients
.
Clin Pediatr (Phila)
.
2011
;
50
(
5
):
442
446
15.
Miller
M
,
Miller
JL
,
Hagemann
TM
,
Harrison
D
,
Chavez-Bueno
S
,
Johnson
PN
.
Vancomycin dosage in overweight and obese children
.
Am J Health Syst Pharm
.
2011
;
68
(
21
):
2062
2068
16.
Le
J
,
Capparelli
EV
,
Wahid
U
, et al
.
Bayesian estimation of vancomycin pharmacokinetics in obese children: matched case-control study
.
Clin Ther
.
2015
;
37
(
6
):
1340
1351
17.
Rybak
MJ
,
Le
J
,
Lodise
T
, et al
.
Executive summary: therapeutic monitoring of vancomycin for serious methicillin-resistant staphylococcus aureus infections: a revised consensus guideline and review of the American Society of Health-System Pharmacists, the Infectious Diseases Society of America, the Pediatric Infectious Diseases Society, and the Society of Infectious Diseases Pharmacists
.
J Pediatric Infect Dis Soc
.
2020
;
9
(
3
):
281
284
18.
Kuczmarski
RJ
,
Ogden
CL
,
Guo
SS
, et al
.
2000 CDC growth charts for the United States: methods and development
.
Vital Health Stat 11
.
2002
;(
246
):
1
190
19.
Mosteller
RD
.
Simplified calculation of body-surface area
.
N Engl J Med
.
1987
;
317
(
17
):
1098
20.
Traub
SL
,
Kichen
L
.
Estimating ideal body mass in children
.
Am J Hosp Pharm
.
1983
;
40
(
1
):
107
110
21.
Green
B
,
Duffull
SB
.
What is the best size descriptor to use for pharmacokinetic studies in the obese?
Br J Clin Pharmacol
.
2004
;
58
(
2
):
119
133
22.
Le
J
,
Bradley
JS
,
Murray
W
, et al
.
Improved vancomycin dosing in children using area under the curve exposure
.
Pediatr Infect Dis J
.
2013
;
32
(
4
):
e155
e163
23.
Madigan
T
,
Sieve
RM
,
Graner
KK
,
Banerjee
R
.
The effect of age and weight on vancomycin serum trough concentrations in pediatric patients
.
Pharmacotherapy
.
2013
;
33
(
12
):
1264
1272
24.
Eiland
LS
,
Sonawane
KB
.
Vancomycin dosing in healthy-weight, overweight, and obese pediatric patients
.
J Pediatr Pharmacol Ther
.
2014
;
19
(
3
):
182
188
25.
Patel
J
,
Lucas
CJ
,
Ryan
J
,
Jenkins
M
,
Martin
JH
.
Vancomycin therapeutic drug monitoring in paediatrics
.
J Paediatr Child Health
.
2020
;
56
(
4
):
563
570

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest relevant to this article to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.