Best Practices for Managing the Life-Cycle of Single-Use Systems

Single-use systems (SUS) have become increasingly prevalent in biopharmaceutical manufacturing due to their flexibility, reduced contamination risk, and cost-effectiveness. The thing is, management of the life-cycle of single-use systems becomes critical and is an area organizations can truly screw up by cutting corners. To do it right requires careful collaboration between all stakeholders in the supply chain, from raw material suppliers to end users.

Design and Development

Apply Quality by Design (QbD) principles from the outset by focusing on process understanding and the design space to create controlled and consistent manufacturing processes that result in high-quality, efficacious products. This approach should be applied to SUS design.

ASTM E3051 “Standard guide for specification, design, verification, and application of SUS in pharmaceutical and biopharmaceutical manufacturing” provides an excellent framework for the design process.

Make sure to conduct thorough risk assessments, considering potential failure modes and effects throughout the SUS life-cycle.

Engage end-users early to understand their specific requirements and process constraints. A real mistake in organizations is not involving the end-users early enough. From the molecule steward to manufacturing these users are critical.

Raw Material and Component Selection

Carefully evaluate and qualify raw materials and components. Work closely with suppliers to understand material properties, extractables/leachables profiles, and manufacturing processes.

Develop comprehensive specifications for critical materials and components. ASTM E3244 is handy place to look for guidance on raw material qualification for SUS.

Manage the Supplier through Manufacturing and Assembly

Implementing robust supplier qualification and auditing programs and establish change control agreements with suppliers to be notified of any changes that could impact SUS performance or quality. It is important the supplier have a robust quality management system and that they apply Good Manufacturing Practices (GMP) through their facilities. Ensure they have in place appropriate controls to

Validate sterilization processes
Conduct routine bioburden and endotoxin testing
Design packaging to protect SUS during transportation and storage. Shipping methods need to protect against physical damage and temperature excursions
Establish appropriate storage conditions and shelf-life based on stability studies
Provide appropriate labeling and traceability
Have appropriate inventory controls. Ideally select suppliers who understand the importance of working with you for collaborative planning, forecasting and replenishment (CPFR)

Testing and Qualification

Develop a comprehensive testing strategy, including integrity testing and conduct extractables and leachables studies following industry guidelines. Evaluate the suppliers shipping and transportation studies to evaluate SUS robustness and determine if you need additional studies.

Implementation and Use

End users should have appropriate and comprehensive documentation and training to end users on proper handling, installation, and use of SUS. These procedures should include how to perform pre-use integrity testing at the point of use as well as how to perform thorough in-process and final inspections.

Consider implementing automated visual inspection systems and other appropriate monitoring.

Implement appropriate environmental monitoring programs in SUS manufacturing areas. While the dream of manufacturing outdoors is a good one, chances are we aren’t even close yet. Don’t short this layer of control.

Continuous Improvement

Ensure you have appropriate mechanisms in place to gather data on SUS performance and any issues encountered during use. Share relevant information across the supply chain to drive improvements.

Conduct periodic audits of suppliers and manufacturing facilities.

Stay updated on evolving regulatory guidance and industry best practices. There is still a lot changing in this space.

Health of the Validation Program

In the Metrics Plan for Facility, Utility, System and Equipment that is being developed a focus is on effective commissioning, qualification, and validation processes.

To demonstrate the success of a CQV program we might brainstorm the following metrics.

Deviation and Non-Conformance Rates

Track the number and severity of deviations related to commissioned, qualified and validated processes and FUSE elements.
The effectiveness of CAPAs that involve CQV elements

Change Control Effectiveness

Measure the number of successful changes implemented without issues
Track the time taken to implement and qualify validate changes

Risk Reduction

Quantify the reduction in high and medium risks identified during risk assessments as a result of CQV activities
Monitor the effectiveness of risk mitigation strategies

Training and Competency

Measure the percentage of personnel with up-to-date training on CQV procedures
Track competency assessment scores for key validation personnel

Documentation Quality

Measure the number of validation discrepancies found during reviews
Track the time taken to approve validation documents

Supplier Performance

Monitor supplier audit results related to validated systems or components
Track supplier-related deviations or non-conformances

Regulatory Inspection Outcomes

Track the number and severity of validation-related observations during inspections
Measure the time taken to address and close out regulatory findings

Cost and Efficiency Metrics

Measure the time and resources required to complete validation activities
Track cost savings achieved through optimized CQV approaches

By tracking these metrics, we might be able to demonstrate a comprehensive and effective CQV program that aligns with regulatory expectations. Or we might just spend time measuring stuff that may not be tailored to our individual company’s processes, products, and risk profile. And quite frankly, will they influence the system the way we want? It’s time to pull out an IMPACT key behavior analysis to help us tailor a right-sized set of metrics.

The first thing to do is to go to first principles, to take a big step back and ask – what do I really want to improve?

The purpose of a CQV program is to provide documented evidence that facilities, systems, equipment and processes have been designed, installed and operate in accordance with predetermined specifications and quality attributes:

To verify that critical aspects of a facility, utility system, equipment or process meet approved design specifications and quality attributes.
To demonstrate that processes, equipment and systems are fit for their intended use and perform as expected to consistently produce a product meeting its quality attributes.
To establish confidence that the manufacturing process is capable of consistently delivering quality product.
To identify and understand sources of variability in the process to better control it.
To detect potential problems early in development and prevent issues during routine production.

The ultimate measure of success is demonstrating and maintaining a validated state that ensures consistent production of safe and effective products meeting all quality requirements.

Focusing on the Impact is important. What are we truly concerned about for our CQV program. Based on that we come up with two main factors:

The level of deviations that stem from root causes associated with our CQV program
The readiness of FUSE elements for use (project adherence)

Reducing Deviations from CQV Activities

First, we gather data, what deviations are we looking for? These are the types of root causes that we will evaluate. Of course, your use of the 7Ms may vary, this list is to start conversation.

Means	Automation or Interface Design Inadequate/Defective	Validated machine or computer system interface or automation failed to meet specification due to inadequate/defective design.
Means	Preventative Maintenance Inadequate	The preventive maintenance performed on the equipment was insufficient or not performed as required.
Means	Preventative Maintenance Not Defined	No preventive maintenance is defined for the equipment used.
Means	Equipment Defective/Damaged/Failure	The equipment used was defective or a specific component failed to operate as intended.
Means	Equipment Incorrect	Equipment required for the task was set up or used incorrectly or the wrong equipment was used for the task.
Means	Equipment Design Inadequate/Defective	The equipment was not designed or qualified to perform the task required or the equipment was defective, which prevented its normal operation.
Media	Facility Design	Improper or inadequate layout or construction of facility, area, or work station.
Methods	Calibration Frequency is Not Sufficient/Deficiency	Calibration interval is too long and/or calibration schedule is lacking.
Methods	Calibration/Validation Problem	An error occurred because of a data collection- related issue regarding calibration or validation.
Methods	System / Process Not Defined	The system/tool or the defined process to perform the task does not exist.

Based on analysis of what is going on we can move into using a why-why technique to look at our layers.

Why 1	Why are deviations stemming from CQV events not at 0% Because unexpected issues or discrepancies arise after the commissioning, qualification, or validation processes Success factor needed for this step: Effectiveness of the CQV program Metric for this step: Adherence to CQV requirements
Why 2 (a)	Why are unexpected issues arising after CQV? Because of inadequate planning and resource constraints in the CQV process. Success Factor needed for this step: Appropriate project and resource planning Metric for this Step: Resource allocation
Why 3 (a)	Why are we not performing adequate resource planning? Because of the tight project timelines, and the involvement of multiple stakeholders with different areas of expertise Success Factor needed for this step: Cross-functional governance to implement risk methodologies to focus efforts on critical areas Metric for this Step: Risk Coverage Ratio measuring the percentage of identified critical risks that have been properly assessed and and mitigated through the cross-functional risk management process. This metric helps evaluate how effectively the governance structure is addressing the most important risks facing the organization.
Why 2 (b)	Why are unexpected issues arising after CQV? Because of poorly executed elements of the CQV process stemming from poorly written procedures and under-qualified staff. Success Factor needed for this step: Process Improvements and Training Qualification Metric for this Step: Performance to Maturity Plan

There were somethings I definitely glossed over there, and forgive me for not providing numbers there, but I think you get the gist.

So now I’ve identified the I – How do we improve reliability of our CQV program, measured by reducing deviations. Let’s break out the rest.

	Parameters	Executed for CQV
IDENTIFY	The desired quality or process improvement goal (the top-level goal)	Improve the effectiveness of the CQV program by taking actions to reduce deviations stemming from verification of FUSE and process.
MEASURE	Establish the existing Measure (KPI) used to conform and report achievement of the goal	Set a target reduction of deviations related to CQV activities.
Pinpoint	Pinpoint the “desired” behaviors necessary to deliver the goal (behaviors that contribute successes and failures)	Drive good project planning and project adherence. Promote and coach for enhanced attention to detail where “quality is everyone’s job.” Encourage a speak-up culture where concerns, issues or suggestions are shared in a timely manner in a neutral constructive forum.
ACTIVATE the CONSEQUENCES	Activate the Consequences to motivate the delivery of the goal (4:1 positive to negative actionable consequences)	Organize team briefings on consequences Review outcomes of project health Senior leadership celebrate/acknowledge Acknowledge and recognize improvements Motivate the team through team awards Measure success on individual deliverables through a Rubric
TRANSFER	Transfer the knowledge across the organization to sustain the performance improvement	Create learning teams Lessons learned are documented and shared Lunch-and-learn sessions Create improvement case studies

From these two exercises I’ve now identified my lagging and leading indicators at the KPI and the KBI level.

Not all Equipment is Category 3 in GAMP5

I think folks tend to fall into a trap when it comes to equipment and GAMP5, automatically assuming that because it is equipment it must be Category 3. Oh, how that can lead to problems.

When thinking about equipment it is best to think in terms of “No Configuration” and ” Low Configuration” software. This terminology is used to describe software that requires little to no configuration or customization to meet the user’s needs.

No Configuration(NoCo) aligns with GAMP 5 Category 3 software, which is described as “Non-Configured Products”. These are commercial off-the-shelf software applications that are used as-is, without any customization or with only minimal parameter settings. My microwave is NoCo.

Low Configuration(LoCo) typically falls between Category 3 and Category 4 software. It refers to software that requires some configuration, but not to the extent of fully configurable systems. My PlayStation is LoCo.

The distinction between these categories is important for determining the appropriate validation approach:

Category 3 (NoCo) software generally requires less extensive validation efforts, as it is used without significant modifications. Truly it can be implicit testing.
Software with low configuration may require a bit more scrutiny in validation, but still less than fully configurable or custom-developed systems.

Remember that GAMP 5 emphasizes a continuum approach rather than strict categorization. The level of validation effort should be based on the system’s impact on patient safety, product quality, and data integrity, as well as the extent of configuration or customization.

When is Something Low Configuration?

Low Configuration refers to software that requires minimal setup or customization to meet user needs, falling between Category 3 (Non-Configured Products) and Category 4 (Configured Products) software. Here’s a breakdown of what counts as low configuration:

Parameter settings: Software that allows basic parameter adjustments without altering core functionality.
Limited customization: Applications that permit some tailoring to specific workflows, but not extensive modifications.
Standard modules: Software that uses pre-built, configurable modules to adapt to business processes.
Default configurations: Systems that can be used with supplier-provided default settings or with minor adjustments.
Simple data input: Applications that allow input of specific data or ranges, such as electronic chart recorders with input ranges and alarm setpoints.
Basic user interface customization: Software that allows minor changes to the user interface without altering underlying functionality.
Report customization: Systems that permit basic report formatting or selection of data fields to display.
Simple workflow adjustments: Applications that allow minor changes to predefined workflows without complex programming.

It’s important to note that the distinction between low configuration and more extensive configuration (Category 4) can sometimes be subjective. The key is to assess the extent of configuration required and its impact on the system’s core functionality and GxP compliance. Organizations should document their rationale for categorization in system risk assessments or validation plans.

Attribute	Category 3 (No Configuration)	Low Configuration	Category 4
Configuration Level	No configuration	Minimal configuration	Extensive configuration
Parameter Settings	Fixed or minimal	Basic adjustments	Complex adjustments
Customization	None	Limited	Extensive
Modules	Pre-built, non-configurable	Standard, slightly configurable	Highly configurable
Default Settings	Used as-is	Minor adjustments	Significant modifications
Data Input	Fixed format	Simple data/range input	Complex data structures
User Interface	Fixed	Basic customization	Extensive customization
Workflow Adjustments	None	Minor changes	Significant alterations
User Account Management	Basic, often single-user	Limited user roles and permissions	Advanced user management with multiple roles and access levels
Report Customization	Pre-defined reports	Basic formatting/field selection	Advanced report design
Example Equipment	pH meter	Electronic chart recorder	Chromatography data system
Validation Effort	Minimal	Moderate	Extensive
Risk Level	Low	Low to Medium	Medium to High
Supplier Documentation	Heavily relied upon	Partially relied upon	Supplemented with in-house testing

Here’s the thing to be aware of, a lot of equipment these days is more category 4 than 3, as the manufacturers include all sorts of features, such as user account management and trending and configurable reports. And to be frank, I’ve seen too many situations where Programmable Logic Controllers (PLCs) didn’t take into account all that configuration from standard function libraries to control specific manufacturing processes.

Your methodology needs to keep up with the technological growth curve.

Risk Assessments as part of Design and Verification

Facility design and manufacturing processes are complex, multi-stage operations, fraught with difficulty. Ensuring the facility meets Good Manufacturing Practice (GMP) standards and other regulatory requirements is a major challenge. The complex regulations around biomanufacturing facilities require careful planning and documentation from the earliest design stages.

Which is why consensus standards like ASTM E2500 exist.

Central to these approaches are risk assessment, to which there are three primary components:

An understanding of the uncertainties in the design (which includes materials, processing, equipment, personnel, environment, detection systems, feedback control)
An identification of the hazards and failure mechanisms
An estimation of the risks associated with each hazard and failure

Folks often get tied up on what tool to use. Frankly, this is a phase approach. We start with a PHA for design, an FMEA for verification and a HACCP/Layers of Control Analysis for Acceptance. Throughout we use a bow-tie for communication.

Aspect	Bow-Tie	PHA (Preliminary Hazard Analysis)	FMEA (Failure Mode and Effects Analysis)	HACCP (Hazard Analysis and Critical Control Points)
Primary Focus	Visualizing risk pathways	Early hazard identification	Potential failure modes	Systematically identify, evaluate, and control hazards that could compromise product safety
Timing in Process	Any stage	Early development	Any stage, often design	Throughout production
Approach	Combines causes and consequences	Top-down	Bottom-up	Systematic prevention
Complexity	Moderate	Low to moderate	High	Moderate
Visual Representation	Central event with causes and consequences	Tabular format	Tabular format	Flow diagram with CCPs
Risk Quantification	Can include, not required	Basic risk estimation	Risk Priority Number (RPN)	Not typically quantified
Regulatory Alignment	Less common in pharma	Aligns with ISO 14971	Widely accepted in pharma	Less common in pharma
Critical Points	Identifies barriers	Does not specify	Identifies critical failure modes	Identifies Critical Control Points (CCPs)
Scope	Specific hazardous event	System-level hazards	Component or process-level failures	Process-specific hazards
Team Requirements	Cross-functional	Less detailed knowledge needed	Detailed system knowledge	Food safety expertise
Ongoing Management	Can be used for monitoring	Often updated periodically	Regularly updated	Continuous monitoring of CCPs
Output	Visual risk scenario	List of hazards and initial risk levels	Prioritized list of failure modes	HACCP plan with CCPs
Typical Use in Pharma	Risk communication	Early risk identification	Detailed risk analysis	Product Safety/Contamination Control

At BOSCON this year I’ll be talking about this fascinating detail, perhaps too much detail.

Retrospective Validation Doesn’t Really Exist

A recent FDA Warning Letter really drove home a good point about the perils of ‘retrospective validation’ and how that normally doesn’t mean what folks want it to mean.

“In lieu of process validation studies, you attempted to retrospectively review past batches without scientifically establishing blend uniformity and other critical process performance indicators. You do not commit to conduct further process performance qualification studies that scientifically establish the ability of your manufacturing process to consistently yield finished products that meet their quality attributes.”

The FDA’s response here is important for three truths:

Validation needs to be done against critical quality attributes and critical process parameters to scientifically establish that the manufacturing process is consistent.
Batch data on its own is rather useless.
Validation is a continuous exercise, it is not once-and-done (or rather in most people’s view thrice-and-done).

I don’t think the current GMPs really allow the concept of retrospective validation as most people want it to mean (including the recipient of that warning letter). It’s probably a term we should go into the big box of Nope.

Retrospective validation as most people mean it is a type of process validation that involves evaluating historical data and records to demonstrate that an existing process consistently produces products meeting predetermined specifications. As an approach retrospective validation involves evaluating historical data and records to demonstrate that an existing process consistently produces products meeting predetermined specifications.

The problem here is that this really just tells you what you were already hoping was true.

Retrospective validation has some major flaws:

Limited control over data quality and completeness: Since retrospective validation relies on historical data, there may be gaps or inconsistencies in the available information. The data may not have been collected with validation in mind, leading to missing critical parameters or measurements. It rather throws out most of the principles of science.
Potential bias in existing data: Historical data may be biased or incomplete, as it was not collected specifically for validation purposes. This can make it difficult to draw reliable conclusions about process performance and consistency.
Difficulty in identifying and addressing hidden flaws: Since the process has been in use for some time, there may be hidden flaws or issues that have not been identified or challenged. These could potentially lead to non-conforming products or hazardous operating conditions.
Difficulty in recreating original process conditions: It may be challenging to accurately recreate or understand the original process conditions under which the historical data was generated, potentially limiting the validity of conclusions drawn from the data.

What is truly called for is to perform concurrent validation.