Decision Rules During In-Study Phase and Associated Risks

There are two types of risks when making this decision the consumer perspective [26], that is, the risk of validating/accepting an invalid method, and the producer perspective, that is, the risk of invalidating/failing a valid method. The consumer risk is typically more important to control as this can impact the selection of an appropriate dose for the patients or disease diagnoses based on a biomarker depending on the context. The producer risk results in extra cost to the sponsor due to the additional development and optimization that may be necessary if the method is declared invalid. 

Taking a decision based on the measurement error profile with the tolerance intervals as described above provides better control over the consumer risk [27]. If the method is tmly incapable of providing a desired proportion of the results that are close enough to their tme value, it will most likely be declared invalid based on this measurement error profile. 

However, if the analytical bias is relatively high, but the method is truly valid (borderline situation), the producer risk can be quite high. Use of a larger sample size will not reduce this risk. 

6 DECISION RULES DURING IN-STUDY PHASE AND ASSOCIATED RISKS 

Unlike the prestudy validation phase, where the experiments are relatively time consuming and need to be rigorous, validation rules during the in-study phase should be simple and inexpensive. An in-study rule that is largely accepted in the bioanalytical community is the 4 6 15 rule and is defined in the FDA guidance [3] as ... Atleast four of every six QC samples should be within 15% of their respective nominal values. .. to accept a run. This mle provides a simple and practical guide for routine follow-up. 

---