EN
PL(Article based on material published by StatSoft GmbH. Original available at: https://www.statsoft.de/en/vetter-cpv-case-study/)
The pharmaceutical industry is caught between the pressure to innovate, growing regulatory complexity, and the demand to make quality measurable and controllable. In the commercial production of sterile dosage forms, Continued Process Verification (CPV) is a central tool for ensuring control over manufacturing processes—not just at specific points, but throughout the entire product lifecycle. Continuous monitoring is also required by FDA and EU regulations.
1. Introduction: Data-Based Continuous Process Verification in the Pharmaceutical Industry
For the CDMO company Vetter, which fills and packages a wide range of injectable drugs for international clients, CPV had previously meant a significant amount of manual effort. Data from various systems, Excel-based evaluations, and diverse customer requirements led to a system of growing complexity. Increasing demand for data availability and automation, combined with the need to improve efficiency, pushed processes to their limits—this is where the joint project began.
Together with the data science and analytics specialist StatSoft, Vetter developed a fully automated, validatable CPV solution based on the Statistica platform within two years. The goal was not just to make CPV more efficient, but to implement it systematically, in a standardised, and future-proof manner—with a solution that combines technical robustness, regulatory security, and operational usability.
2. Initial Situation and Challenges at Vetter
Vetter supports pharmaceutical manufacturers from process development to clinical and commercial filling, as well as a variety of assembly and packaging options for vials, syringes, and cartridges. The company produces medicines in a highly regulated GMP environment under aseptic conditions, serving a wide variety of customer and product constellations. This is reflected in CPV use: Over 100 CPV programs, spread across 21 cleanrooms, are continuously maintained. The challenge: The related reports were based on a semi-automated, largely manual process in which data from different sources were manually selected, visualized, and compiled.
A particularly time-consuming step was the extensive evaluation of process performance for each individual parameter, which formed the basis for deciding which parameters to include in regular CPV reporting. Manual steps introduced a risk of errors, and the requirement for the four-eyes principle in critical evaluation steps significantly slowed down the entire process.
Our existing CPV processes had grown historically—this led to high manual effort and product-specific approaches. It was clear to us: We need a future-proof, scalable solution that stands up both regulatorily and practically.
Bettina Schroeder, Team Lead QA Trending Process Data, Vetter
This was the starting point for close cooperation between Vetter and StatSoft. The goal was to develop a standardized, validatable, and automated CPV solution that met the diversity of customer/product/cleanroom combinations as well as regulatory requirements—while integrating existing systems and data sources.
3. Project Goals and Approach
Given the many challenges and the need to evaluate CPV-relevant data efficiently and securely, a clear joint objective was defined at the start of the project:
- Automate statistical data evaluation for CPV programs to reduce manual steps.
- Standardize analyses to ensure consistent and traceable evaluations—regardless of product, customer, or cleanroom.
- Ensure the solution’s validity in compliance with GxP guidelines.
- Create a central yet flexible platform enabling key users to define CPV reports independently, without programming knowledge.n
The project was a strategic milestone on the path to a digitally integrated quality management system for us. What mattered most was that we didn’t just get a technical solution—but one that genuinely improved our internal processes and would be sustainable in the long term.
Dr. Philip Hörsch, Director Quality Assurance Process Performance & System Compliance, Vetter
The approach was iterative, collaborative, and technologically ambitious from the outset. The first step was to develop a shared vision for the automated CPV process—from data source through statistical evaluation to automated report creation. The focus was not on a generic standard solution but on a modular, configurable architecture tailored to Vetter’s specific requirements. The system was designed to accommodate future developments (new customer/product/cleanroom combinations).
StatSoft’s development phase lasted about a year of intensive work—with multiple test cycles, close feedback from Vetter, and regular approvals from the project team. In parallel, Vetter prepared for rollout: technical integration, governance models, and validation strategy. Following development, another year was dedicated to implementing and validating the solution in the regulated production environment.
The foundation for success was open, solution-oriented collaboration—characterized by mutual trust, a clear understanding of regulatory requirements, and the ability of both sides to respond flexibly to new insights.
Based on our experience from a prior PoC project, Vetter created a URS tailored for implementation in Statistica. This allowed all customer requirements to be addressed efficiently during the main project. Direct, close, and trusting contact between Vetter’s subject matter experts and our developers ensured continuous optimization, resulting in an optimal solution with efficient use of the purchased standard software and minimal development effort.
Michael Busch, Lead Data Engineer, StatSoft
4. The Solution in Detail: Automated CPV Workspaces with Validated Architecture
The developed solution is based on creating product-, customer-, and cleanroom-specific workflows in Statistica for statistical evaluation of production data—within a validated software architecture aligned with Vetter’s existing system landscape.
Integration of Existing Systems
A key goal was seamless integration with existing data sources, especially Vetter’s SAP-BW system, which serves as the primary source for in-process control (IPC) data. The new solution automatically retrieves this data and structures it according to defined parameters—based on product- and customer-specific configurations.
Product- and Customer-Specific Parameterization
For each product/customer/cleanroom scenario, a dedicated parameter table was stored. This defines which CPPs (Critical Process Parameters) and CQAs (Critical Quality Attributes) are analyzed with which methods (e.g., run charts, control charts, process capability analyses) and over what time periods. This creates high flexibility within a standardized framework—crucial given the wide variety of scenarios.
Statistical Evaluation and Visualization
The automated evaluation uses various statistical methods, including:
- Run charts and modified run charts with moving average
- Control charts (Shewhart)
- Histograms
- Calculation of process capabilities
Evaluations are generated automatically and presented in a clear reporting layout—including placeholders for author commentary. Reports can be generated at the push of a button in a Word template, including all relevant results, graphics, and metadata.
What made working with StatSoft special was their ability to understand our complex environment. It wasn’t just about software—it was about creating a system that fully integrates our internal requirements and regulatory specifications. That’s been achieved completely.
Jonas Portele,Team Lead Operational Excellence, Vetter
5. Results and Added Value: Efficiency, Security, and Scalability
The introduction of the automated CPV solution has had measurable effects at Vetter—both in terms of day-to-day efficiency gains and the quality of reporting processes. The added value is particularly evident in three key areas:
5.1 Time and Resource Savings
Reducing manual effort in CPV report creation was a core project goal—and it was clearly achieved. Automation of data aggregation, analysis, and reporting reduced the resources needed for regular reports by up to 50%. This applies not only to document creation but also to preparatory and coordination steps.
The saved time is now used for qualitative evaluation of results—focusing on what truly matters: in-depth analysis of process behavior, deviations, and optimization opportunities.
5.2 Standardization with High Flexibility
The solution enables a consistent CPV reporting approach across all products, customers, and cleanrooms. This ensures transparency, traceability, and security in internal and external audits.
At the same time, flexibility is preserved through parameter tables that adapt to each scenario—standardization without forced uniformity, a challenging balance in regulated environments.
5.3 Validation and GxP Compliance
The solution is fully embedded in a three-tier system landscape (Development – Qualification – Production) and follows a clearly documented validation process
Responsibility for configuring and transporting parameter tables is strictly separated and subject to role-based access rights—important for data integrity and security. Reports meet all GxP documentation requirements, including audit trails, electronic signatures, and version history..
5.4 User-Friendliness and Enablement
Through a jointly developed key-user model, selected specialists can maintain product- and customer-specific parameters independently—without programming skills, but with clear guidelines. This allows quick reactions to changes and relieves central IT resources.
6. Conclusion and Outlook: From Project to Platform
By completing the CPV project, Vetter has not only implemented a technical solution but initiated lasting change. Automating and standardizing CPV evaluations is a major step toward data-driven quality assurance based on scalability, transparency, and regulatory security.
Success Factors in Retrospect
The project’s success rests on several factors:
- A clearly defined vision with shared responsibility
- A collaborative partnership between subject matter experts, IT, and project management
- Iterative development with room for feedback and adjustments
- Consistent alignment with regulatory requirements and internal processes
This combination made it possible to create a solution that is both technically robust and practically usable—across a complex organizational structure.
Outlook: Expansion and Integration
The project also lays the foundation for future developments: Plans are underway to integrate additional data types into the CPV platform. Integration of data from quality management systems—such as deviation management or CAPA systems—is also conceivable. The goal is an even more comprehensive assessment of process stability and quality across all relevant sources.
Juliane Berek, StatSoft GmbH
Juliane Berek has been exploring for over 15 years how to generate knowledge and value from data. As a Client Partner, she advises StatSoft clients on Data Science, Machine Learning, and Artificial Intelligence.