1969. Comparison of AUC-guided vs trough-guided vancomycin dosing in adult patients.
Session: Poster Abstract Session: Antimicrobial Pharmacokinetics and Pharmacodynamics
Saturday, October 29, 2016
Room: Poster Hall
Background: IDSA guidelines suggest vancomycin 40-60 mg/kg day and troughs of 10-20 mg/L. We hypothesized that Bayesian-controlled dosing with an AUC target would permit lower doses and troughs.

Methods: This was a single center, prospective study of inpatient adults receiving IV vancomycin with ≥1 trough prior to enrollment. Cohorts were as follows. Year 1: Predict vancomycin levels using a non-parametric population model and the BestDose software, with no dose control. Year 2: BestDose control of dosing to achieve AUC:MIC ≥400 (max AUC 800), with unknown MIC set to 1 mg/L. All prior troughs generated each subject's Bayesian posterior model for each dose calculation. Year 3: Use BestDose's MMopt algorithm to obtain individually, optimally timed samples for dosing control. We calculated the prediction bias of each level by excluding it (i.e. using all prior levels in a subject) when generating the Bayesian posterior.

Results: In cohorts (years) 1, 2 and 3 there were 83, 90 and 70 evaluable subjects. The median (range) age was 50 (18-93) years, with 62% males. Weight was 76 (15-194) kg. Age (P=0.10), sex (P=0.36) and weight (P=0.63) were similar across years. Planned MMopt sample times in cohort 3 were 10 peaks, 29 troughs, and 30 others.


1: Pred


3:MMopt Pred+Control


Prediction %bias, median (IQR)a

-9% (-26 to 12)

-2% (-18 to 9)


Daily dose mg/kg, median (IQR)

23.3 (17.0-28.9)

21.8 (17.7-26.6)

22.6 (16.3-27.5)


%Trough therapeutic (N)b

30% (112)

24% (107)

15% (29)


%AUC therapeutic (N)c

80% (112)

74% (107)

82% (70)


%Trough >20 mg/L (N)

21% (112)

8% (107)

11% (29)


%Nephrotoxicity (N)d

6% (83)

0% (90)

1% (70)


%Treatment Failure (N)e

1% (83)

0% (90)

0% (70)


a %bias = (pred-obs)/obs, IQR=interquartile range

b By infection and guidelines

c AUC:MIC ≥ 400 or ≥300 and responding and AUC < 800

d Creatinine rise >0.5 mg/dL or >50% from baseline

e Relapse in resolved symptoms ≤72 hours after stopping vancomycin

Conclusion: Our institution underdoses vancomycin with 70% troughs sub-therapeutic by guidelines. However by AUC, 80% are therapeutic. BestDose Bayesian control nevertheless significantly reduced the number of troughs >20 mg/L and nephrotoxicity. Hospitals with more aggressive dosing would see even greater benefit.

Lauren Kato, Pharm.D.1, Gilmer Youn, MD1, Emi Minejima, PharmD2 and Michael Neely, MD3, (1)University of Southern California, Los Angeles, CA, (2)Pharmacy, University of Southern California, Los Angeles, CA, (3)Division of Pediatric Infectious Diseases, Childrens Hospital Los Angeles, Los Angeles, CA


L. Kato, None

G. Youn, None

E. Minejima, None

M. Neely, None

Findings in the abstracts are embargoed until 12:01 a.m. CDT, Wednesday Oct. 26th with the exception of research findings presented at the IDWeek press conferences.