Combination product testing questions

What is combination product testing?

Combination products deliver biologics and other pharmaceutical agents accurately and dependably to the point of need. Validation and batch testing for these systems examine those properties. Is the dose delivered accurately as intended without any extraneous material? Studies therefor revolve around stability, chemical migration and mechanical performance. That performance can include diffusion across membranes, particle generation, liquid dose volumes, dissolution, user interaction and many other aspects.

ISO 11608-1:2022 How to Define and Test Primary Functions


With the new version of ISO 11608-1:2022 being released, it introduces the concept of Primary Functions of Needle Injection System (NIS) devices, a similar concept (Essential Performance) is detailed in IEC 60601-1. ISO 11608-1:2022 defines Primary Function as:

‘Functions or operation of an NIS which, if it does not perform to specifications during use would directly result in failure to accurately deliver the medicinal product via the correct route and/or directly result in unacceptable harm to the patient’.

In summary the Primary Function of a NIS device is a function, that if it were to not perform as designed would inaccurately deliver its intended does and potentially cause harm to the end user.

Primary Function Determination

When determining the Primary Function of a NIS device two criteria need to be addressed:

  • If the Primary Function was to not perform as intended would it directly affect the NIS in its capability to accurately deliver the required dose? (Criteria 1)
  • If the Primary Function was to not perform as intended would it directly lead to direct harm to the end user? (Criteria 2)

Each individual function should be assessed against both of these criteria, if ‘Yes’ is answered to either criteria then that function is deemed as a Primary Function.

Primary Functions are generally determined early on in the design process, this means they are subject to expanded verification. This will include ensuring that the Primary Functions in-use are fully maintained throughout their in-use life cycle, at the end of their shelf life, after transport simulation (for example ASTM D4169) and after certain pre-conditioning (for example Cool, Standard and Warm conditions). Device functionality not designated as the Primary Function has no need for extensive pre-conditioning and can be tested at standard conditions.

Examples of Primary Functions

Dose Accuracy – This can be determined to be a Primary Function from Criteria 1 alone as under-dosing, overdosing or complete dosage failure is a failure to accurately deliver the required dose by the correct route.

Needle Safety Features – This has no impact on the delivery of the required dose nor does it cause unacceptable harm if the safety features fails to work. It will be unlikely to be designated a Primary Function by either criteria. Although ISO 11608-1:2022 recommends testing the safety features against ISO 23908.

Dose Dial Indication – This does not direct impact the accurate delivery of the dose when compared to Criteria 1, as the audible or visual provide feedback to the end user on the status of the NIS device. If the end user were to not complete the injection before removal this could potentially lead to significant harm, this means under Criteria 2 this may be considered a Primary Function.

ISO 11608-1:2022 Sample Size Rationale

When it comes to prefilled syringe and auto-injector testing there is a fine balancing act required in determining the minimum number of test samples and acceptable levels of quality and risk.

Device testing can be performed as either an Attribute test or a Variable test. Just because numerical data is gathered this does not automatically mean it is a variable test. If numerical data is gathered and this data is compared to a set specification this is an Attribute test (Pass / Fail). For variable tests, the data is analysed to generate a k-value to determine if the specification is met.

Generally for Attribute testing there will be a defined number of allowable failures based on a binominal distribution (a distribution with two possible outcomes, in this case Pass or Fail). The allowable number of failures determines the probability content, if you only consider systematic failures then the probability of a single sample failing is 50%. This means it is possible for every second sample to fail. The risk category plays heavily into statistical failures with values that can range from 1% to 5%. To establish sample numbers (according to ISO 11608-1:2022) for Attribute tests, the probability content (the probability of failure) will need to be determined, via assessment of risk.

For Variable testing with a normal distribution typically uses a confidence interval of 95%. ISO 11608-1:2022 recommends using a sample size of n = 20 for a 95% probability content and n = 30 for a 97.5% probability content (both with a 95% probability interval). If the data is not normally distributed then the required statistical calculations should be performed to justify the sample size. If the sample size differs from the recommended numbers several factors need to be considered. Firstly, justification for the differing sample size needs to be representative of the device design. Also, with lower sample numbers calculated uncertainties are much high than higher sample sizes. Using a smaller sample size will in turn mean using a higher k-value, which is more difficult to achieve. Larger samples sizes can reduce any small variances within the data, but the disadvantage to larger samples sizes is the potential of considerably higher costs coming from manufacturing materials and the cost of testing.

ISO 11608-1:2022 provides the basic information to determine the required sample sizes for your device. There are advantages and disadvantages to using samples sizes that differ from the recommended values, in general if the samples size has been risk assessed and can be justified any can be used.

Is ISO 18562 Testing Required for Metered Dose Inhalers (MDIs)?

We all know that inhalers are used intermittently to deliver specific drugs to a specific location. Does that make them a medicinal product or a medical device?

The EU regulations are crystal clear on this. Prefilled inhalers are cited as an example of an Integral Combination Device. That is the medicinal product and the medical device form a single integrated product. They are absolutely intended and designed to be used together and exclusively in this combination.

This means that they need to follow both EU pharmaceutical legislation and the MDR. To put it another way, the combination must have both a marketing authorisation and a CE mark.

This makes the inhaler a ‘breathing system component’ and hence subject to ISO 18562.

Its contact time on each usage is of course very short, just a second or two. But, if it used four times a day over 25 years its cumulative contact time would be 20 hours. That is short term use in the eyes of the standard. Go up to 30 years and it would be a long term device.

The biocompatibility requirements for a breathing component cover particles, gases, and leachates. Particulate testing (ISO 18562-2) testing is most relevant when the inhaler is removed from its packaging. The situation will change after it has been in someone’s pocket for a few days. The gas testing (ISO 18562-3 Test for emissions of volatile organic compounds (VOCs)) is again most relevant on a new device, but the profile may change over its lifetime. Finally, leachate testing (ISO18562-4) applies only if there is an aqueous fluid path from the device to the patient. This is probably not the case for inhalers. Dry powder inhalers clearly no, gas propelled inhalers if they have a fluid connection it is not aqueous. But, ISO 10993 does come into play here. There is transient, mucosal membrane contact when the inhaler or expansion chamber touches the patient’s lips. This should not require specific testing as long as the materials used meet the toxicity requirements of a Biological Evaluation Plan (BEP) and chemical characterisation. There is cumulative contact but if the transfer of materials of concern is negligible to zero with each contact it remains neglible over time.

The conclusion is that MDIs require as a minimum both particulate and VOC testing.

Are human factors studies required for combination devices?

Drug delivery combination products can be convenient, accurate and targeted. They can save time in clinics and help avoid clinical errors. They are also subject to in use errors. Human factors studies are required to assess the effectiveness of instructions and markings, the handling and application of the device, in summary the scope for avoidable errors. Studies should be designed to test the representative user interaction, in the possible environments of use and use the information garnered to deliver safe and effective outcomes.

How are combination products regulated?

EMA draft guidance (EMA/CHMP/QWP/BWP/259165/2019) was published for consultation in the summer of 2019 with a planned final publication date of 26 May 2020.

The American Food and Drug Administration (FDA) applies the pre-market review of combination products through its Office of Combination Products (OCP). Here medical a device is not necessarily included in the combination product. It could comprise two therapeutic materials (drug/biologic).

All regulatory authorities will want to see evidence drug dosages are accurately administered from a device which is not harmful and is stable throughout its life


FAQs home < Previous FAQ