"The statistical approach underlying an assay failure risk assessment can take many forms. Populating the various fields of an FMEA can be supported by experimental designs such as OFAT (One Factor At a Time) or multifactorial designs (MFD), using statistical analysis to determine significant cause-and-effect relationships, their impact on method performance, and to identify optimal and robust conditions. Additionally, approaches such as total error analysis and setting acceptance criteria often rely on prediction interval calculations to quantify uncertainty and ensure reliability.

The risk scores themselves can reflect or be directly derived from experimental results. Ideally, one could aim for a more formalized system of absolute Risk Priority Numbers (RPNs), quantitatively linking experimental data to each scoring element (e.g., severity, occurrence, detectability).

In practice, however, a wide variety of approaches exist for translating knowledge and data into risk scores. These range from qualitative to quantitative methods, and from simple three-point scales to more granular ten-point systems. "

Learn more from Dr. Lars Geurink by viewing his presentation "Analytical Procedure Control Strategy (ACPS)" during the Expert Forum: Overcoming key challenges in understanding and implementing guideline ICH Q14 for Analytical Procedure Development