Qualitative Versus Quantitative Risk Assessment in Good Manufacturing Practice (GMP) Audits: How European-Based Auditors Interpret Probability Under International Council for Harmonisation Guideline (ICH) Q9 (R1)

Abstract Purpose Under ICH Q9 (R1), risk is defined as a function of probability and severity of harm. Although probability may be expressed quantitatively, little empirical evidence exists regarding how EU GMP auditors assess probability in practice. This study explored probability assessment approaches among auditors and the use of qualitative and quantitative methods. Methods A descriptive cross-sectional questionnaire was administered to 52 EU-based GMP auditors with at least two years of auditing experience. Participants (21 in total) completed structured multiple-response questions on probability and optional free-text comment fields addressing probability assessment practices (response rate: 40.4%). Data were analysed using descriptive statistics and cross-tabulation; no inferential analyses were performed. Results Qualitative probability assessment approaches were the most frequently reported method (14/21), while 8/21 respondents reported using quantitative approaches (multiple responses permitted). Probability assessment was commonly described as context-dependent and informed by professional judgement rather than purely numerical models. Qualitative approaches were reported across all experience categories and levels of English proficiency. Participants highlighted limited availability of robust historical data, contextual complexity, and reliance on prior audit experience as key factors influencing probability assessment. Respondents (4/21) suggested that no single clear definition of probability assessment was consistently applied in practice. Conclusion Although some participants reported using quantitative approaches, probability assessment remained predominantly qualitative and experience-based in audits. Participants frequently described using structured risk matrices and categorical probability scales (e.g., High/Medium/Low) to support probability assessments. The findings highlight variability in how probability is operationalised and suggest that additional practical guidance may be useful.