eduKateSG · Why Science?
Use risk thinking to focus effort where quality failures could matter most to patients and medicine supply
Read FDA’s final May 2023 ICH Q9(R1) guidance through risk assessment, control, communication, review, formality, subjectivity, knowledge, uncertainty, product availability and decision-making.
Pharmaceutical teams make choices every day: which deviation needs urgent escalation, which supplier change deserves deeper study, and which control protects the patient most. Quality risk management provides a disciplined way to focus attention without pretending uncertainty has disappeared.
FDA’s final May 2023 ICH Q9(R1) guidance updates the 2006 framework. The revision addresses subjectivity, appropriate formality, risk-based decision-making and risks to product availability caused by manufacturing quality problems.
Did you know? Q9(R1) says the level of effort, formality and documentation should be commensurate with risk, uncertainty, importance and complexity. Not every question needs the same tool or meeting.
For students, Q9(R1) is structured judgment. Good risk work makes assumptions visible, invites relevant knowledge and connects assessment to action.
Chapter index
Chapters 1–9: May 2023 final guidance, hazards, probability, severity, detectability and uncertainty
- 1. Quality risk management supports patient protection
- 2. The May 2023 Q9(R1) guidance is final
- 3. A clear question prevents unfocused scoring
- 4. Hazard and risk are not the same
- 5. Risk identification gathers plausible failure modes
- 6. Risk analysis estimates likelihood and consequence
- 7. Risk evaluation compares results with criteria
- 8. Subjectivity should be recognised and managed
- 9. Formality should match the situation
Chapters 10–17: formality, subjectivity, risk control, communication, review, product availability and governance
- 10. Risk control asks whether reduction is needed
- 11. Risk communication keeps decisions aligned
- 12. Risk review responds to new knowledge
- 13. Quality risk management supports availability
- 14. Tools should serve the question
- 15. Data quality limits precision
- 16. Governance prevents risk work from becoming theatre
- 17. A complete Q9(R1) receipt makes judgment reproducible
CHAPTER 1 OF 17 · Frame the question
1. Quality risk management supports patient protection
Q9(R1) provides principles and tools for identifying, analysing, evaluating, controlling, communicating and reviewing risks to quality across development, manufacturing, distribution and lifecycle activities.
The scientific value of quality risk management supports patient protection comes from linking knowledge to a controlled decision. Q9(R1) provides principles and tools for identifying, analysing, evaluating, controlling, communicating and reviewing risks to quality across development, manufacturing, distribution and lifecycle activities. Keep this boundary visible: A precise score can still rest on weak evidence or arbitrary scales.
Try this
Build a cause-and-control map for quality risk management supports patient protection. Define the decision and risk question for quality risk management supports patient protection before choosing a tool. Add the evidence and owner. Keep this limit: A precise score can still rest on weak evidence or arbitrary scales.
A strong science note separates observation from interpretation. Define the decision and risk question for quality risk management supports patient protection before choosing a tool. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Define the decision and risk question for quality risk management supports patient protection before choosing a tool. Keep this boundary visible: A precise score can still rest on weak evidence or arbitrary scales.
FDA lists Q9(R1) as final guidance under docket FDA-2022-D-0705, issued by CDER and CBER. It replaces the 2022 draft and targets implementation weaknesses observed since the original Q9.
For the may 2023 q9(r1) guidance is final, a useful model explains both the expected process and the response to variation. FDA lists Q9(R1) as final guidance under docket FDA-2022-D-0705, issued by CDER and CBER. It replaces the 2022 draft and targets implementation weaknesses observed since the original Q9. Preserve this guardrail: Risk acceptance cannot override legal requirements or conceal patient harm.
Try this
Challenge the assumptions behind the may 2023 q9(r1) guidance is final. Re-score the may 2023 q9(r1) guidance is final under plausible alternative assumptions and compare the decision. Test a plausible failure mode. Respect this boundary: Risk acceptance cannot override legal requirements or conceal patient harm.
A strong science note separates observation from interpretation. Re-score the may 2023 q9(r1) guidance is final under plausible alternative assumptions and compare the decision. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Re-score the may 2023 q9(r1) guidance is final under plausible alternative assumptions and compare the decision. Keep this boundary visible: Risk acceptance cannot override legal requirements or conceal patient harm.
The process should define the decision, scope, assumptions, stakeholders and available information. A vague prompt produces long hazard lists that may not help anyone choose an action.
A trustworthy account of a clear question prevents unfocused scoring keeps evidence, assumptions and lifecycle action together. The process should define the decision, scope, assumptions, stakeholders and available information. A vague prompt produces long hazard lists that may not help anyone choose an action. Remember: Detectability is not a substitute for preventing a serious failure.
Try this
Audit the lifecycle record for a clear question prevents unfocused scoring. Assign controls for a clear question prevents unfocused scoring with owners, deadlines and effectiveness measures. Mark what changed and why. Preserve this caution: Detectability is not a substitute for preventing a serious failure.
A strong science note separates observation from interpretation. Assign controls for a clear question prevents unfocused scoring with owners, deadlines and effectiveness measures. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Assign controls for a clear question prevents unfocused scoring with owners, deadlines and effectiveness measures. Keep this boundary visible: Detectability is not a substitute for preventing a serious failure.
A hazard is a potential source of harm; risk combines the probability of occurrence with the severity of that harm. Some methods also consider detectability, but a score should not obscure the underlying evidence.
The scientific value of hazard and risk are not the same comes from linking knowledge to a controlled decision. A hazard is a potential source of harm; risk combines the probability of occurrence with the severity of that harm. Some methods also consider detectability, but a score should not obscure the underlying evidence. Keep this boundary visible: Low observed frequency may reflect poor detection or limited opportunity.
Try this
Build a cause-and-control map for hazard and risk are not the same. Review hazard and risk are not the same after a deviation or supply disruption changes the evidence. Add the evidence and owner. Keep this limit: Low observed frequency may reflect poor detection or limited opportunity.
A strong science note separates observation from interpretation. Review hazard and risk are not the same after a deviation or supply disruption changes the evidence. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Review hazard and risk are not the same after a deviation or supply disruption changes the evidence. Keep this boundary visible: Low observed frequency may reflect poor detection or limited opportunity.
CHAPTER 5 OF 17 · Assess the risk
5. Risk identification gathers plausible failure modes
Process maps, historical data, scientific knowledge, deviations, complaints and expert input can reveal what might go wrong. The search should include conditions outside normal experience without turning every remote possibility into equal priority.
For risk identification gathers plausible failure modes, a useful model explains both the expected process and the response to variation. Process maps, historical data, scientific knowledge, deviations, complaints and expert input can reveal what might go wrong. The search should include conditions outside normal experience without turning every remote possibility into equal priority. Preserve this guardrail: A precise score can still rest on weak evidence or arbitrary scales.
Try this
Challenge the assumptions behind risk identification gathers plausible failure modes. Define the decision and risk question for risk identification gathers plausible failure modes before choosing a tool. Test a plausible failure mode. Respect this boundary: A precise score can still rest on weak evidence or arbitrary scales.
A strong science note separates observation from interpretation. Define the decision and risk question for risk identification gathers plausible failure modes before choosing a tool. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Define the decision and risk question for risk identification gathers plausible failure modes before choosing a tool. Keep this boundary visible: A precise score can still rest on weak evidence or arbitrary scales.
CHAPTER 6 OF 17 · Assess the risk
6. Risk analysis estimates likelihood and consequence
Qualitative categories or quantitative models can be used depending on the decision. Estimates should state data sources, time horizon, population, assumptions and uncertainty.
A trustworthy account of risk analysis estimates likelihood and consequence keeps evidence, assumptions and lifecycle action together. Qualitative categories or quantitative models can be used depending on the decision. Estimates should state data sources, time horizon, population, assumptions and uncertainty. Remember: Risk acceptance cannot override legal requirements or conceal patient harm.
Try this
Audit the lifecycle record for risk analysis estimates likelihood and consequence. Re-score risk analysis estimates likelihood and consequence under plausible alternative assumptions and compare the decision. Mark what changed and why. Preserve this caution: Risk acceptance cannot override legal requirements or conceal patient harm.
A strong science note separates observation from interpretation. Re-score risk analysis estimates likelihood and consequence under plausible alternative assumptions and compare the decision. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Re-score risk analysis estimates likelihood and consequence under plausible alternative assumptions and compare the decision. Keep this boundary visible: Risk acceptance cannot override legal requirements or conceal patient harm.
Teams decide which risks are acceptable, need reduction or require escalation. Criteria should be established before seeing a preferred outcome where practical and should reflect patient impact and regulatory obligations.
The scientific value of risk evaluation compares results with criteria comes from linking knowledge to a controlled decision. Teams decide which risks are acceptable, need reduction or require escalation. Criteria should be established before seeing a preferred outcome where practical and should reflect patient impact and regulatory obligations. Keep this boundary visible: Detectability is not a substitute for preventing a serious failure.
Try this
Build a cause-and-control map for risk evaluation compares results with criteria. Assign controls for risk evaluation compares results with criteria with owners, deadlines and effectiveness measures. Add the evidence and owner. Keep this limit: Detectability is not a substitute for preventing a serious failure.
A strong science note separates observation from interpretation. Assign controls for risk evaluation compares results with criteria with owners, deadlines and effectiveness measures. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Assign controls for risk evaluation compares results with criteria with owners, deadlines and effectiveness measures. Keep this boundary visible: Detectability is not a substitute for preventing a serious failure.
Risk assessments can differ because people frame problems, choose scales and interpret evidence differently. Diverse expertise, calibrated definitions, facilitators and transparent rationale reduce uncontrolled bias.
For subjectivity should be recognised and managed, a useful model explains both the expected process and the response to variation. Risk assessments can differ because people frame problems, choose scales and interpret evidence differently. Diverse expertise, calibrated definitions, facilitators and transparent rationale reduce uncontrolled bias. Preserve this guardrail: Low observed frequency may reflect poor detection or limited opportunity.
Try this
Challenge the assumptions behind subjectivity should be recognised and managed. Review subjectivity should be recognised and managed after a deviation or supply disruption changes the evidence. Test a plausible failure mode. Respect this boundary: Low observed frequency may reflect poor detection or limited opportunity.
A strong science note separates observation from interpretation. Review subjectivity should be recognised and managed after a deviation or supply disruption changes the evidence. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Review subjectivity should be recognised and managed after a deviation or supply disruption changes the evidence. Keep this boundary visible: Low observed frequency may reflect poor detection or limited opportunity.
High uncertainty, complexity, importance or potential harm can justify a structured cross-functional process and detailed record. Routine low-risk decisions may use simpler methods while still retaining enough reasoning to be accountable.
A trustworthy account of formality should match the situation keeps evidence, assumptions and lifecycle action together. High uncertainty, complexity, importance or potential harm can justify a structured cross-functional process and detailed record. Routine low-risk decisions may use simpler methods while still retaining enough reasoning to be accountable. Remember: A precise score can still rest on weak evidence or arbitrary scales.
Try this
Audit the lifecycle record for formality should match the situation. Define the decision and risk question for formality should match the situation before choosing a tool. Mark what changed and why. Preserve this caution: A precise score can still rest on weak evidence or arbitrary scales.
A strong science note separates observation from interpretation. Define the decision and risk question for formality should match the situation before choosing a tool. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Define the decision and risk question for formality should match the situation before choosing a tool. Keep this boundary visible: A precise score can still rest on weak evidence or arbitrary scales.
CHAPTER 10 OF 17 · Control and communicate
10. Risk control asks whether reduction is needed
Controls can avoid, reduce, detect or accept risk. The team should consider whether the residual risk is acceptable and whether a control introduces new risks elsewhere.
The scientific value of risk control asks whether reduction is needed comes from linking knowledge to a controlled decision. Controls can avoid, reduce, detect or accept risk. The team should consider whether the residual risk is acceptable and whether a control introduces new risks elsewhere. Keep this boundary visible: Risk acceptance cannot override legal requirements or conceal patient harm.
Try this
Build a cause-and-control map for risk control asks whether reduction is needed. Re-score risk control asks whether reduction is needed under plausible alternative assumptions and compare the decision. Add the evidence and owner. Keep this limit: Risk acceptance cannot override legal requirements or conceal patient harm.
A strong science note separates observation from interpretation. Re-score risk control asks whether reduction is needed under plausible alternative assumptions and compare the decision. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Re-score risk control asks whether reduction is needed under plausible alternative assumptions and compare the decision. Keep this boundary visible: Risk acceptance cannot override legal requirements or conceal patient harm.
CHAPTER 11 OF 17 · Control and communicate
11. Risk communication keeps decisions aligned
Relevant assumptions, results, actions and residual uncertainty should reach the people who operate, approve or are affected by the decision. Communication continues through handoffs and change control, not just the final report.
For risk communication keeps decisions aligned, a useful model explains both the expected process and the response to variation. Relevant assumptions, results, actions and residual uncertainty should reach the people who operate, approve or are affected by the decision. Communication continues through handoffs and change control, not just the final report. Preserve this guardrail: Detectability is not a substitute for preventing a serious failure.
Try this
Challenge the assumptions behind risk communication keeps decisions aligned. Assign controls for risk communication keeps decisions aligned with owners, deadlines and effectiveness measures. Test a plausible failure mode. Respect this boundary: Detectability is not a substitute for preventing a serious failure.
A strong science note separates observation from interpretation. Assign controls for risk communication keeps decisions aligned with owners, deadlines and effectiveness measures. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Assign controls for risk communication keeps decisions aligned with owners, deadlines and effectiveness measures. Keep this boundary visible: Detectability is not a substitute for preventing a serious failure.
Signals, deviations, trends, audits, complaints, technology and process changes can alter likelihood or consequence. Review frequency should match the speed at which the risk and evidence can change.
A trustworthy account of risk review responds to new knowledge keeps evidence, assumptions and lifecycle action together. Signals, deviations, trends, audits, complaints, technology and process changes can alter likelihood or consequence. Review frequency should match the speed at which the risk and evidence can change. Remember: Low observed frequency may reflect poor detection or limited opportunity.
Try this
Audit the lifecycle record for risk review responds to new knowledge. Review risk review responds to new knowledge after a deviation or supply disruption changes the evidence. Mark what changed and why. Preserve this caution: Low observed frequency may reflect poor detection or limited opportunity.
A strong science note separates observation from interpretation. Review risk review responds to new knowledge after a deviation or supply disruption changes the evidence. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Review risk review responds to new knowledge after a deviation or supply disruption changes the evidence. Keep this boundary visible: Low observed frequency may reflect poor detection or limited opportunity.
CHAPTER 13 OF 17 · Review and improve
13. Quality risk management supports availability
Q9(R1) explicitly considers manufacturing quality issues that can threaten product availability. Supply vulnerability, capacity, recovery time and alternative sources may be relevant when quality failures interrupt access to medicines.
The scientific value of quality risk management supports availability comes from linking knowledge to a controlled decision. Q9(R1) explicitly considers manufacturing quality issues that can threaten product availability. Supply vulnerability, capacity, recovery time and alternative sources may be relevant when quality failures interrupt access to medicines. Keep this boundary visible: A precise score can still rest on weak evidence or arbitrary scales.
Try this
Build a cause-and-control map for quality risk management supports availability. Define the decision and risk question for quality risk management supports availability before choosing a tool. Add the evidence and owner. Keep this limit: A precise score can still rest on weak evidence or arbitrary scales.
A strong science note separates observation from interpretation. Define the decision and risk question for quality risk management supports availability before choosing a tool. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Define the decision and risk question for quality risk management supports availability before choosing a tool. Keep this boundary visible: A precise score can still rest on weak evidence or arbitrary scales.
FMEA, fault trees, hazard analysis, HACCP, ranking and statistical methods each have strengths. The tool is not the objective; a simpler method can be better when it gives a clear, defensible decision.
For tools should serve the question, a useful model explains both the expected process and the response to variation. FMEA, fault trees, hazard analysis, HACCP, ranking and statistical methods each have strengths. The tool is not the objective; a simpler method can be better when it gives a clear, defensible decision. Preserve this guardrail: Risk acceptance cannot override legal requirements or conceal patient harm.
Try this
Challenge the assumptions behind tools should serve the question. Re-score tools should serve the question under plausible alternative assumptions and compare the decision. Test a plausible failure mode. Respect this boundary: Risk acceptance cannot override legal requirements or conceal patient harm.
A strong science note separates observation from interpretation. Re-score tools should serve the question under plausible alternative assumptions and compare the decision. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Re-score tools should serve the question under plausible alternative assumptions and compare the decision. Keep this boundary visible: Risk acceptance cannot override legal requirements or conceal patient harm.
Sparse events, inconsistent coding and changing processes can make numerical rankings look more certain than the evidence. Sensitivity analysis and ranges can show whether the decision survives plausible alternatives.
A trustworthy account of data quality limits precision keeps evidence, assumptions and lifecycle action together. Sparse events, inconsistent coding and changing processes can make numerical rankings look more certain than the evidence. Sensitivity analysis and ranges can show whether the decision survives plausible alternatives. Remember: Detectability is not a substitute for preventing a serious failure.
Try this
Audit the lifecycle record for data quality limits precision. Assign controls for data quality limits precision with owners, deadlines and effectiveness measures. Mark what changed and why. Preserve this caution: Detectability is not a substitute for preventing a serious failure.
A strong science note separates observation from interpretation. Assign controls for data quality limits precision with owners, deadlines and effectiveness measures. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Assign controls for data quality limits precision with owners, deadlines and effectiveness measures. Keep this boundary visible: Detectability is not a substitute for preventing a serious failure.
CHAPTER 16 OF 17 · Review and improve
16. Governance prevents risk work from becoming theatre
Owners, timelines, escalation thresholds, effectiveness checks and management oversight turn assessment into control. Closing an action item is different from showing that risk actually fell.
The scientific value of governance prevents risk work from becoming theatre comes from linking knowledge to a controlled decision. Owners, timelines, escalation thresholds, effectiveness checks and management oversight turn assessment into control. Closing an action item is different from showing that risk actually fell. Keep this boundary visible: Low observed frequency may reflect poor detection or limited opportunity.
Try this
Build a cause-and-control map for governance prevents risk work from becoming theatre. Review governance prevents risk work from becoming theatre after a deviation or supply disruption changes the evidence. Add the evidence and owner. Keep this limit: Low observed frequency may reflect poor detection or limited opportunity.
A strong science note separates observation from interpretation. Review governance prevents risk work from becoming theatre after a deviation or supply disruption changes the evidence. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Review governance prevents risk work from becoming theatre after a deviation or supply disruption changes the evidence. Keep this boundary visible: Low observed frequency may reflect poor detection or limited opportunity.
CHAPTER 17 OF 17 · Review and improve
17. A complete Q9(R1) receipt makes judgment reproducible
The receipt records question, scope, participants, hazards, data, assumptions, scales, uncertainty, analysis, evaluation criteria, controls, residual risk, communications, owners, review triggers and effectiveness evidence. It lets another team understand why the decision was reasonable.
For a complete q9(r1) receipt makes judgment reproducible, a useful model explains both the expected process and the response to variation. The receipt records question, scope, participants, hazards, data, assumptions, scales, uncertainty, analysis, evaluation criteria, controls, residual risk, communications, owners, review triggers and effectiveness evidence. It lets another team understand why the decision was reasonable. Preserve this guardrail: A precise score can still rest on weak evidence or arbitrary scales.
Try this
Challenge the assumptions behind a complete q9(r1) receipt makes judgment reproducible. Define the decision and risk question for a complete q9(r1) receipt makes judgment reproducible before choosing a tool. Test a plausible failure mode. Respect this boundary: A precise score can still rest on weak evidence or arbitrary scales.
A strong science note separates observation from interpretation. Define the decision and risk question for a complete q9(r1) receipt makes judgment reproducible before choosing a tool. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Define the decision and risk question for a complete q9(r1) receipt makes judgment reproducible before choosing a tool. Keep this boundary visible: A precise score can still rest on weak evidence or arbitrary scales.
Q9(R1) risk-management cycle
| Question | Evidence to capture | Boundary to retain |
|---|---|---|
| What can go wrong? | Hazards, causes and affected quality attributes | Frame the decision first |
| How serious and likely? | Evidence, estimates and uncertainty | Scores are not facts |
| What should change? | Avoidance, reduction, detection or acceptance | Check residual and new risks |
| When is it revisited? | Signals, changes, trends and effectiveness | Risk knowledge evolves |
Science learning becomes useful when a student can show where a claim came from, why it applies and where it stops. Continue with the eduKate Science Learning Hub, compare choices in the Education Hub, and connect these habits to career pathways built from subject capabilities.
Q9(R1) brings humility and focus to quality decisions. It asks teams to use the best available knowledge, expose uncertainty, apply effort in proportion to consequence and return to the decision when the world changes.
Primary sources and verification date
Source status checked 26 October 2026. FDA guidance describes the agency’s current thinking and does not replace applicable statutes, regulations or product-specific communication with FDA.
- FDA: ICH Q9(R1) Quality Risk Management — Final guidance, May 2023; docket FDA-2022-D-0705.
- FDA: Full Q9(R1) guidance PDF — Official risk-management principles, formality and tools.
- FDA: Q8, Q9 and Q10 Questions and Answers (R5) — Final Level 2 implementation guidance, May 2026.