Single-Use Infusion Sets: Where GB 8368 Applies and Where It Does Not

The hard part is not the number of tests, it is inconsistent scoping

A single-use infusion set is a textbook case of "simple structure, complicated testing". It is assembled from a closure-piercing device, an air-inlet component, a drip chamber, tubing, a fluid filter, a flow regulator, an injection site and conical fittings. Taken individually, none of these parts is difficult to make; together they have to satisfy flow, sealing, cleanliness, sterility and biological safety requirements on one and the same fluid path, so a process deviation anywhere gets amplified at the test bench.

When we take on infusion set projects, the deficiency letters and requests for additional information are rarely about a missing test. Most registration test lists are copied from peers and are broadly complete. The real problems cluster into three types: the wrong standard was chosen for the device type; the samples submitted do not cover the configurations being registered; and the test report does not line up with the product technical requirements. This article works through these in the order they actually arise. Specific parameters and acceptance limits referred to below are governed by the currently effective version of the standard text.

Sort the device type first, then sort out what each standard governs

Before discussing any test list, confirm that the product in hand is the type GB 8368 governs. GB 8368 covers gravity-feed single-use infusion sets - sets in which flow is driven by the height difference of the fluid column, used in conventional infusion scenarios. Sets with fine filtration, light protection, pressure infusion, or other special-purpose functional structures each have a corresponding industry standard within the system, and cannot all be forced under GB 8368.

The way to decide is to look at the driving mechanism first and the functional components second. If the drive is not gravity, or if the product carries a functional structure GB 8368 does not address - fine filtration media, a light-protective layer, tubing intended to carry infusion pump pressure - go back to the standard that corresponds to that type and confirm the scope and item set there. Which industry standard applies and which edition governs is determined by the currently effective version of the standard text and by the standards catalogue published by the competent authority; applying one from memory is not advisable. Getting this step wrong means redoing everything downstream: once the wrong reference standard is written on the test request, the report is no longer fit for its purpose - and this is typically discovered on the day of sample submission, while the request details are being checked.

Once the type is confirmed, the remaining references fall into three lines with clear boundaries. A lot of technical documentation reads badly because product standards, method standards and evaluation-framework standards get cited as if they were the same kind of thing.

The point of the table is this: the product standard says what the product has to achieve, the method standard says how to measure it, and the evaluation framework says how to judge risk once the data are in. Technical requirements that cite only the product standard without stating where the methods come from, or that pile up method standards with no evaluation conclusion, leave the reviewer unable to see where the data came from.

Flow and drop count: the functional items unique to infusion sets

Flow and drop count are functional items specific to infusion sets, and they are also the data set most likely to disagree on retest. Flow is not determined by the tubing alone; it is the combined result of fluid column height, tubing bore and length, the resistance of the fluid filter, and the state of the flow regulator. The relationship between drop count and delivered volume depends on the geometry of the drip tube and on the surface tension of the test liquid.

Several practical consequences follow. First, if the test rig is built to a different reading of the standard - a different reference point for liquid level, a regulator not set fully open, tubing that is coiled or kinked - the measured flow will deviate systematically, and nothing in the reported figures will reveal the source of that deviation. Second, the clamping area of the flow regulator compresses the tubing during storage, so flow measured after prolonged clamping differs from flow measured on freshly assembled samples; the storage condition of the samples therefore needs to be stated on the request form. Third, once the fluid filter is partially blocked by particles introduced during extraction or rinsing, flow drops noticeably, so flow testing should not reuse samples that have already been through destructive tests.

When submitting samples, state on the request form how the test rig is to be built and what condition the samples are in, and retain untouched reserve samples from the same batch so that anomalous results can be reproduced. Producing a fresh batch for retest is a weak answer to a deficiency letter, because the new batch is no longer the batch the original report covers.

Fluid filter and particulate matter: retention capability and self-cleanliness are two different things

The fluid filter and particulate contamination are often treated as one topic; in fact they sit at two different levels. The filter governs retention capability - whether particles in the expected size range are held back on the far side from the patient. Particulate contamination governs self-cleanliness - whether the set releases particles into the fluid during use. The first is a function, the second is a cleanliness attribute; their failure mechanisms are entirely different, and the technical requirements should express them separately.

Particle sources are not hard to locate in engineering terms: flash and demoulding debris on injection-moulded parts, burrs on cut tubing ends, adhesive fragments from solvent bonding, airborne dust in the assembly environment, fibres from packaging material. Most of these relate to cleanroom management and mould condition rather than to resin grade, which is why changing the resin rarely fixes a failing particulate result. The investigation belongs with mould maintenance, cutting process and assembly environment.

On the testing side, watch the sampling and rinsing scope. Particulate testing is sensitive to the working environment and to the background of the glassware; if the laboratory has not run or not recorded a background check, there is no way to tell whether excess particles came from the product or from the environment, and the report should show the background data. On the filter side, check whether the sample is a complete fluid path. Submitting only the filter assembly rather than the whole set yields a conclusion that cannot be applied directly to the technical requirements of the complete set, and that is hard to defend in the submission.

Closure-piercing device and protective cap: sharpness and sealing pull in opposite directions

The closure-piercing device has to do two conflicting things in a single piercing action: pierce the closure cleanly, and then form a reliable seal against it without generating debris. Taper, edge geometry and material stiffness together determine that balance. Too sharp, and it shaves rubber particles off the closure, becoming one more particle source. Too blunt, and piercing resistance rises, which in use shows up as closure deformation, a bent piercing device, or even fracture.

The air-inlet component and the protective cap are the two matching details. An air inlet with an air filter has to admit air freely while blocking environmental microorganisms, and changes in the state of its filter medium during ethylene oxide sterilisation and aeration are easily overlooked. The protective cap determines the sterile condition of the product before use: a loose fit lets the cap come off during distribution, while too tight a fit deforms the piercing device when the cap is removed. Neither part looks significant on a drawing, yet between them they drive three classes of conclusion - sterility, particulate matter and usability.

The testing recommendation is to run piercing-related items against closure material matching clinical use; results obtained on a substitute material cannot be extrapolated directly to the real combination. Read the post-piercing debris observation together with the particulate results: when both are out of specification at the same time, they usually point to a single process cause, and investigating them separately wastes time.

Tubing tensile strength and connection leakage: the load path and the seal path overlap

What follows is engineering analysis based on structure and materials rather than a statistical conclusion, offered as a direction for development troubleshooting.

Sealing failures in an infusion set are usually not in the tubing itself but at a few interfaces: the fit between piercing device and closure, the bonded interface between tubing and rigid components, and the taper fit between the conical fitting and external devices. The tubing is typically plasticised polyvinyl chloride or a thermoplastic elastomer; the rigid parts are typically polypropylene or polycarbonate, with very different moduli. Assembly forms the seal through solvent bonding or an interference fit, and that interface carries both axial tensile load and radial sealing duty at once. The load path and the seal path overlap, so damage in one place causes both functions to fail together - which is why tensile and leakage problems so often appear at the same time.

Several failure scenarios follow from this. First, if the solvent in a bonded joint has not fully evaporated before packaging and sterilisation, residual solvent continues to soften the substrate, interface strength declines with storage time, and tensile and leakage items that pass on initial samples fail on aged samples. Second, excessive interference improves initial sealing but leaves sustained hoop stress in the rigid part; polycarbonate-type materials under residual stress combined with sterilant exposure are prone to stress cracking, which shows up as seepage after a period in storage. Third, demoulding marks or flash on the taper of a conical fitting create local leak paths that may not appear under positive pressure and only reveal themselves under negative pressure.

The testing implication is that leakage and tensile items must be run on samples that have been through the complete process chain, including sterilisation and packaging, not on semi-finished parts taken off the line. Within one sample set, cover different mould cavities rather than drawing everything from a single cavity; otherwise cavity-to-cavity variation is averaged out and only surfaces once volume production starts.

Chemistry and biology: conclusions here, methodology in a separate article

Chemical items lose marks in the extraction step almost every time, not in the instrumental analysis. Extraction medium, how the sample is taken, the relationship between liquid volume and sample, and the combination of temperature and time - unless the request form states which set was chosen, data from different laboratories are simply not comparable. How to choose extraction conditions, and how the chemical line connects to the biological line, is generic methodology, and we deal with it in a dedicated article on the GB/T 14233 methods, available in the technical knowledge base.

Ethylene oxide residue has one peculiarity in infusion sets: the complete set is assembled from parts that aerate at very different rates. The piercing device protective cap, the drip chamber, the tubing and the filter medium cannot aerate in step, so sampling the tubing alone does not represent the whole set. Residue levels are also tied directly to the aeration process after sterilisation - change the ventilation of the aeration room or the stacking pattern and the result changes. Residue should therefore be read together with sterilisation process validation rather than treated as an isolated chemical item; see our sterilisation validation service.

The same applies on the biological side, and the conclusion comes first: an infusion set is an externally communicating device forming a path to circulating blood, and the nature of that contact sets the starting point of the evaluation. Following the logic of ISO 10993-1, first determine the device category and contact duration category, then judge which endpoints require test data and which can be argued from existing information, and only then arrive at specific items. Blood-contact endpoints are run to the corresponding parts of the ISO 10993 series; which part applies and which edition governs is determined by the currently effective version of the standard text and by the reasoning in the evaluation route. Where cytotoxicity is run along the ISO 10993-5 route, failures more often originate in the process than in the formulation - incomplete ethylene oxide aeration, migration of printing ink additives, residual bonding solvent, use of reground material. None of these is usually solved by switching resin grade. For how the biological work is organised overall, see our biocompatibility testing service.

The sterile barrier and shelf life are not "the next phase"

Many manufacturers schedule packaging and shelf-life studies after product testing, and only discover at the point of assembling the submission that the two do not match. Once the sterile barrier system of an infusion set changes - a new supplier for the paper-plastic pouch, or an adjustment to the sealing process - the product testing already completed strictly needs its applicability reassessed, because sterilisation and ageing act on the product through the packaging.

Under the ISO 11607 framework, packaging has to demonstrate sterile barrier integrity and the ability to maintain it throughout shelf life, including through distribution. Under the ISO 11135 framework, ethylene oxide sterilisation validation treats product, packaging, loading pattern and aeration conditions as one system. For infusion sets these two lines mesh tightly: packaging permeability affects aeration rate, loading density affects residue level, and residue level in turn affects cytotoxicity results. Put all three on the same timeline when the project is scheduled; our packaging validation service page sets out the scope.

Common deficiency types and the self-checks that prevent them

What to prepare before submitting samples

Sample quantity needs headroom for items running in parallel: physical, chemical and biological items generally cannot share the same units, and destructive items need retest reserves as well. Samples should be finished product that has been through the complete sterilisation and packaging process, not work in progress taken off the line. These generic requirements are the same as for syringes, catheters and other sterile infusion-related products, so we will not repeat them here.

What is worth supplying alongside the samples specifically for infusion sets is: an assembly drawing of the complete set with the material grade of each part marked; a description of the drip chamber and drip tube geometry; information on the fluid filter medium; the bonding process and the solvent used; the sterilisation method and aeration conditions; the packaging format and sealing process; and a draft of the product technical requirements. The draft technical requirements matter most - aligning them with the test plan before samples ship avoids the rework that follows when the report comes back and item names and wording do not match the technical requirements.

What SUNGO Lab can do

SUNGO Lab (Shanghai Shage Medical Technology Co., Ltd.) operates laboratories in Shanghai and Hefei and is accredited by CNAS, CMA and IAS (USA). We handle physical performance, chemical performance, biological evaluation, ethylene oxide sterilisation residue, packaging and shelf-life testing and validation for single-use infusion sets and comparable sterile infusion devices, and at the planning stage we can help establish the relationship between device type and reference standard and align typicality of sampling, choice of test conditions and the product technical requirements. To be clear, an accreditation mark only demonstrates that the laboratory holds the corresponding technical competence within its accredited scope; it does not constitute a commitment as to market access outcomes.

If you are preparing registration testing for an infusion set or responding to a deficiency letter, bring your structural drawings and the draft technical requirements and we will run a submission feasibility review first, so that device type and reference standard are settled before anything else moves. Call +86 132 4819 8029 or request a quote.