What is Continued Process Verification

# Ongoing Process Verification: A European GMP Inspector's Essential Guide

In the evolving landscape of pharmaceutical manufacturing, **ongoing process verification** (also known as **continued process verification**) remains a cornerstone of compliance. Introduced as part of the **process validation life cycle**, this stage ensures processes stay robust amid real-world variables. Drawing from insights shared by Dr. Franz Schönfeld, a seasoned European GMP inspector, during an ECA Academy course, this post explores what inspectors expect—and how your team can meet those standards. If you're navigating GMP regulations, these perspectives could refine your validation strategy.

## Understanding the Process Validation Life Cycle: The PDAC Framework

At its core, **process validation** follows a structured life cycle, often visualized as the PDAC cycle (Plan-Do-Check-Act). Using an active ingredient synthesis as a practical example, Dr. Schönfeld breaks it down:

- **Plan**: Identify and mitigate potential contamination risks through thorough risk assessments.

- **Do**: Execute validation within proven acceptable ranges, whether via traditional, hybrid, or continuous approaches.

- **Check**: Analyze validation run results against the plan, scrutinizing deviations for root causes.

- **Act**: Approve the process or implement enhancements, such as intensified sampling.

This cyclical model isn't a one-off; it's designed for perpetual improvement, making **ongoing process verification** indispensable for long-term control.

## Detecting Anomalies in the Commercial Phase: Key Triggers

One of the biggest challenges? Spotting anomalies post-launch, when production scales up. Dr. Schönfeld stresses that **continued process verification** is your frontline defense. Watch for these common triggers:

- Shifts in personnel, including leadership changes that could alter oversight.

- Equipment maintenance, repairs, or upgrades that subtly impact performance.

- Manufacturing deviations, no matter how minor.

- Emerging trends in analytical testing results.

- Rising customer complaints signaling quality drifts.

- Evolving regulatory requirements that demand process tweaks.

By leveraging extensive commercial data, **ongoing process verification** not only flags issues early but also drives data-backed refinements—transforming potential pitfalls into opportunities for optimization.

## Moving Beyond Routine Revalidation: A Smarter Approach

Gone are the days of periodic revalidations, especially in non-sterile manufacturing. **Ongoing process verification** steps in as a dynamic alternative, guided by an approved plan that outlines:

- The processes under review.

- Verification scope and frequency, tailored to your process maturity and incremental changes.

- Integration with broader quality systems.

This shift emphasizes proactive monitoring over reactive checks, aligning with EU GMP Annex 15 guidelines.

## Harnessing Statistics and Product Quality Reviews (PQR)

Statistics aren't optional—they're expected. Dr. Schönfeld highlights their role in **product quality reviews (PQR)**, where trend analysis is already commonplace. However, PQR serves products, not processes, and arrives too late in the cycle to substitute for dedicated verification.

Pro tip: Embed statistical tools in your **validation master plan (VMP)** to:

- Mandate change control evaluations for process stability.

- Set thresholds for cumulative changes that prompt deeper reviews.

- Adopt rolling reviews to catch trends in real time.

The result? A **continued process verification report** that clearly states if your process is controlled—and flags any need for extra monitoring or sampling.

## EU vs. US: Harmonized Yet Distinct

Comparing notes with the US FDA's Process Validation Guidance, Dr. Schönfeld notes semantic differences but a unified life cycle philosophy. Both regions prioritize stage 3 (continued verification) for sustained control, ensuring global harmonization without compromising local nuances.

## Common GMP Pitfalls: Lessons from Inspections

From his inspection tenure, Dr. Schönfeld flags recurring deficiencies in **process validation**, including gaps in anomaly detection and inadequate statistical rigor. While specifics vary, the takeaway is clear: Robust documentation and forward-thinking plans are non-negotiable to avoid citations.

## Final Thoughts: Elevate Your GMP Compliance Today

**Ongoing process verification** isn't just a regulatory checkbox—it's a strategic imperative for pharmaceutical excellence. As Dr. Schönfeld's expertise reveals, embracing this stage with data-driven vigilance can safeguard quality, streamline operations, and impress inspectors.

Ready to audit your processes? Review your VMP against these insights and consider ECA resources for deeper dives. Share your experiences with **continued process verification** in the comments—how has it shaped your GMP strategy?

*For more on GMP trends, subscribe to our newsletter and stay ahead of compliance curves.*

Investigation Models

When conducting an investigation, the type of reasoning you use depends on the nature of the problem, the available evidence, and the goals of your inquiry. The two primary types of reasoning—**deductive** and **inductive**—are commonly used, alongside others like **abductive reasoning**. Here's a breakdown to help you choose the appropriate reasoning method for your investigation:

---

### 1. Deductive Reasoning

**Definition**: Deductive reasoning starts with general premises or rules and applies them to specific cases to reach a logically certain conclusion. It moves from the general to the specific.

**When to Use**:

- When you have established facts, rules, or principles that are known to be true.

- When you need a definitive, logically certain conclusion based on those premises.

- Common in investigations where you test a hypothesis against specific evidence.

**Example**:

- **Premise 1**: All employees with keycard access can enter the restricted area.

- **Premise 2**: John has keycard access.

- **Conclusion**: John can enter the restricted area.

- In an investigation, you might use deductive reasoning to confirm whether a suspect fits a specific profile based on established criteria (e.g., access logs, qualifications).

**Strengths**:

- Provides clear, certain conclusions if the premises are true.

- Useful for ruling out possibilities or confirming hypotheses with concrete evidence.

**Weaknesses**:

- Requires accurate and complete premises; if premises are false, the conclusion will be invalid.

- Limited to what is already known—doesn’t generate new hypotheses.

**Use in Investigation**:

- Testing specific hypotheses (e.g., "If the crime required technical expertise, only suspects with that expertise could be involved").

- Analyzing forensic evidence against known standards (e.g., matching DNA profiles).

---

### 2. Inductive Reasoning

**Definition**: Inductive reasoning involves observing specific instances or patterns and drawing general conclusions based on those observations. It moves from the specific to the general.

**When to Use**:

- When you have incomplete information and need to form hypotheses or generalizations.

- When analyzing patterns or trends in evidence to predict future outcomes or identify causes.

- Common in early stages of investigations where you’re gathering evidence and forming theories.

**Example**:

- **Observation**: Three burglaries occurred at night in the same neighborhood, targeting unlocked homes.

- **Conclusion**: Burglaries in this neighborhood are likely to occur at night and target unlocked homes.

- In an investigation, you might use inductive reasoning to identify a pattern in a series of incidents (e.g., a serial offender’s methods).

**Strengths**:

- Allows you to generate hypotheses or theories based on limited data.

- Useful for identifying trends or predicting future events.

**Weaknesses**:

- Conclusions are probabilistic, not certain—new evidence may contradict them.

- Risk of overgeneralization based on insufficient or biased data.

**Use in Investigation**:

- Identifying patterns in criminal behavior (e.g., a thief always strikes on weekends).

- Developing profiles of suspects based on observed behaviors or evidence.

---

### 3. Abductive Reasoning

**Definition**: Abductive reasoning involves forming the most likely explanation based on incomplete or uncertain information. It’s often described as "inference to the best explanation."

**When to Use**:

- When you have incomplete evidence and need to form a plausible hypothesis.

- When deductive certainty isn’t possible, and inductive generalizations are too broad.

- Common in complex investigations where you’re piecing together partial clues.

**Example**:

- **Observation**: A window is broken, and valuables are missing from a home.

- **Conclusion**: The most likely explanation is that a burglar broke the window to gain entry and steal the valuables.

- In an investigation, abductive reasoning helps you propose the most plausible scenario based on available clues, even if not all facts are known.

**Strengths**:

- Useful for generating working hypotheses in the absence of complete data.

- Balances creativity and logic to explain complex or ambiguous situations.

**Weaknesses**:

- Conclusions are tentative and may need revision as new evidence emerges.

- Relies on subjective judgment about what’s "most likely."

**Use in Investigation**:

- Developing theories about a crime when evidence is incomplete (e.g., hypothesizing a motive based on partial clues).

- Reconstructing events based on fragmented evidence (e.g., crime scene analysis).

---

### Comparing the Approaches

| **Reasoning Type** | **Starting Point** | **Conclusion** | **Certainty** | **Best for** |

|--------------------|--------------------|----------------|---------------|--------------|

| **Deductive** | General rules | Specific fact | Certain | Testing hypotheses, confirming facts |

| **Inductive** | Specific observations | General rule | Probabilistic | Identifying patterns, forming theories |

| **Abductive** | Incomplete evidence | Most likely explanation | Tentative | Explaining ambiguous or incomplete data |

---

### Other Reasoning Approaches

In addition to deductive, inductive, and abductive reasoning, you might consider:

- **Causal Reasoning**: Identifying cause-and-effect relationships (e.g., determining what caused a system failure in an investigation).

- **Analogical Reasoning**: Drawing comparisons to similar cases or situations (e.g., comparing a current crime to past cases with similar patterns).

- **Bayesian Reasoning**: Updating probabilities based on new evidence (e.g., adjusting the likelihood of a suspect’s involvement as new clues emerge).

---

### How to Choose the Right Reasoning for Your Investigation

1. **Assess Available Evidence**:

   - If you have clear, reliable facts or rules, use **deductive reasoning** to reach certain conclusions.

   - If you’re observing patterns or trends, use **inductive reasoning** to form hypotheses.

   - If evidence is incomplete or ambiguous, use **abductive reasoning** to propose the most likely explanation.

2. **Consider the Investigation Stage**:

   - **Early Stage**: Use inductive or abductive reasoning to generate hypotheses based on initial evidence or patterns.

   - **Mid-Stage**: Use abductive reasoning to refine hypotheses and deductive reasoning to test them against new evidence.

   - **Late Stage**: Use deductive reasoning to confirm conclusions or finalize the investigation.

3. **Combine Approaches**:

   - Most investigations benefit from combining reasoning types. For example:

     - Use **inductive reasoning** to identify a pattern in crime scenes.

     - Use **abductive reasoning** to hypothesize a suspect’s motive.

     - Use **deductive reasoning** to confirm the hypothesis with forensic evidence.

4. **Account for Uncertainty**:

   - If evidence is limited, lean on abductive reasoning to form working theories, but remain open to revising them.

   - Use Bayesian reasoning to update probabilities as new evidence emerges.

---

### Practical Tips for Applying Reasoning

- **Document Evidence Clearly**: Organize your evidence to distinguish between facts (for deductive reasoning), observations (for inductive reasoning), and incomplete clues (for abductive reasoning).

- **Test Hypotheses**: Use deductive reasoning to test hypotheses generated through inductive or abductive reasoning.

- **Avoid Bias**: Be cautious of confirmation bias (favoring evidence that supports your hypothesis) or overgeneralization in inductive reasoning.

- **Iterate**: Revisit your reasoning as new evidence emerges, refining or discarding hypotheses as needed.

---

### Example Scenario

**Investigation**: A series of thefts in an office building.

- **Inductive Reasoning**: You notice all thefts occur on Fridays when the cleaning crew is present. You hypothesize the thief is someone on the cleaning crew.

- **Abductive Reasoning**: A broken lock is found, and only small items are stolen. You infer the thief is likely an opportunist who entered through the broken lock.

- **Deductive Reasoning**: The company’s access logs show only three employees were present during the thefts. You conclude one of them must be involved, assuming the logs are accurate.

---

By understanding the strengths and limitations of each reasoning type, you can strategically apply them to different phases of your investigation, ensuring a thorough and logical approach to uncovering the truth. If you have a specific investigation in mind, I can help tailor these methods to your case!

Annex 22: Artificial Intelligence in GMP – What Pharma Needs to Know

Annex 22: Artificial Intelligence in GMP – What Pharma Needs to Know

Artificial Intelligence (AI) is rapidly transforming pharmaceutical manufacturing, but regulatory compliance remains critical. To guide the safe and effective use of AI, the European Commission has introduced Annex 22: Artificial Intelligence under EU GMP (Good Manufacturing Practice). This annex complements Annex 11 (Computerised Systems) and outlines expectations for the use of machine learning (ML) models in regulated pharma environments.

Scope of Annex 22

Applies to AI/ML models that are trained on data, not hard-coded.

Covers only static, deterministic models (those that do not adapt once deployed).

Excludes dynamic learning systems, probabilistic models, Generative AI, and Large Language Models (LLMs) for critical GMP applications.

Such advanced AI tools may only be used in non-critical GMP processes with strong human-in-the-loop (HITL) oversight.

Core Principles for AI in GMP

Cross-functional collaboration: QA, process SMEs, IT, and data scientists must work together.

Quality risk management: Risk to patient safety, product quality, and data integrity drives all activities.

Strong documentation: Model development, validation, and testing records are mandatory.

Qualified personnel: Staff must be trained to understand AI risks and responsibilities.

Intended Use & Acceptance Criteria

The intended use of an AI model must be clearly defined and documented.

Input data, variations, and possible limitations should be characterized.

AI must perform as well as or better than the process it replaces.

Acceptance metrics may include accuracy, sensitivity, specificity, precision, and F1 score.

Test Data & Validation

Test data must be representative, independent, and verified.

Pre-processing, exclusions, or synthetic data use must be justified.

Independence is key: training, validation, and testing datasets must remain separate.

All testing requires a formal plan, documentation, and deviation handling.

Explainability & Confidence in AI Decisions

AI systems must provide explainable results using techniques like SHAP, LIME, or heat maps.

Confidence scores should be logged; low-confidence predictions flagged as “undecided” instead of forcing unreliable decisions.

Ongoing Operations & Monitoring

AI models must be under change control and configuration control.

Continuous performance monitoring is required to detect data drift or system changes.

Human review remains essential for AI-assisted decision-making in GMP.

Why Annex 22 Matters

Annex 22 marks a regulatory milestone in pharma AI adoption. It emphasizes:

Safety first: patient safety and product quality cannot be compromised.

Transparency: AI decisions must be explainable.

Accountability: human oversight and strong governance remain non-negotiable.

For pharmaceutical companies, this guidance provides a clear compliance roadmap for AI implementation. While Generative AI and LLMs are not permitted in critical processes, their use in supportive, non-critical applications is acknowledged — as long as there is qualified human oversight.

👉 In short, Annex 22 bridges innovation and regulation, ensuring that pharma can leverage AI responsibly, without risking GMP compliance.

EU Aneex 22 AI/ML Artificial intelligence and Machine learning

Understanding "Human-in-the-Loop" in Pharmaceutical Manufacturing

In the evolving landscape of pharmaceutical production, the integration of artificial intelligence (AI) and machine learning (ML) is transforming how active substances and medicinal products are manufactured. The forthcoming EU GMP Annex 22 outlines stringent guidelines for implementing these technologies, emphasizing validation, training, and oversight. Central to these guidelines is the principle of "Human-in-the-Loop" (HITL), which ensures that human expertise remains integral to AI-driven processes, safeguarding quality and safety.

Defining "Human-in-the-Loop"

At its core, "Human-in-the-Loop" describes a collaborative framework where AI or ML systems provide insights or recommendations, but final decisions rest with qualified human operators. This approach prevents fully autonomous AI operations in regulated environments, particularly where patient safety, product quality, or data integrity could be at stake. Instead, it positions humans as the ultimate validators, reviewing and approving AI outputs to align with established standards.

Notably, the guidelines exclude generative AI and large language models from critical Good Manufacturing Practice (GMP) areas, deeming them unsuitable due to inherent uncertainties. However, in non-critical applications—those with minimal direct impact on core GMP principles—these tools may be employed under strict human supervision.

Implementing Human Oversight in Practice

To operationalize HITL, organizations must clearly define the roles and responsibilities of human operators within the system's intended use. For instance, when an AI model analyzes data to suggest process adjustments, the operator is tasked with evaluating the recommendation, verifying its accuracy, and documenting the rationale for any actions taken. This oversight mirrors traditional manual processes but leverages AI to enhance efficiency.

Key implementation steps include:

Operator Training and Qualification: Personnel must possess relevant expertise and undergo targeted training on the specific AI tools, ensuring they can critically assess model outputs for appropriateness.

Monitoring and Review: Continuous evaluation of AI performance is essential, with operators reviewing outputs on a regular basis. In higher-risk scenarios, every output may require individual testing or approval to mitigate potential errors.

Record-Keeping: Comprehensive documentation of human interventions, model decisions, and validation steps is mandatory, providing an audit trail that demonstrates compliance and accountability.

These measures ensure that AI serves as a supportive tool rather than a replacement for human judgment, maintaining the reliability of pharmaceutical manufacturing workflows.

Benefits of the Human-in-the-Loop Approach

Adopting HITL offers several advantages in a GMP-compliant setting. It balances technological innovation with regulatory caution, allowing AI to streamline routine tasks while preserving human control over nuanced decisions. This reduces operational risks, enhances decision-making accuracy, and upholds the highest standards of product integrity. Ultimately, HITL fosters a culture of responsibility, where AI augments rather than supplants skilled professionals, leading to more robust and traceable manufacturing processes.

Navigating Challenges

While HITL provides essential safeguards, it introduces complexities that organizations must address. Human operators require ongoing training to stay proficient with evolving AI systems, and the need for meticulous monitoring can increase administrative demands. In non-critical areas, validating AI outputs—potentially for each instance—demands resources comparable to traditional methods. By proactively designing systems with clear protocols, pharmaceutical firms can overcome these hurdles, ensuring seamless integration without compromising compliance.

Conclusion

The "Human-in-the-Loop" concept, as highlighted in the draft EU GMP Annex 22, represents a forward-thinking strategy for incorporating AI into pharmaceutical manufacturing. By mandating human oversight, it reinforces the pillars of patient safety, product quality, and data integrity. As the industry advances, embracing HITL will be key to harnessing AI's potential responsibly, paving the way for innovative yet compliant production environments.

EU GMP Chapter 1 Revision: Key Updates and Stakeholder Consultation

The European Commission has announced a targeted revision to Chapter 1 of EudraLex Volume 4, the EU Good Manufacturing Practice (GMP) Guidelines, focusing on the Pharmaceutical Quality System (PQS). This update, developed collaboratively by the European Medicines Agency's (EMA) Good Manufacturing and Distribution Practice Inspectors Working Group (GMDP-IWG) and the Pharmaceutical Inspection Co-operation Scheme (PIC/S), aims to align the guidelines with contemporary advancements in regulatory science and risk management principles.

The primary objective of the revision is to enhance the efficiency of regulatory frameworks by incorporating evidence-based approaches, including the recently updated ICH Q9(R1) guideline on Quality Risk Management  It places greater emphasis on knowledge management (KM) and fostering a proactive quality culture within pharmaceutical manufacturing operations. These changes build on the existing foundation of Chapter 1, which already mandates that a PQS must manage product and process knowledge across all lifecycle stages, ensuring compliance with modern GMP expectations.

A key aspect of the revision is its integration with ICH Q10, the guideline on Pharmaceutical Quality Systems. KM and QRM are positioned as essential enablers for an effective PQS.

Notably, a new paragraph (1.4(xviii)) introduces the application of KM in conjunction with QRM to create an early warning system for identifying and addressing emerging quality and manufacturing risks.

Additionally, paragraph 1.13 underscores the role of knowledge in driving informed decision-making, prompting re-evaluations, and promoting continuous improvement

The proposed amendments address three core areas: Quality Risk Management (QRM), Knowledge Management (KM), and Product Quality Review (PQR)

For QRM, the revision incorporates ICH Q9(R1) through seven new sections that promote risk-based decision-making, a proactive and evidence-based quality culture, scientific rationale in risk assessments, appropriate formality in QRM processes, awareness of subjectivity in evaluations, and the proactive identification of manufacturing risks to prevent shortages and strengthen supply chain resilience.

KM is elevated as a foundational element in GMP-regulated environments, working synergistically with QRM to improve oversight and risk mitigation

. Regarding PQR, the updates provide clearer requirements, such as incorporating trending data from prior reviews for products with limited batch production in a 12-month period, specifying minimum content for PQRs when no batches are manufactured, and permitting adjusted review timelines with proper justification.

### Key Takeaways

- **Consultation Deadline:** Comments on the revised draft must be submitted by December 3, 2025.

- **Core Focus Areas:** The revision targets enhancements in Quality Risk Management (QRM), Knowledge Management (KM), and Product Quality Review (PQR).

- **Integration with ICH Guidelines:** Aligns with ICH Q9(R1) for QRM and ICH Q10 for PQS, emphasizing KM and QRM as enablers for proactive risk oversight.

- **New Provisions:** Includes an early warning system via KM and QRM (paragraph 1.4(xviii)) and the use of knowledge for decision-making and improvement (paragraph 1.13).

- **PQR Clarifications:** Requires trending data from previous reviews for low-volume products, minimum content for zero-batch periods, and justified adjustments to review timelines.

- **Development Collaboration:** Joint effort by EMA GMDP-IWG and PIC/S to ensure science-based, risk-managed regulatory frameworks.