Statistical Methods for Validation in Laboratory Environments

Validation is widely used in laboratory situations to confirm that a method or piece of equipment is functioning correctly and/or providing the necessary improvements. Statistical methods give the evidence required by regulators to support any changes. The use of hypothesis testing to decide whether to accept or reject a change and applying confidence intervals are particularly valuable.

Typical Validation projects include:

  • In the manufacturing sector, the price of a raw key material increased substantially. Applying statistical methods to laboratory results showed that the quality of the same raw materials obtained from a new supplier were equivalent with 99% confidence. A change of supplier resulted in substantial cost savings.
  • In a laboratory a new procedure for testing was developed. It was quicker and cheaper than the current procedure – but did it provide the same reliable analytical results? In the validation, statistical methods showed that the new method was equivalent to the previous method and actually reduced the analytical variability. This produced savings and also increased confidence in results.

In Validation, statistical methods can be used to answer questions such as:

  • Is raw material from a new supplier equivalent to the current supplier?
  • Are there differences in analytical results between 2 laboratories?
  • Does the new procedure reduce variability?
  • Is this batch of bulk material homogeneous?

The course includes a range of statistical hypothesis tests – t-test, F-test, equivalence testing, one-way analysis of variance – and confidence intervals showing their application to validation.

Both our standard and bespoke courses are suitable for those with no prior knowledge of hypothesis testing or for those who require a refresher programme.

For further details contact ISRU.

 

Copyright © ISRU, Newcastle University 1984-2010 | Privacy Policy