Article Text

Download PDFPDF
CP-118 Neonatal vancomycin: exploring levels at NHS Tayside, Scotland
  1. K Petrie1,
  2. C O’Brien2,
  3. S Bhushan3,
  4. A Tonna4
  1. 1The Royal Women’s Hospital, Pharmacy, Melbourne, Australia
  2. 2NHS Tayside, Pharmacy, Dundee, UK
  3. 3NHS Tayside, Neonatal Unit, Dundee, UK
  4. 4Robert Gordon University, School of Pharmacy and Life Sciences, Aberdeen, UK


Background Neonatal vancomycin target trough levels are not always achieved, increasing the risk of morbidity and mortality due to treatment failure. The British National Formulary for Children (BNFc) dosing guidance, for which there is no published review of its target level achievement, is used widely. In 2007, the BNFc changed its target recommendations from 5–10 mg/L to 10–15 mg/L, due to increased rates of resistance, but did not change its dosage recommendations. Potentially neonates are being under-dosed post 2007.

Purpose To determine the current trough level achievement for neonatal vancomycin (based on BNFc dosing) and then, by using pharmacokinetic modelling, provide recommendations for an improved dosing regimen.

Materials and methods Retrospective data was used, of all neonates with a vancomycin level analysed between 1 January 2009 and 30 June 2012, from NHS Tayside, Scotland. Medical notes for each patient were reviewed to obtain the gestational and postnatal ages, birth and dosing weights, initial vancomycin dose and level, and serum creatinine. An audit determined target trough achievement. Various published pharmacokinetic population models were evaluated for predicting vancomycin levels in the Tayside sample. The association between prediction errors and clinical characteristics was analysed using multiple regression to assist in formulating a new dosing regimen to improve target achievement.

Results Only 13.3% of initial vancomycin levels lay within the target range of 10–15 mg/L in the Tayside sample (n = 83).

Evaluation of published pharmacokinetic models found that the model based on serum creatinine (Grimsley and Thomson, 1999) most accurately predicted vancomycin levels for the Tayside sample with an average 51.16% unsigned percentage prediction error. Serum creatinine was the only clinical characteristic significantly (p < 0.05) related to the model’s prediction error.

Conclusions Modifying the Grimsley and Thomson (1999) model to account for the systematic bias found enabled a new dosing strategy to be developed based on serum creatinine. This new dosing strategy needs to be prospectively audited to measure potential improvements in target levels.

Reference Grimsley C, Thomson A. Pharmacokinetics and dose requirements of vancomycin in neonates. Archives of Disease in ChildhoodFetal and Neonatal Edition, 1999;81:pp. F221-F227

No conflict of interest.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.