Article Text

Download PDFPDF
General and Risk Management, Patient Safety (including: medication errors, quality control)
Investigation into the use of rapid (alternative) microbial detection methods to determine bioburden of NHS pharmacy-manufactured non-sterile medicines†
  1. S. Hiom,
  2. C. Talbot,
  3. A. Wilkes,
  4. S.D. Denyer
  1. 1Cardiff & Vale University Health Board, St Marys Pharmaceutical Unit, Cardiff, UK
  2. 2Cardiff University, Welsh School of Pharmacy, Cardiff, UK
  3. 3Cardiff University, School of Medicine, Cardiff, UK


Background Rapid microbiological methods (RMMs) have been shown to improve aseptic NHS QA activities. This study investigates the potential to use these new rapid qualitative technologies for quantitative bioburden testing. The principle of dilution to extinction, extrapolation and an inference on original sample bioburden will be investigated. Non-sterile medicines manufactured within NHS pharmacy are required to meet regulatory standards for microbial bioburden of ≤100 cfu/ml, currently measured by reference microbial methods (eg, total aerobic microbial counts (TAMC) after 5 days incubation at 30°C). RMM can give results within 24 h. Initial studies will be required to determine equivalence or non-inferiority between the reference and alternative methods.

GRP014 Table 1

Equivalence evaluation data for RMM and TAMC

Purpose To investigate the potential to use RMM (BacT/ALERT (bioMerieux)) to determine the bioburden of non-sterile products.

Materials and methods A non-sterile oral product, tranexamic acid 5% w/v mouthwash, was inoculated to 100 cfu/ml with Staphylococcus aureus (n=8). Serial dilutions of 1 in 10 to 1 in 1000 were then prepared (n=10) and split sampling methods used to test TAMC and RMM simultaneously. Equivalence of the alternative method was initially assessed using EN ISO 16140:2003 and subsequently with non-inferiority statistics.

Results Equivalence evaluation data for RMM and TAMC is presented below.

Data analysed using ISO 16140 indicates that the 1 in 100 dilution is used for analysis and then concludes that the two methods are not equivalent. However further analysis with non-inferiority statistics and investigation of 95% CI demonstrates that BacT/ALERT is superior to TAMC (p=0.003, McNemar test).

Conclusions The authors propose that BacT/ALERT can be used as a presence/absence test to determine the dilution to extinction point of a contaminant and thus infer an original sample bioburden.

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.