Media fill bracketing approach: three words that appear more and more often in Aseptic Process Simulation (APS) protocols and, unfortunately, in inspection findings too. Anyone manufacturing sterile products with multiple container formats, line speeds and presentations knows that simulating every combination at every six-monthly requalification is materially impossible. Bracketing — and its more structured cousin, the matrix approach — is the regulatory answer to this problem: simulate the extremes while documenting why they cover everything in between. The key word is "documenting": bracketing without a written rationale is one of the fastest ways to turn a compliant media fill into a deviation.
What the media fill bracketing approach is
The media fill bracketing approach means selecting, within a family of aseptic process configurations, the worst case combinations to challenge in the APS, assuming — with scientific justification — that success at the extremes also covers the intermediate configurations. The classic example is container formats: if a line fills 2R, 10R and 50R vials with the same equipment train and the same types of intervention, the protocol can alternate the largest and the smallest format instead of simulating all three in every cycle.
The matrix approach extends the logic to several variables at once (format, speed, container/closure combination, shift), distributing worst-case conditions across different runs within the APS programme. The practical distinction: bracketing works on the extremes of one variable, while a matrix combines several variables into a multi-run plan that, taken as a whole, must cover the entire operating space.
What Annex 1 and PIC/S PI 007-6 say
EU GMP Annex 1 (2022 revision) dedicates paragraphs 9.32 to 9.49 to the APS. The core principle is that the simulation must imitate the routine aseptic process as closely as possible, including all critical steps and known interventions (par. 9.33–9.36), and must never be used to justify practices that add contamination risk (par. 9.35). Bracketing and matrix approaches for container/closure combinations are considered acceptable only where process equivalence between configurations is scientifically justified: worst cases must be selected on the basis of process impact — for example the vial with the widest opening at the slowest speed, because it maximises exposure time in Grade A.
PIC/S recommendation PI 007-6 is even more explicit operationally: simulations should use appropriate combinations of container size, opening and processing line speed, preferably at the extremes (§5.1.1), and should represent a worst-case situation (§5.1.2). The document reminds us that the worst case is not always the largest container: small ampoules run at the highest speed can be more critical, because of more frequent jams and therefore more interventions (§5.1.3). On frequency, the references converge: initial qualification with three consecutive satisfactory simulations per shift (PI 007-6 §5.5.3) and repetition at least twice a year for each process, line and shift (§5.5.6), with Annex 1 adding that every operator should take part in at least one successful APS per year, plus stricter requirements for manual operations (par. 9.39).
Topics like this one — APS, Annex 1, sterility assurance — are the daily bread of The Pragmatic GMP, GuideGxP's free weekly newsletter: every week one GMP topic broken down into operational, audit-ready decisions. Subscribe here.
How to select worst cases: formats, speeds and interventions
Worst-case selection is the technical heart of bracketing, and it must be built variable by variable, not by gut feeling. The questions to ask: which configuration maximises product exposure to the environment? Which generates the most Grade A interventions? Which puts operators under the greatest strain?
| Variable | Typical worst case | Rationale |
|---|---|---|
| Container format | The largest and the smallest in the range | The large one has the widest opening and the slowest fill (maximum exposure); the small one runs at top speed, with more jams and interventions |
| Line speed | Low and high extremes of the operating range | Minimum speed extends dwell time in the critical zone; maximum speed increases the frequency of corrective interventions |
| Container/closure combination | The combination with the most manipulations | Every additional component (stopper, cap, prefilled syringe) introduces its own aseptic steps |
| Number of formats | An additional intermediate format where there are more than three | With wide ranges, the extremes alone do not adequately represent the central configurations |
| Duration and shifts | End of campaign, shift changeover, breaks | Operator fatigue and end-of-production activities are known unfavourable conditions |
Beware of a widespread conceptual error: piling every extreme into a single run. A simulation that combines every unfavourable condition at once may no longer represent any real process, and is as hard to defend as unjustified bracketing. It is better to distribute worst cases across different runs of the programme, ensuring that each significant variable is challenged at defined intervals.
Building a defensible APS matrix
A bracketing/matrix programme stands up in inspection when the protocol makes three elements explicit. First: the complete inventory of configurations — every authorised product/container/closure/line/shift combination must appear in the matrix, even when it is not simulated directly. Second: the equivalence rationale — which process elements are common to all configurations (sterilisation train, material entry, setup) and which are format-specific; only genuinely equivalent elements can be "covered" by a run on a different format. Third: the rotation rule — how often each extreme configuration comes back into simulation and where the remaining variables are challenged.
The matrix must also be kept alive by change control: a new format, a line modification or a new container/closure combination falling outside the justified perimeter requires a new assessment and, where necessary, an additional APS before routine production. And remember PI 007-6's minimum numbers: for batches below 3,000 units the APS should equal the commercial batch size (§5.1.5), and the acceptance criteria (§6.2) target zero contaminated units, with an investigation already required at the first positive unit.
Typical findings raised in audits
- Bracketing declared but not justified: the protocol alternates formats without a written equivalence rationale — the most frequent finding.
- Worst case chosen once and for all: the rationale is never reviewed after changes to line, speed or product mix.
- Orphan configurations: authorised combinations that appear neither in the runs nor in the matrix justification.
- All extremes at once: a single "super worst case" run that represents no real process.
- Misaligned interventions: the list of simulated interventions does not reflect those actually recorded for the format under test.
GuideGxP recommendation
Treat bracketing as a risk assessment deliverable, not a scheduling shortcut. Start from the configuration inventory, rank the variables by impact on aseptic exposure, write the equivalence rationale format by format and define the rotation in the APS master protocol — before an inspector asks for it. And at every six-monthly requalification, check that yesterday's worst case is still today's: intervention data and environmental monitoring results are the best evidence to confirm or disprove it.
To design, execute and defend your APS programme — bracketing, matrices, interventions, incubation and failure management — see the GuideGxP operational guide Aseptic Process Simulation (Media Fill): design, execute and defend your APS per Annex 1 and FDA, with a ready-to-use Excel/Word toolkit.
