Manual Visual Inspection (MVI) of Injectables

The Impact of Particulate Matter in Injections

The impact of particulate matter in injections on clinical outcomes is influenced by several factors, including the size, shape, colour, and quantity of particles, the potential for microbiological contamination, the composition of the material, the target patient population, and the specific clinical condition of the patient. To minimise these risks, Manual Visual Inspection plays a critical role in identifying and removing particulate contamination during the manufacturing process.

To ensure that injectable products are free from particulate matter, each product should undergo 100% visually inspection as part of the routine manufacturing process.

The purpose of visual inspection is to detect the following three types of particulate matter in the product and to reject contaminated products.

• Extrinsic particles – Foreign particles not associated with the manufacturing process, such as hair, insect parts, non-process related fiber, etc. Typically, extrinsic particles are a one-time occurrence; however, elevated levels may indicate a broader contamination issue.
• Intrinsic particles – Particles within the process, which are associated with production process, such as glass pieces, stainless steel, process-related fibers, etc. Any process-related intrinsic particles should have appropriate controls established during process development.
• Inherent particles – Particles associated with specific products or their formulations, such as proteinaceous particulates, liposomes, or agglomerates etc. Any product or formulation related inherent particles should have appropriate controls established during the product development and stability studies.

Each unit of injectable product should undergo one of the following processes as part of manufacturing:

• Manual Visual Inspection (MVI)
• Semi-Automated Visual Inspection (SAVI)
• Automated Visual Inspection (AVI)

The most commonly used process in pharmaceutical manufacturing is Manual Visual Inspection (MVI), which can be developed through a series of standard steps.

How to develop a Manual Visual Inspection Process

Product Appearance / Defect Standard Library

Establish a product appearance library and develop a list of typical defects to represent realistic defects for particulates, container closure and cosmetic defects.

Defect Classification

Defects are classified based on patient safety and compliance risk. Critical defects are those that may cause a serious adverse reaction or death. Major defects present a risk of temporary impairment, medically reversible reaction, or a remote probability of a serious adverse reaction. Minor defects do not affect product performance or compliance and are often cosmetic in nature.

Defect Standard Preparation

Develop a procedure to generate the defects in a reproducible manner. The particle size should be near the visual detection threshold 80µm – 100µm to not more than (NMT) 200µm.

The defects should include a blend of seeded intrinsic particles, including simulated or naturally occurring particulates. The production rejects or naturally occurring particles that are anticipated from the process or environment can be used in a placebo solution with the same characteristics as the product.

A key part of the development of the MVI process is to identify the Critical Process Parameters (CPP).

Critical Process Parameters

• Light Intensity – Light levels of not less than (NLT) 2,000 -3,750 lux at the point of inspection are recommended. Increased light levels are recommended for translucent plastic containers or for containers made of amber glass.
• Background and Contrast – Contrast between the unit under inspection and the surrounding background is required for particle detection. The use of white and black backgrounds provides a good contrast for a wide range of particulate and container defects.
• Inspection Rate – Ten seconds per container (five seconds each against black and white backgrounds) should be allowed to enable a thorough inspection of each container.
• Container Handling and Movement – Use a technique of careful swirling or inversion of the liquid product in the container to set particles into motion while minimising air bubble formation.
• Magnification – Use a magnifier to enlarge the image and thereby increase the probability of detecting units with defects.

Inspector Training

It is important to acknowledge that the quality of an inspector’s vision plays a significant role in the Manual Visual Inspection (MVI) process.

A potential visual inspector’s near vision should be equivalent to 20/20, either naturally or corrected, with no impairment in colour vision. Initially, all potential visual inspectors should undergo eye testing to confirm their visual capability. Subsequently, annual eye testing should be conducted to ensure there is no deterioration in vision.

Potential visual inspectors should be trained on GMP principles, inspection techniques, defect types, lighting conditions, and handling procedures.

The training involves allowing potential visual inspectors to inspect known defective and acceptable (good) units. It is recommended to use a staged approach and have three Training Challenge Kits for training potential visual inspectors. The total number of units of the Training Challenge Kits should be based on the number of units that can be inspected within one inspection interval.

• Training Challenge Kit 1 – 100% defects
• Training Challenge Kit 2 – 30-50% defects and 50-70% acceptable units
• Training Challenge Kit 3 – 20% defects and 80% acceptable units

First, trainees should familiarise themselves with acceptable units and the defect library, including:

• Created defects
• Actual product defects

Then, trainees should inspect each Training Challenge Kit a minimum of three times and record the results. The Probability of Detection (POD) for each defect and for each trainee should be calculated and compared against the predefined acceptance criteria. The trainees who meet the predefined acceptance criteria will be considered to have passed the training.

Detection Rate Determination – Knapp’s Methodology (Human Baseline)

Once the defects have been created and the critical process parameters (CPPs) determined, the defects should be inspected by trained inspectors in accordance with Knapp’s methodology (USP <1790>) to determine the detection frequency.

Knapp’s methodology is based on developing the statistical detection probability for each individual defect. For establishing a human baseline, each unit needs to be inspected multiple times by multiple inspectors to develop a Probability of Detection (POD):

• Units detected less than 30% of the time are classified as “Accept Zone” units
• Units detected 30–70% of the time are classified as “Grey Zone” units
• Units detected more than 70% of the time are classified as “Reject Zone” units.

Next determine the total number of units to be assessed based on the number of units that can be inspected within one inspection interval. 80% of the total units to be assessed should be certified as acceptable units and 20% should be certified as defective.

As per USP<1790>:

• 50-60% of defects should be classified as critical defects,
• 30-35% as major defects and
• 10-15% as minor defects.

This example contains a total of 180 units. Of these, 144 units are certified as acceptable, 20 units have critical defects, 12 units have major defects, and 4 units have minor defects. The inspection is conducted by multiple inspectors on several occasions and Probability of Detection (POD) is calculated for each unit. Each unit is then catergorised as “Accept”, “Gray”, or “Reject”.

Inspector Qualification

The trained inspector qualification will involve allowing potential visual inspectors to inspect Qualification Challenge Kit. This kit contains defects identified with a Probability of Detection (POD) greater than 70% during the Knapp study, along with some acceptable units. As per USP <1970>, it should have no more than (NMT) 10% rejects to prevent rejection bias. The total number of units in the Qualification Challenge Kit should be determined based on the number of units that can be inspected within one inspection interval.

It is important to note the potential for falsely rejecting good units. Therefore, an acceptable False Reject Rate (FRR) needs to be established, typically set at less than 5%.

The trained inspectors should inspect the Qualification Challenge Kit a minimum of three times. The Probability of Detection (POD) for each defect and for each inspector should be calculated and compared with the Knapp Study results. Inspectors who achieve a POD equal to or greater than that of the Knapp Study are considered fully qualified.

Additionally, a False Reject Rate (FRR) should be established for each inspector. Inspectors who have a False Reject Rate (FRR) greater than 5% may require further inspection of the Qualification Challenge Kit.

Once the inspectors are fully qualified, they can participate in routine product inspections, with requalification performed at six-month intervals using an updated defect library.

Routine Manual Visual Inspection

A detailed Standard Operating Procedure (SOP) for Manual Visual Inspection (MVI) should be developed, outlining the critical process parameters of MVI, including fatigue break for inspectors of not less than (NLT) 5-10 minutes per hour. Once the 100% Manual Visual Inspection (MVI) is completed, AQL sampling and inspection should be conducted on acceptable units as per ISO 2849 level II sampling plan.

Typical AQL Acceptable Values for Manual Visual Inspection (MVI) process:

If the AQL inspection results meet the acceptance criteria of the sampling plan, the batch is deemed as suitable for release.

If the acceptance criteria of the sampling plan are not met, a Root Cause Analysis (RCA) should be conducted. After the RCA, if the inspection process is deemed capable of detecting the defects, the batch may be reinspected. Following a second 100% inspection (i.e. a reinspection), a new AQL sample of acceptable units should be taken and inspected. Reinspection should only be conducted according to the pre-approved procedure and the acceptance criteria should be the same as the primary inspection or more stringent.

In cases where a high False Rejection Rate (FRR) is observed due to assignable causes such as air bubbles or specific container defects, the implementation of a second inspection stage for rejects from the first inspection stage may be considered. The limitations of the first inspection stage and the reasons for conducting a second inspection stage should be clearly defined and documented. The units accepted by the first and second inspection stage should be sampled separately and evaluated and confirmed as accepted before recombining into a single batch.

As part of an inherent GMP mindset, it is best practice to characterise all defects and update the defect library for continuous improvement of the manufacturing and packaging process, as well as to enhance the Manual Visual Inspection (MVI) process. Additionally, collecting and trending inspection data for each inspector will help identify retraining needs and further develop inspector training and qualification activities over time.

PharmOut can help you with commissioning, qualification, and inspector training of Manual Visual Inspection.