ECA released Version 3.0 of "The GMP/GDP Questions & Answers Guide."

Sep. 18, 2026

In September 2026, ECA Academy officially released Version 3.0 of "The GMP/GDP Questions & Answers Guide." This guide compiles common and challenging questions in GMP/GDP practices in a Q&A format, covering multiple compliance areas from manufacturing and quality to supply chain, serving as a crucial reference for the pharmaceutical industry to address differences in regulatory interpretation.

In this latest edition, Section 6.1, "Questions and Answers on Visual Inspections," specifically focuses on visual inspection of injectables and visible particulate testing. It covers topics including manual and automated inspection, qualification/validation/requalification, test sets, AQL testing, defect classification, special products, regulatory affairs, and process control/SPC—nearly encompassing all key aspects of compliant visual inspection. This section is highly recommended for personnel involved in visual inspection within pharmaceutical companies.

6.1 Questions and Answers on Visual Inspections

Manual Inspection

MI1 According to Annex 1 of the EU-GMP-Guideline, operators doing the inspection for particles and other defects should make frequent breaks from inspection. Are there any regulations concerning rest times and time intervals?

An interpretation of this requirement is not trivial. Sometimes, the companies interpret it in a significantly different way. In practice, for example, the following variants can be found:

• 55 minutes inspecting, 5 minutes break over a period of 8 hours

• 20 minutes inspecting, 20 minutes other activities over a period of 8 hours

• 20 minutes inspecting, 5 minutes break over a period of 4 hours

It is a known fact that this activity normally can be carried out successfully for a maximum period of 15 to 20 minutes. Insofar the provision "55 minutes inspecting, 5 minutes break over a period of 8 hours" doesn't seem to be appropriate. Ideally, the activity of inspection will be limited to 15 to 20 minutes. Afterwards a break for the eyes should be taken without carrying out any other activities.

MI2 Which requirements should there be concerning the requalification of the personnel carrying out the optical inspection? Are there specific time intervals to be observed and should the requalification be announced?

The pharmaceutical companies have different approaches concerning the time intervals for requalification. The intervals reach from each month to every two years. Every two years certainly is too long. This period of time is especially problematic if the person carrying out the inspections doesn't pass the requalification any more. Which consequences does this have for the batches inspected by this operator? Insofar it seems to be appropriate to choose a time interval that is as short as possible and to control the personnel for example by means of an AQL testing. Are test sets used in the context of the requalification this should not be announced before the qualification run. For this reason, it is advisable to requalify not exactly every six/twelve months, for example, but to do this on a random basis i.e. to lay down irregular time intervals. If test kits are processed in the context of requalification you need to think carefully about the question whether the test kits should be inspected directly in the morning or at the end of the day in the case of an eight-hour shift. Naturally, it is advisable to choose the worst case situation and therefore, to let the inspection be carried out at the end of the shift.

MI3 How often have employees of the manual (or semi-automatic) inspection to be trained? What does "regularly" mean in this context?

With regard to visual inspection, the difference between initial qualification and requalification must basically be made. Strictly speaking, requalification is an examination of the status of a person's qualification. It is recommended to re-qualify employees at least every 12 months.

MI4 Should the set of samples for the qualification (training) of the staff be taken directly form the production rejects or should it be produced artificially based on typical defects?

Usually, the training set should be taken directly from the production rejects. Unfortunately, some failures occur only very rarely so that you have to produce some of them yourself. Finally, you have a set with failures that are typical for the production that has to undergo a formal release procedure described in an SOP. This guarantees that the failures are typical and that you can use them for training.

MI5 Are there empirical values about how long do lamps in semi-automatic testing stations keep their intensity?

Here, the certificates of the lamp manufacturers should be consulted. Usually, 2-3 years are indicated. In many companies, the lamps are exchanged routinely after one year to never have to question the test results due to possible lamp weakness retrospectively.

MI6 We have trained our inspection team to inspect containers in less than 5 seconds against a white and black background. Moreover we do an AQL testing afterwards and if the AQL fails the qualification status of the inspector will also be evaluated within a deviation. So is it allowed so shorten the inspection time?

Ph.Eur. and USP requirements for particulates are clear – manual visual inspection for 5 sec in front of black and white background is required. Annex 1* of the GMP guidance document states to use a given time and given set-up for inspection. Any modification must be validated and should show to be equal or better than the compendial approach. *(Vs. 2008 rev.)

MI7 According to Ph.Eur the light intensity should be between 2000-3750 Lux but according to USP 790 it should be minimum 2000-3750 Lux. Is it ok then that the light intensity is higher than 3750 Lux?

The “minimum intensity” in USP <790> refers to 2000 Lux. But higher intensities as 3750 Lux or higher are possible and acceptable for certain types of products (e.g. Blow-Fill-Seal container). But operators fatigue will be higher and you have to adopt the eye breaks and the inspection time per day accordingly.

MI8 Is the wearing of anti reflective eyewear by manual inspectors a common practice among pharmaceutical companies during Manual or Semi automatic inspection as a measure against fatigue?

Wearing anti reflective eyewear is not the way for improving the quality of the visual inspection. On one hand a high light intensity can be necessary, on the other hand the tiring of the human eye seems to work contra productive to the detecting of the defects. One has to find a way between both effects meaning: use more eye breaks when using light with a higher intensity.

MI9 Is there a requirement in regards to the size of the visual inspection booth, e.g. width, length, etc?

MI10 How can foreign particulate matter be distinguished from a micro bubble in a vial that can form during vial stirring during manual inspection?

Up to now there are limited technologies for distinguishing air bubbles and real particles. Cameras and the human eye cannot see a difference when the particles/air bubbles are small. The only way is to avoid air bubbles. One approach is the use of two cameras. In case particle-camera1 detects something but particle-camera2 (under the same conditions as camera1) does not, it is likely an air bubble.

MI11 What exactly is meant by point 2.1 of the best practice paper ("The relative humidity and air velocity should be controlled and ensure comfortable working conditions.")? Is a permanent room monitoring required?

It is a fundamental GMP requirement: the fulfilment of a requirement has to documented. A permanent monitoring of temperature and humidity may not be necessary, but how to prove during an inspection, that the working conditions for the human inspectors have been adequate? This will not be possible without measuring temperature and humidity and some sort of system, meaning the documentation and evaluation of measuring data. If this is not part of the building control system anyway, a manual system may also be possible.

Automated Inspection

AV1 The grey portion of fully automatic control is often checked manually, to return not clearly or fully tested products back to the inspection process. Is it allowed to carry out this testing with the automated inspection machine?

From a GMP view, there are no restrictions. It is also important here that at the end a yield calculation and evaluation in the batch record appears. And there are also automated inspection systems that have already integrated the double inspection with multiple cameras.

AV2 Can one reject test be considered as a good after two "good" inspection on the same machine?

This is possible in a few cases where - for example - the machine stopped and goods were therefore ejected. Otherwise, "reject" should always remain "reject". This is particularly applicable to "bad" goods which have been rejected because of particles or opacity. Containers sorted out due to cosmetic defects are however usually being re-inspected.

AV3 We produce a lot of products with different formats. Until now we use a rejection rate of <2% as acceptance criterion and sets of samples from production to control the functionality of the machine.

An AQL test should be carried out for each batch. In the meantime this is expected by the authorities and inspectors. This is also described in the ECA Good Practice Guide on Visual Inspection. A control of the rejection rate of the 100% inspection is also expected but it rather serves for recognizing whether a batch differs from the normal unobtrusive production. This trending limits should be product specific. A generic limit eg <2% would need a rationale.

AV4 In highly automated manufacturing lines for LVP flexible containers, the visual inspection process may/cannot comply to the standard visual inspection criteria e.g.: 5 sec inspection time, agitation of the container etc. Is this a compliance problem?

The requirements like 5 sec inspection time required by pharmacopoeias are addressing manually performed visual inspection. If the visual inspection is performed automatically, it is the company’s responsibility to ensure that the inspection via camera systems is as effective as a manual visual inspection via a validation (e.g. Knapp Test).

AV5 Do we have to perform challenge test before production / shift to check the 100% automatic visual inspection? And how often?

A function test kit (system suitability test kit) used before and after the inspection of each batch to demonstrate the functionality of the fully automated inspection system. It may contain an abridged set of more apparent defects such as big particles, cracked or empty containers.

Qualification /Validation

QV1 In the course of validation and during operation there are recurring problems with false reject rates in the case of fully automated systems. Are there any GMP requirements concerning this?

Due to the fact that the systems are able to detect also considerably smaller particles than human operators there are repeatedly emerging more or less big amounts of objects in the part with defects that have been assessed by a human operator as being good. Furthermore, a fully automated system might also get problems with air bubbles and reject these objects as having defects. In the end, the trick is to configure the system in such a way that no objects containing very small particles are rejected. In some companies the objects rejected by automated systems are again inspected by human operators. But this method entails the risk that objects actually having defects are suddenly classified as having no defects by the human operator. Two conclusions can be drawn from the point of view of GMP. Rejecting objects without defect does not entail a risk for patients and thus seems to be practicable. On the other hand one could also criticise the qualification as such since a system making errors is not sufficiently qualified. In any case, acceptance criteria should be defined for the part with defects. In the case of exceedance certain measures are to be taken such as an additional 100 % inspection for example. To sum it up: there is no requirement to set a limit for false rejects. But on the other hand it is advisable to also set a limit for the false rejects, usually during the qualification/5000 test. Because having no limit or a very high number of false rejects may question the whole qualification during a GMP inspection.

QV2 We have about 50 different aqueous solutions for a few hundred products (one solution for several formats). So far we have prepared one set of samples (function set, qualification set including Knapp set) from production rejects. This is very time consuming. Do you think it might be useful to measure parameters such as viscosity, surface tension of the 50 solutions in order to group the products (bracketing)?

Bracketing is useful here. The rationale using viscosity is ok.

QV3 Which statistical tests should be applied to demonstrate equivalence of different visual inspection processes (e.g. manual versus automatic inspection process)?

The goal of the AIM qualification is to show that the machine is equal or better than the “gold standard” that is the human inspection. This can be done by comparing the overall detection rates for particles (this set also includes non-particle objects) and at least 10 inspections runs of this set for manual inspection (normally 3 operators) and the same set of objects on the machine. For non-particle defects (scratches, missing stopper,…) one could use a predefined limit or also a man/machine comparison.

Test Sets

TS1 What are the differences between qualification, particle (Knapp test) and function test set?

The function test set serves for a sort of system suitability test, i.e. this test is used to test before and after each batch whether the cameras are functioning correctly. Usually, no challenging samples are used for this, but rather unities with particles having a detection rate of 100% such as particles with a size of 1000 µm, vials with missing stoppers.

Qualification test set: the qualification test set consists of product specific containers containing the product and having all known "static" defects (scratches, wrong flip-off, missing stopper,...). Usually, about 10-20% of the containers of the set have a defect. New failures or defects are added to the qualification test set.

Particle test set (Knapp-Test): Sets that contain only particles. These are particles having the size from 50µm to 1000µm and consisting of different materials (plastic, the material stoppers are made of, glass, metal). Hence, they are "non-static", i.e. the defect is in the container or in the drug solution. Particle test sets are part of the qualification test set at the same time.

TS2 Should all these test sets be prepared artificially? Should this be done, for instance, by an external laboratory with defined failures?

We prepare the static defects in-house and a part of the non-static defects (particle defects) are produced externally. But it is also possible to have everything produced externally. But the static defects should be the same that are generated in the worst case by your production machines. Therefore, I advise to make these in-house and get the release from QA. In this way you would have representative bad samples from your process. In the best case you take bad samples from your process but it is not possible to do this for all samples.

TS3 The problem is that sets of samples have to be remade regularly since they have expired. How do you prepare the function test set?

Usually, our sets keep several years. Single samples have to be replaced again and again. These single samples are then released by QA and we introduce them into the set. The sets should be controlled at least once a year and be released again afterwards.

TS4 How long can a training kit be used?

Training kits should contain all kind of defects and must be updated constantly with new evolving defects out of production. Expiry of specific defects depends on nature (a crack will not expire; small particles may clot together…). The set must be regularly released and reinspected by a supervisor.

TS5 How should qualification sets be prepared for a new product?

All representative defects coming from a production line should be included in the set. Some of the defects need to be generated but if possible should be taken from production. These defects need to be classified. See chapter 5 of the Best Practice Guide.

Requalification

RQ1 If systems for the fully automatic visual inspection are used, regular functional testing is carried out. Must the system be requalified nevertheless?

According to EU-GMP a difference must be made between the following activities in the case of a fully automated system:

• Qualification: in the case of new equipment

• Requalification: Assessment of equipment within defined time intervals

• Periodic evaluation: According to EU-GMP Annex 11.

Functional testing addresses the sensor functionality only and is typically carried out before and after batch inspection. Due to its limited scope, functional testing cannot replace requalification of an equipment. Requalification should be performed according to the EU-GMP Annex 15 to safeguard that the whole equipment remains in a state of control. Requalification programs and intervals shall be planned and justified according to the guideline. In the context of the assessment of the facility, the functional testing carried out until the date of requalification will certainly be part of the assessment.

RQ2 Requalification: If the regular requalification of the automatic inspection machine is carried out by means of the test of 5,000 does this mean that the 5,000 vials from one production batch must all be controlled manually and by the automatic inspection machine?

I.e. is it necessary to repeat the man-machine-comparison that was carried out in PQ? Almost correct. PQ consists of the particle (Knapp test) run, the run concerning the "static" defects and also of a run of 5,000 of GOOD vials. This run of 5,000 is carried out once a year for each product. The Knapp test and the "static" test are not repeated except when there have been changes at the particle detection stations. Without changes (something that occurs only very seldom) only the test of 5,000 is carried out, i.e. 5,000 vials from the automatic inspection machine are inspected by a person and by the machine

AQL-Testing

AQ1 Is AQL testing mandatory as a part of the visual inspection?

A direct requirement cannot be derived from the EU GMP. Yet, the AQL tests correspond to the state of the art in science and technology. Since we know that neither a manual nor an automated visual inspection can guarantee a 100% particle-free batch, an additional measure - like the AQL tests - is certainly appropriate. Another method would be of course a second 100% inspection. Or you could show in the validation that the test method used is a 100% flawless and complete, what will hardly be possible in practice. For US, the AQL testing has been included in the USP, chapter <790>.

AQ2 Do the European agencies follow the rules for AQL testing given in USP <790>?

There is no written guidance in Europe requiring an AQL test according USP <790>. Companies have to define their own way to implement a check of the visual inspection efficacy.

AQ3 Should the AQL be inspected by QC or production

AQL manual inspection may be carried out by production staff (to avoid setting up a separate visual inspection team in QC) under a quality oversight or the quality unit. If performed by production operators, the AQL test should not be done by members of the team that was performing the 100 % visual inspection of the batch.

AQ4 Is the AQL acceptance criteria of 0.65 for the particles only or for the overall defects?

USP <790> addresses visual inspection for particles. So the requirement to apply an AQL of 0.65 or sample plans with better protection applies to particles. Other defects like cracks or stopper failures may be addressed with tighter AQLs.

AQ5 Which sample size should be taken if various AQL-levels depending on the defect criticality are used (for example batch size of 22.000, AQL = 0 for critical defects, AQL = 0.65 for major defects, AQL = 2.0 for minor defects)

There is a misunderstanding. The sample size is depending on the Quality Level one wants to use (e.g. level II). With this level and the batch size one gets the size of the sample which must be drawn. The AQL level (e.g. 0.65) then defines how many defects are allowed to accept the defined batch quality. Of course the AQL level then depends on the criticality because the more critical the less defects are allowed to be found in the AQL procedure.

AQ6 For vials packaged in separated sub-batches, when the 100% visual inspection is located at the beginning of the packaging process, should the acceptance sampling be statistically significant on the full manufacturing batch or on each single sub-batch?

The AQL sampling should be based on the sub-batch meaning the batch size that is inspected. One is checking with this test if the incoming quality of the batch is according the quality which is expected. The levels do not need to be the same like for the initial batch.

Defect Categorisation

DC1 The USP in sections 790 and 1790 classifies glass defects as a major defect. There are many types of glass defects, e.g. glass particulates adhered to the side of the vial, and loose glass within a vial (in both incidents if product integrity/sterility assurance is not compromised), could you please tell me what the defect classification levels (critical, major or minor), for either of the former mentioned defects should be.

Both USP chapters do not make any classification of defects. This needs to be done by the pharmaceutical company and has to be based on a quality risk assessment.

DC2 How should freeze-drying defects such as collapse and melt-back be classified regarding to their criticality

In the scientific literature a melt-back is seen to be critical. The difference of a collapsed and a melt back is difficult to define. A collapsed lyo cake could also be a melt back. One would need to explain this difference and its criticality.

Special Products

SP1 Our DP is a powder. Is visual inspection of the reconstituted solution sufficient?

100% visual inspection of the delivered DP is a general requirement. The inspection of the reconstituted solution would be part of the release process. According USP<1790> special sampling plans according S-3 and S-4 (ANSI/AQS Z1.4) should be used.

SP2 How should the operators performing the supplemental testing be qualified?

Supplemented testing is performed on reconstituted vials and thus has to be performed in a lab environment. The reconstituted vials are tested either by manufacturing operators or QC staff based on the same criteria as training for 100 % visual inspection or AQL testing.

SP3 What is the recommendation for AQL setting for the testing of samples after reconstitution?

This depends on the nature of the product. Taking 0.65 will in many cases disqualify nearly all batches. In some cases it might be an option to allow a second level testing to avoid that 1 single particle in one vial out of 20 requires rejection of your batch.

Regulatory Affairs

RA1 If the 100% visual inspection is followed by an AQL testing, do we have to perform also an test on visual particles at the release (testing) of the batch?

The AQL testing is intended to replace testing for visible particles at release testing in the lab. However, if this test is part of your filing, this has to be done. Or you have to file a variation.

RA2 Is the carrying out of a 100% inspection of parenterals to be understood as IPC testing or as final product testing? Is it possible to carry it out under the responsibility of production or must it be done by QC?

Assigning the activity of a 100% inspection is not trivial at first sight. Formally, this activity is allocated to production. By the way, this is also FDA's point of view. Hence, it also is an activity which requires a manufacturing authorisation within the meaning of § 13 German Medicine Act (AMG). Due to the fact that it is a 100% inspection, it is neither a real IPC nor a final inspection on a random basis. The 100% inspection being attributed to production, it is carried out under the responsibility of the head of production

RA3 Are there legal acceptance criteria or provisions regarding the size of (visible) particles?

According to studies by Jules Knapp, people can recognise particles under optimal conditions from 30 - 50 µm, taking into account the colour of the particle. Other studies have shown that eventually particles from approximately 200 µm can be surely detected - in other sources, values of 50-80 µm can be found. There are (still) no statutory limits on the size of the visible particles

RA4 What does essentially free from particles mean with respect to the US and EU pharmacopeia? Is there a difference?

Up to now there is no official statement from the EMA about the interpretation of “essentially free from particles”.

RA5 How has the limit for inherent particles to be set in a registration of a new protein product which is known to form particles?

The guidance according USP<1790> would be to define a limit and to monitor this limit. Of Course this limit needs then to be described within the dossier of the DP.

Process Control / SPC

PC1 How can Process Control Limits in the Visual inspection process be defined?

Typical limits for product and production line need to be established. Therefore see chapter 6 of the Best-Practice Paper. This follows the ASTM E2587-2 standard (Standard Practice for Use of Control Charts in Statistical Process Control).