How to Set a Containment Performance Target Before SMEPAC Testing

Setting acceptance criteria after SMEPAC data has already been collected is one of the more damaging sequencing errors a project can make. The data may look numerically reasonable, but without a pre-approved containment performance target, there is no objective benchmark that makes any pass or fail judgment defensible — and a protocol that cannot support a formal containment conclusion cannot support regulatory review either. The downstream cost is not just a retest; it is a structural mismatch between the conditions under which data was collected and the acceptance logic applied to it, which may require an entirely new test campaign under corrected conditions. What resolves this is agreement on the containment performance target before sampling design begins, because that single number drives every subsequent technical decision in the protocol.

CPT Inputs Required Before Sampling Design

The containment performance target is not a value that can be estimated in hindsight and applied retroactively. It is derived from health-based reference values — typically an occupational exposure limit, an acceptable daily exposure, or a banded limit — and the derivation must be complete before surrogate selection, sample volume calculation, or analytical method validation begins. According to ISPE SMEPAC guidance, safety factors for mandatory PPE can be incorporated into the CPT, which means the target already reflects operational context, not just a raw toxicological limit.

A common planning figure from SMEPAC guidance is to set the CPT as low as one-tenth of the lowest OEL of any API the facility will process, specifically when three or more test iterations are planned. That 0.1× figure is a design criterion tied to iteration count, not a universal regulatory minimum, and it should be confirmed against the specific test campaign structure before it is written into the protocol. One source of negotiation risk at this stage is OEL variability: when two organizations derive OELs for the same compound independently, a factor-of-3 difference between sources is generally treated as acceptable within SMEPAC practice, while a factor-of-10 difference requires documented justification. If internal and supplier OELs diverge beyond that range without documented resolution, the CPT itself becomes contestable before the first sample is taken.

Surrogate selection cannot be separated from CPT stringency. For OEB5 targets, naproxen sodium is commonly used because its detection limit falls in the range of 0.05–0.2 ng, which is necessary to measure at very low concentrations. Lactose and mannitol, with detection limits in the range of 1–2.5 ng, are appropriate for lower-band applications where the CPT is less demanding. Choosing a surrogate without first confirming that its assay sensitivity is compatible with the target detection limit produces a sampling plan that is analytically mismatched to the goal.

One boundary that must be clear before any CPT is finalized: SMEPAC applies only to powders. Liquids, vapors, and gases are outside scope, and the methodology does not account for respiratory protection. Respiratory PPE must be managed through a separate exposure control logic, and non-powder processes require different assessment methods entirely.

How the Target Changes Method Sensitivity and Sample Volume

Once the CPT is fixed, it functions as a design input for every analytical and logistical decision in the protocol. Sample volume, collection time, and analytical method sensitivity are all calculated backward from the target — which means a CPT set too low without sufficient analytical justification does not simply require a more sensitive method; it may require a method that does not yet exist at validated scale.

A practical illustration from SMEPAC planning: if an API is processed at 100% concentration and the CPT is set at 10 µg/m³, then processing a 50% blend of the same API proportionally relaxes the target to 20 µg/m³. That adjustment is not automatic — it requires explicit documentation of the formulation basis and agreement that the adjusted target still protects against the relevant exposure risk. The value here is that formulation dilution can legitimately reduce the analytical burden, but only if the CPT adjustment is approved before sampling begins, not applied after results are in.

The dependency on process conditions is stricter than it may appear. SMEPAC results are specific to the API amount processed, the equipment used, and the duration of the activity. A test conducted with a 200g charge under a 10-minute operation does not transfer to a 500g charge over 30 minutes — even if the same equipment is involved. This is not a margin issue; it is a conditions-of-applicability issue. If the test conditions deviate from the CPT basis on which the protocol was designed, the resulting data is unusable for the intended containment conclusion, not merely less precise. Teams that treat these parameters as adjustable after the fact are generating data that cannot be defended as representative of the production process.

For equipment operating in OEB4 or OEB5 bands, where CPTs are most demanding, this dependency on conditions is also the primary reason why surrogate powder testing methods must be selected and validated in direct response to the approved CPT, not chosen generically and applied to whatever target is eventually agreed upon.

Protocol Failure When Acceptance Logic Is Chosen After Testing

The failure mode here is specific: a team completes a SMEPAC campaign, reviews the data, and then begins constructing acceptance criteria around what the numbers show. This produces criteria that the data appears to satisfy, but the process is inverted. Without a CPT defined and approved before the test, the result cannot be characterized as a pass against an objective health-protective standard — it can only be characterized as a measurement. Those are not the same thing for regulatory or QA purposes.

The second failure mode compounds the first. If the actual production conditions — API charge, equipment configuration, activity duration — differ from the conditions under which the SMEPAC test was run, then any acceptance logic derived from that test does not apply to the real process. The test results describe a challenge scenario that does not match what operators will actually encounter. Applying post-hoc acceptance logic to mismatched conditions does not recover the validity of the data; it adds another layer of unjustifiable reasoning to an already structurally flawed conclusion.

Both of these failure modes are avoidable. They require only that the CPT is approved in writing before the test begins, and that the tested operation — API amount, equipment, and duration — exactly mirrors the intended production parameters. The entire team, including toxicology, EHS, QA, and the test provider, must agree that the target is fixed before any sampling takes place.

Tradeoff Between Conservative Targets and Practical Equipment Design

Pushing the CPT lower than the documented health risk justifies introduces real engineering and procurement consequences that are often underweighted at the planning stage. A very low CPT combined with a correspondingly low limit of quantitation creates measurement burdens that are not proportional to the protective benefit: sampling becomes more cumbersome, collection times lengthen, and analytical methods require validation at sensitivity levels that may not be routinely available. In some cases, mandatory PPE remains required even when the equipment is specifically designed to provide engineered containment — meaning the operational burden of a conservative target is not offset by reduced PPE reliance.

The more significant consequence involves containment design itself. Applying large safety factors to derive a stricter CPT can trigger equipment design changes — additional containment features, tighter pressure cascades, modified transfer systems — that the actual process risk does not justify. Those changes are not without cost, and they are difficult to reverse once a supplier has begun detailed design or fabrication. For isolator and cRABS procurement in OEB4/OEB5 applications, where containment architecture is already highly engineered, an overstated CPT can produce specification requirements that exceed what the hazard demands, without improving operator protection in any measurable way.

The balancing principle is straightforward to state but requires cross-functional discipline to implement: CPT stringency should be proportional to documented process risk, iteration count, and operational practicability. A CPT set lower than the risk warrants is not inherently more protective if the analytical infrastructure cannot reliably measure at that level or if the equipment design changes it triggers cannot be practically validated. Neither is it a reason to weaken health-based limits — it is a reason to match the target to the actual exposure basis and document that alignment before equipment design begins.

Approval Gate for CPT, Test Operation and Retest Rule

The CPT approval gate is not a single-function review. It is the point at which four interdependent elements must be locked simultaneously: the target itself, the tested operation parameters, the analytical method with its validated detection limit, and the retest rule. Releasing the protocol before all four are agreed renders the test campaign structurally incomplete, regardless of how well the sampling is executed.

Embedding the CPT in the User Requirements Specification is the most direct way to make this gate enforceable. When the CPT is defined in the URS, suppliers can design to a containment performance limit that demonstrably stays below the target during both initial and recurrent SMEPAC tests. This creates a traceable link between the health-based design input and the equipment performance specification — a link that supports inspection readiness and qualification documentation. Without it, the supplier is designing to an unvalidated containment assumption, and the CPT becomes something that must be negotiated retrospectively against delivered equipment rather than verified against a committed design criterion. For OEB4/OEB5-Isolatoren where containment performance is a primary specification driver, the URS is where that commitment is established before fabrication begins.

One gap in current SMEPAC guidance that teams must address internally is the absence of a specified retest interval. The guidance acknowledges that containment equipment degrades or fails over time and that periodic retesting is necessary, but it does not define how frequently that testing must occur. This is a process detail gap, not a compliance deficiency in the standard, but it creates a real operational risk if teams treat the initial SMEPAC result as indefinitely valid. The retest rule — interval, trigger conditions, and what constitutes a recurrence of the approval gate — should be defined in the protocol before the initial test is run. Defining it after a potential failure has occurred is a significantly harder position to defend.

For a broader view of how retesting frequency interacts with OEB classification and monitoring strategy, the discussion of real-time containment monitoring versus annual SMEPAC testing provides useful framing for setting an internal retest rule where the standard is silent.

The most concrete implication across these planning steps is that the CPT functions as a design input at every stage — surrogate selection, sample volume, analytical method, equipment specification, and retest interval all depend on a value that must be approved before any of those decisions are made. A target confirmed after the fact cannot be retrofitted into a protocol that was already structured around different assumptions, and data collected under those conditions cannot reliably support a formal containment conclusion.

Before releasing a SMEPAC protocol, the questions worth confirming are: Is the CPT documented in the URS and agreed by toxicology, EHS, and QA? Does the analytical method’s limit of quantitation match the target? Do the test operation parameters — API charge, equipment configuration, activity duration — exactly mirror the intended production process? And has a retest interval been defined internally, even though the current guidance does not specify one? If any of those items remains open, the protocol is not ready to release, and the risk of structural data failure is already present before the first sample is collected.

Häufig gestellte Fragen

Q: Does the CPT framework apply if the facility has not yet finalized which APIs will be processed?
A: The protocol cannot be responsibly released in that state. The CPT is derived from the lowest OEL of any API the facility will process, so an unresolved API list means the numeric basis for the target is still open. The practical resolution is to identify a worst-case API — the one with the lowest expected OEL — and design the CPT to that compound, with a documented assumption that any future API added to the scope will be evaluated against the same threshold before its campaign begins.

Q: Once the initial SMEPAC test passes, what is the next decision the team must make before the equipment enters routine production?
A: The immediate next decision is defining the retest interval and trigger conditions internally, since current SMEPAC guidance does not specify a required frequency. That interval should be documented in the protocol before the initial test is run, not assigned afterward. Relevant inputs for setting it include equipment age, frequency of high-potency operations, any observable changes in pressure cascade performance, and whether real-time containment monitoring is in place between formal SMEPAC campaigns.

Q: At what point does a CPT become too conservative to be practically defensible?
A: When the analytical infrastructure cannot reliably measure at the required detection limit, or when the equipment design changes triggered by the target exceed what the documented process risk justifies. A CPT that mandates a surrogate assay sensitivity not routinely available at validated scale, or that forces containment architecture changes disproportionate to the actual OEL-based hazard, is no longer proportional to its protective purpose. The threshold is not a fixed number — it is the point where the measurement burden and design cost cannot be justified against a documented health-based rationale.

Q: How does the CPT-setting process differ when an OEL for the compound has not been formally established by the facility’s toxicology team?
A: Occupational exposure banding — such as the NIOSH banding framework — provides a structured alternative for assigning a provisional OEL when compound-specific data is insufficient for a full toxicological derivation. The banded limit can serve as the CPT basis, but this requires documented agreement from toxicology and EHS that the band assignment is appropriate for the compound’s hazard profile. A banded CPT carries more inherent uncertainty than a compound-specific OEL, which is a reason to apply more conservative safety factors rather than fewer, and to flag the basis explicitly in the URS so it can be revisited if better data becomes available.

Q: Is a single SMEPAC campaign sufficient to support CPT qualification across multiple similar isolator units of the same model?
A: Not without documented justification for the equivalence claim. SMEPAC results are specific to the equipment tested, the API charge, and the activity duration. Extending a single campaign’s conclusion to nominally identical units requires, at minimum, a documented risk assessment confirming that the units share the same containment-critical design parameters and have been installed and maintained under equivalent conditions. Where containment performance is a primary specification driver — such as in OEB4/OEB5 isolators — regulators and QA reviewers will expect either individual test data or a formally justified bracketing strategy, not a blanket extrapolation from a single test campaign.